VOL. 47, NO 3-4, SEPTEMBER-DECEMBER 2004

ISSN 0031-1480

VOL. 47, NO 3-4, SEPTEMBER-DECEMBER 2004

Medical Society of Papua New Guinea

Executive 2004

President: Mathias SapuriVice-President: Nicholas MannSecretary: Sylvester LaheTreasurer: Harry AigeelengExecutive Member: Brunie Christian-Dangga

Published quarterly by the Medical Society of Papua New Guinea

Papua New Guinea Medical Journal

ISSN 0031-1480

September-December 2004, Volume 47, Number 3-4

EDITORS: PETER M. SIBA, JOHN REEDER, NAKAPI TEFUARANI

Editorial Committee

I. Kevau G. Mola A. Saweri

Editorial Assistant: Cynthea LeahyEmeritus Editor: Michael Alpers

Email: [email protected] page: http://www.pngimr.org.pg

Registered at GPO, Port Moresby for transmission by Post as a Qualified Publication.

Printed by PNG Printing for the Medical Society of Papua New Guinea.

Authors preparing manuscripts for publication in the Journal should consult ‘Information for Authors’ inside back cover.

Papua New Guinea Medical Journal Volume 47, Number 3-4, September-December 2004

EDITORIAL Malaria epidemics in the highlands of Papua New Guinea L. Dapeng

ORIGINAL ARTICLESThe management of children with cancer in Papua New Guinea: a review of

children with cancer at Port Moresby General Hospital M. Kiromat, J.D. Vince, G. Oswyn and N. Tefuarani

Zinc in human health V.J. Temple and A. Masta

The epidemiology of malaria in the Papua New Guinea highlands: 3. SimbuProvince I. Mueller, J. Kundi, S. Bjorge, P. Namuigi, G. Saleu, I.D. Riley and J.C.Reeder

A comparison of booked and unbooked mothers delivering at the Port MoresbyGeneral hospital: a case-control study F. Failing, P. Ripa, N. Tefuarani and J.Vince

Strongyloides fuelleborni kellyi and other intestinal helminths in children fromPapua New Guinea: associations with nutritional status and socioeconomicfactors S.E. King and C.G.N. Mascie-Taylor

How the poor die in the settlements of Port Moresby, 2003-2004 P. Sims

Antenatal care in Goroka: issues and perceptions G.L. Larsen, S. Lupiwa, H.Paito Kave, S. Gillieatt and M.P. Alpers

Rehabilitation services in Papua New Guinea A. Shaw

MEDICAL RESEARCH PROJECTS IN PAPUA NEW GUINEA

MEDLARS BIBLIOGRAPHY

CONTENTS

135

138

146

159

174

181

192

202

215

228

231

PNG Med J 2004 Sep-Dec;47(3-4):

135

EDITORIAL

Malaria epidemics in the highlands of Papua New Guinea

135-137

The highlands region of Papua NewGuinea (PNG) comprises over 15% of thetotal land mass of the country and about 2million people (over one-third of the nationalpopulation) in the five provinces, namelyEastern Highlands, Simbu, WesternHighlands, Enga and Southern HighlandsProvinces. The elevation of the habitablearea ranges from 800 metres to 3000 metresabove sea level. The rainy season is notclear-cut, but heavy precipitation isexperienced during the December-Marchperiod. Before the 1950s, malariatransmission was unusual in most parts ofthe highlands, particularly above 1200metres. Malaria extended into new areas,along with the opening-up of the highlandsfor economic development, especially afterthe completion of the road link. Today, thehighlands provinces are experiencingincreasing outbreaks of malaria, after almostcomplete cessation of control activities sincethe late 1980s. Malaria epidemics in thehighlands are characterized by relativelysmall, often isolated outbreaks that takentogether constitute periodic epidemics,resulting in large numbers of deaths in all agegroups among the non-immune population,often in the late rainy and early dry seasons.This is fundamentally different from malariain the coastal and island areas, where thereis constant transmission throughout the year,and, as the population is partially immune,outbreaks are rare and their victims are mostoften young children.

During 1995 and early 1996, more than1000 malaria deaths and over 10,000hospitalizations were reported from thehighlands. From February to April 2004 therewas a malaria epidemic in the Omarua area,Arona Valley, Eastern Highlands Province,where more than 153 people within apopulation of about 13,000 reportedly diedin eight villages. During the epidemic, theauthor tested 179 villagers with the rapid

diagnostic Malaria Cassette Test (from ICTDiagnostics, South Africa) in Omarua village,in the centre of the epidemic area; 116 (65%)of them were positive, including 75Plasmodium falciparum single and 41 mixedinfection (P. falciparum, P. vivax and P.malariae). 65 children aged 2-9 years wereexamined in the village and 53 (82%) of themshowed palpable enlargement of the spleen,suggesting that the epidemic had lasted along time in the village. A malaria epidemicalso occurred in the Kerowagi area of SimbuProvince, affecting around 30,000 people in40 villages. During my visit to Miunde Village,Kerowagi District, Simbu Province in late May2004, one elder said that he and his familymembers had attended 50 funerals from Aprilto May 2004. Similar devastating malariaepidemics were reported to have claimedmore than 100 lives in the Kagua and EraveDistricts of Southern Highlands Provincefrom April to June 2004.

The increased frequency and severity ofrecent outbreaks in the PNG highlands canbe attributed to a number of important factors:1) increasing population movement from thehighlands to the lowlands, to plant andharvest coffee or undertake other jobs wherethey become infected with malaria and bringmalaria back to their own villages wherepeople have little or no immunity to theinfection; 2) climate change (El Niño event)with longer than normal periods of rainresulting in more breeding places andproducing a larger than normal populationof mosquitoes, and therefore a higher thannormal potential for malaria transmission; 3)high prevalence of drug resistance – althoughthe current first-line antimalarial drugrecommendation is chloroquine combinedwith sulfadoxine-pyrimethamine in PapuaNew Guinea, monotherapy with chloroquineis very common in the highlands region andcan result in a fatal outcome in highlanderswith little or no immunity; 4) lack of an

136

Papua New Guinea Medical Journal Volume 47, No 3-4, Sep-Dec 2004

adequate and efficient malaria controlprogram in the highlands – traditionalpreventive measures consisting of regularinsecticide residual house spraying in thehigh-risk areas were suspended in 1989 andthere is no proper diagnosis of patients anda shortfall of adequate supplies ofantimalarial drugs in most rural healthservices of the highlands; 5) inefficiency ofhealth information systems resulting indelays in detecting malaria epidemics –epidemics were often identified through themedia or political administration only after ahigh death toll; 6) law and order issues, whereclan fighting might result in demolished publicservices, such as health centres and schoolbuildings, for example, the recent malariaepidemic in the Omarua area of EasternHighlands Province was largely due to twoclans fighting and their local clinics beingclosed down for 3 years – it takes 2-3 hoursfor a patient in the Omarua area to go byvehicle to other clinics, resulting in a longdelay of treatment and a consequentaccumulation of infection sources, eventuallycausing the devastating epidemic.

Many operational and technical problemsstanding in the way of the implementation ofan efficient and successful malaria controlprogram in the highlands have beenidentified. Most significant among them isthe absence of reliable information on theextent, distribution and location of malariatransmission areas in the highlands. Inresponse to this shortage of information,Mueller and his team at the PNG Institute ofMedical Research, with support from theWorld Health Organization, have conducteda systematic survey of malaria in thehighlands since 2000, using GeographicalInformation System (GIS) technology toanalyze and map malaria risk in the highlandprovinces. Based on the results of theirsurveys a series of articles in this Journaldescribe the results of their studies in theEastern Highlands, Western Highlands andSimbu Provinces (1-3). Similar reports forSouthern Highlands and Enga Provinces willsoon be published in the Journal. This seriesof articles has re-established a valuable

database of the prevalence of malariainfection in all parts of 5 highland provincesin PNG. This is the first study of the malariasituation in the highlands since the classicstudies in the late 1950s (4-6). These articles(1-3) together with previous publicationsdemonstrate that there are complex patternsof malaria epidemiology in terms of diversityin both parasite and vector population in thehighlands of Papua New Guinea (7), and theburden of malaria on the highlands is now atleast as high if not higher than that in thepre-control era (1,3,8). Most importantly,Mueller and his team have innovativelycategorized whole highland regions in PapuaNew Guinea into four zones of ‘areas ofhighly endemic malaria’, ‘areas of moderatelyendemic malaria’, ‘areas of low, unstablemalaria but high risk of epidemic’ and ‘areasof no or very rare local transmission and onlylimited epidemic risk’, by combining thecriteria of prevalence of parasitaemia, spleenenlargement rate, the occurrence ofepidemics in the epidemiological history andaltitude. The beauty of the study is that theyhave delineated the 4 zones graphically byusing GIS technology and categorized allvillages in the highlands into these zones.The articles also make specificrecommendations to health authorities aboutmalaria control strategies for a specific zone,based on the Papua New Guinea nationalmalaria control policy. Therefore, decision-makers and malaria control personnel canuse the recommendations to guide theircontrol activities in the highlands of PapuaNew Guinea.

The series of articles by Mueller and histeam reflect the impressive state of currentresearch on malaria epidemiology, ecologyand malaria control strategy in the highlandsof Papua New Guinea. The time is here totranslate this commendable research effortto guide malaria epidemic control in thehighlands. The national malaria controlprogram, together with provincial healthoffices in the highlands, should make everyeffort to strengthen their malaria controlprogram to improve malaria diagnosis andtreatment in the highlands region and to

137


conduct annual insecticide house-sprayingin all the areas with unstable malaria and highrisk of epidemics as delineated in the articles(1-3). By doing this, we may avoidunacceptable death tolls and suffering dueto malaria epidemics in the highlands ofPapua New Guinea in the future.

Luo Dapeng

World Health OrganizationPO Box 5896Boroko, NCD 111Papua New Guinea

REFERENCES

1 Mueller I, Taime J, Ivivi R, Yala S, Bjorge S, RileyID, Reeder JC. The epidemiology of malaria in thePapua New Guinea highlands: 1. Western HighlandsProvince. PNG Med J 2003:46:16-31.

2 Mueller I, Bjorge S, Poigeno G, Kundi J,Tandrapah T, Riley ID, Reeder JC. The

epidemiology of malaria in the Papua New Guineahighlands: 2. Eastern Highlands Province. PNG MedJ 2003:46:166-179.

3 Mueller I, Kundi J, Bjorge S, Namuigi P, Saleu G,Riley ID, Reeder JC. The epidemiology of malariain the Papua New Guinea highlands: 3. SimbuProvince. PNG Med J 2004:47:159-173.

4 Peters W, Christian SH, Jameson JL. Malaria inthe highlands of Papua and New Guinea. Med JAust 1958:2:409-416.

5 Peters W, Christian SH. Studies on theepidemiology of malaria in New Guinea. IV. Unstablehighland malaria – the clinical picture. Trans R SocTrop Med Hyg 1960:54:529-537.

6 Peters W, Christian SH. Studies on theepidemiology of malaria in New Guinea. V. Unstablehighland malaria – the entomological picture. TransR Soc Trop Med Hyg 1960:54: 537-548.

7 Mueller I, Taime J, Ibam E, Kundi J, Lagog M,Bockarie M, Reeder JC. Complex patterns ofmalaria epidemiology in the highlands region ofPapua New Guinea. PNG Med J 2002:45:200-205.

8 Mueller I, Namuigi P, Kundi J, Ivivi R, TandrapahT, Bjorge S, Reeder JC. Epidemic malaria in thehighlands of Papua New Guinea. Am J Trop MedHyg 2004, in press.

138


Introduction

The management of children withmalignant disease in Papua New Guinea(PNG) has received relatively little attention.

This is hardly surprising, since malignantdisease accounts for only a small fraction ofthe total caseload of the busy paediatric unitsthroughout the country – Port MoresbyGeneral Hospital (PMGH) statistics indicate

The management of children with cancer in Papua New Guinea: areview of children with cancer at Port Moresby General Hospital

M. KIROMAT1, J.D. VINCE2,3, G. OSWYN1 AND N. TEFUARANI2

Department of Paediatrics, Port Moresby General Hospital, Papua New Guinea andSchool of Medicine and Health Sciences, University of Papua New Guinea, Port

Moresby

SUMMARY

In the period of three and a half years between January 1998 and June 2001, 64 childrenwith cancer were seen at the Paediatric Unit of Port Moresby General Hospital (PMGH).62 children presented for the first time, whilst 2 were under review, having startedtreatment in 1996. The male:female ratio was 1.8:1. The median age was 60 monthswith an interquartile range of 36-84 months. 50% of the children were from the PortMoresby area, 15% from Central Province and 35% were referred from other provinces.Lymphoma, with Burkitt’s lymphoma predominating, was as common as leukaemia. 20(31%) of the children presented either at an advanced stage of disease or with cancerassociated with a poor prognosis with available treatment, and were not offered curativetreatment. 2 children transferred overseas for treatment. Of 42 families offered treatment38 accepted and continued. At review 5 years after the start of the study 19 of the 20children not offered treatment were known to have died and the outcome for 1 wasunknown. Of the 38 children who underwent treatment at PMGH 24 (63%) were knownto have died, 2 (5%) were still under treatment, 7 (18%) were in remission and the outcomefor 5 (13%) was unknown. Of the 24 known to have died, remission induction failed in16, relapse followed remission in 3 and 5 died from infection. The mean (SD) survival ofthose who died was 3.9 (3.4) months. 24 (51%) of the 47 known deceased children diedin hospital, including 7 (32%) of the 22 referred patients. Significant problems wereencountered in patient treatment. Infections occurred in 74% of treated children anddrug shortages were experienced in 26%. The substantial problems faced by the familiesincluded marital discord, major financial hardship and, for those referred from otherprovinces whose children died, major delays and difficulties in repatriation. It issuggested that in Papua New Guinea the most appropriate approach to treatment formost children with cancer is the model in which paediatricians at the child’s nearestappropriately staffed hospital take responsibility. Appropriate drug regimens, readilyavailable drugs, ongoing advice and data collection should be coordinated through acentral source. Accurate data should facilitate rational decisions.

1 Department of Paediatrics, Port Moresby General Hospital, Private Mail Bag, Boroko, NCD 111, Papua NewGuinea

2 Discipline of Child Health, School of Medicine and Health Sciences, University of Papua New Guinea, PO Box5623, Boroko, NCD 111, Papua New Guinea

3 Corresponding author

138-145

139


0.5% (1). The burden of infectious diseasesdwarfs that of childhood cancer. In suchcircumstances the management of paediatriccancer cannot reasonably be regarded as apublic health priority. Nevertheless for theindividual child and the family themanagement of paediatric cancer is of theutmost importance and the doctor lookingafter the child has an obligation to do what isbest for the child and the family within thelimitations imposed by circumstances.

Over the last 30 years remarkableimprovements in the outcome of treatmentof children with cancer have been achievedin the industrialized world where 70-80% ofpaediatric cancers are regarded as curable(2). In contrast the outcome from developingcountries remains poor (3,4). Theachievement of such impressive results inthe developed world, however, depends oninterdisciplinary teamwork with access tohigh technology and the use of veryexpensive treatment modalities. Since mostof the world’s children are in non-industrialized countries, it is estimated thatless than 20% of the 186,000 childrendeveloping cancer each year actually receiveadequate treatment (2). There isconsiderable international concern about thediscrepancy of care and outcome for childrenwith malignant disease between thedeveloped and developing world (5,6). It hasbeen suggested that specialized paediatriccancer units, linked to and supported byinternational centres of excellence, beestablished (2). Whilst such an approachmay be feasible in some developingcountries (7), its feasibility in countries withlow health expenditure is questionable. It isin such countries that most of the 14-16million annual deaths of children fromdiarrhoea and other infectious diseases occur(2). Poor nutritional status, prevalent in suchcountries, together with other socioeconomicindicators of poverty, is adversely associatedwith treatment outcome (8). Care incentralized units often necessitatesexpensive travel from the patient’s home andseparation from normal family supportsystems. For similar reasons long-term

treatment and follow-up is often problematic.In many parts of the world, including PNG, itis culturally very important that a child whodies is buried in the home area. This cancreate major problems in transporting a deadchild home.

The 2001 Annual Symposium of theMedical Society of Papua New Guineafocused attention on cancer and itsmanagement. Whilst reasonable data on thespectrum of paediatric malignant disease areavailable (9,10) there are no published dataon treatment outcome. Paediatric malignantdisease was included in the list of diseaseson the Paediatric Surveillance Unit registerin 1998 and in the medium to long term thiswill provide reasonably accurate nationaldata. For example, 128 children with cancerwere reported to the Unit over the three years2000-2002, giving an approximate annualrate of 40-50 (J. Leslie, PaediatricSurveillance Unit, personal communication).In order to obtain some immediate baselineinformation on treatment outcome and theproblems faced by children and their families,information available from the Paediatric Unitat Port Moresby General Hospital wascollated and analyzed.

Methods

An audit was made of the treatment andoutcome of children up to age 13 yearspresenting or referred to the PaediatricDepartment of PMGH with cancer over theperiod of 3½ years between January 1998and June 2001. The audit, revised in January2003, 5 years after the start of the study,included newly diagnosed children and thoseattending the review clinic during the studyperiod. Two of the authors sequentially tooksole responsibility for the care of affectedchildren. Problems relating to treatment anddifficulties experienced by the families werenoted.

Results

64 children were included in the study, 62of whom were newly diagnosed. The

140


remaining 2 patients had been diagnosed 2years before the study and were underreview. There were 41 males and 23 females(male:female ratio of 1.8:1). Ages rangedfrom 2 weeks to 13 years with a mean of63.6 months, a median of 60 months and aninterquartile range of 36-84 months. 31(50%) of the newly diagnosed children werefrom the Port Moresby area, 9 (15%) werefrom Central Province and 22 (35%) werereferred from other provinces. The children’sdiagnoses are shown in Table 1.

20 (31%) of the children presented eitherwith potentially treatable cancers at a veryadvanced stage, or with cancer associatedwith a very poor prognosis, and curativetreatment was not offered. A further 2children transferred overseas for treatment(1 Non-Hodgkin’s and 1 Hodgkin’slymphoma). Of the 42 families offeredtreatment at PMGH, 2 refused and 2 stoppedtreatment because of family problems.

Five years after the start of the study 19

1ELBAT

DIAGNOSIS OF CHILDREN WITH CANCER AT PORT MORESBY GENERAL HOSPITAL

sisongaiD )%(rebmuN

amohpmyL )03(91

s'ttikruB )71(11

s'nikgdoH-noN )9(6

s'nikgdoH )3(2

aimeakueL )03(91

citsalboleymetucA )71(11

citsalbohpmyletucA )6(4

citycoleymcinorhC )6(4

sruomutSNC )8(5

amotsalborueN )8(5

amotsalboniteR )9(6

ruomuts'mliW )8(5

amocrasoymodbahR )3(2

amocrass'gniwE )2(1

amotycomorhcoeahP )2(1

amoignyrahpoinarC )2(1

latoT 46

metsyssuovrenlartnec=SNC

141


(95%) of the 20 children for whom curativetreatment was not offered were known tohave died as had the 4 children who wereoffered but did not receive treatment. Of the38 children commenced and continued ontreatment at PMGH 24 (63%) had died, 2(5%) were continuing treatment and 7 (18%)were in remission. The outcome for 5 (13%)children was unknown. Details of outcomerelating to the diagnoses are shown in Table2. The mean (SD) survival of those who diedwas 3.9 (3.4) months. 3 children relapsedafter apparent remission (2 with Non-Hodgkin’s lymphoma and one with

Hodgkin’s), and 5 died from infections duringchemotherapy.

24 (51%) of the 47 known deceasedchildren died in hospital including 7 (32%) ofthe 22 children referred from provinces otherthan NCD and Central.

Problems during treatment includedinfections in 28 patients (74%) andunavailability of drugs in 10 (26%).

The parents of more than half of thechildren (20/38) were unemployed. Financial

2ELBAT

OUTCOME BY DIAGNOSIS IN CANCER PATIENTS TREATED IN PORT MORESBY GENERAL HOSPITAL

sisongaiD oN detaerT noissimeR nOtnemtaert

deiD tsoL

ruomuts'mliW 5 5 2 0 3 0

amohpmyls'ttikruB 11 01 3 0 6 2

amotsalboniteR 6 6 0 0 5 1

sruomutSNC 5 0 0 0 5 0

amotycomorhcoeahP 1 0 0 0 1 0

amoignyrahpoinarC 1 0 0 0 1 0

amocrasoymodbahR 2 2 0 0 1 1

amocrass'gniwE 1 1 0 0 1 0

amotsalborueN 5 3 0 0 4 1

amohpmyls'nikgdoH-noN 5 4 0 0 4 1

amohpmyls'nikgdoH 1 1 1 0 0 0

LLA 4 2 1 0 3 0

LMA 11 0 0 0 11 0

LMC 4 4 0 2 2 0

latoT 26 83 7 2 74 6

metsyssuovrenlartnec=SNCaimeakuelcitsalbohpmyletuca=LLAaimeakuelcitsalboleymetuca=LMAaimeakuelcitycoleymcinorhc=LMC

142


and other social problems in this groupresulted in 25% of the children havingrepeated delays in treatment. The parentsand guardians of referred patients facedmajor financial problems during their stay andin some cases this affected their willingnessto remain in hospital for the appropriateduration of treatment. The repatriation of thebodies of the 7 children referred from otherprovinces was a major problem, taking from5 to 60 days, with many phone calls and faxmessages to arrange. Long periods ofadmission resulted in marital disharmonyand, in two cases, family separation.

Discussion

Although the study is small, it hashighlighted a number of issues. Lymphoma,with Burkitt’s accounting for the majority, wasas common as leukaemia, consistent withprevious data (9,10). This is a differentpattern from that reported from mostcountries of the developed world where theleukaemias predominate (10). Acutemyeloblastic leukaemia was the predominantform of leukaemia. The present study showsa major inequality between boys and girls.Hospital data in PNG are consistent inshowing a male predominance in paediatricadmissions. Reasons for this predominance,which is seen in other developing countries,remain speculative. In 2001 the male:femaleratio in paediatric admissions at PMGH was1.36 (1). However, in this study there werealmost twice as many boys as girls. A recentstudy of paediatric cancer registrations incountries throughout the world has shownthat the male:female ratio increases as grossdomestic product decreases and as infantmortality rates increase (11). Our very limiteddata are consistent with this pattern, PNGhaving an infant mortality rate of 79/1000 anda gross national product of $US800 per capita(12,13).

The processes of caring for children withcancer in resource-poor countries presentsmedical staff, families and children with manydifficulties, some of which are listed in Table3. Our case series illustrates many of these

difficulties. Infectious complications occurredin 74%. In the absence of strict isolationfacilities and bone marrow support, reliancemust be placed on the use of methods thatare as sterile as is possible and practical, onearly aggressive use of antibiotics and onantituberculous prophylaxis with isoniazid.The requisite cytotoxic drugs were notavailable at the appropriate time for 26% ofthe children. This illustrates the difficulty inrelying on routine hospital pharmaceuticalsupplies and the necessity of having amechanism for ready availability of thestandard anticancer drugs.

The current ‘best regimens’ in the westerncontext are likely to be disastrous in thedeveloping world. Such regimens, whichobliterate the bone marrow, must besupported by the availability and use of strictisolation facilities, and platelet andgranulocyte infusions in addition to bloodtransfusion and granulocyte growthstimulating factors. Such support is notavailable in most countries of the developingworld. In these countries drug regimenswhich are relatively simple to administer,which use readily available and relativelyinexpensive drugs and which do notcompletely obliterate the marrow arerequired. Such ‘modified’ drug regimens areavailable (14), though they should beregularly reviewed.

Many of the patients presented at a latestage of the disease where it was felt thattreatment would be of no benefit. Many ofthose who were treated were at a relativelyadvanced stage, at which even optimaltreatment, had it been available, would havebeen unlikely to have achieved a high curerate.

The major problems inherent in the referralof patients from other provinces to PMGHincluded major financial hardship, familydisturbances and the distress and hardshipcaused by the delay in repatriation of thosechildren who died. These problems shouldprompt clear thinking on the wisdom of

143


3ELBAT

PROBLEMS IN CARING FOR CHILDREN WITH CANCER IN PAPUA NEW GUINEA

seitilicafcitsongaiddetimiL

gnipytbuslleccitenegotycrolacigolonummioN

seitilicafnacsIRMondnanacsTCdetimilyreV

gnissecorpdnatropsnartyspoibniyaleD

tnemeganamlacinilC

noitatneserpetaL

nosaerdooghtiwnetfo–ecnailpmocrooP

sgurdyparehtomehcfoylppuselbailernU

sgurdyparehtomehcfomurtcepsdetimiL

seitilicafnoitalosifokcaL

aeragnisrundetangisedoN

sgurdlarivitnadnalagnufitna,lairetcabitnafoylppuscitarre/detimiL

noisserppusworramenobroftroppusdetimilyreV

noisufsnartetycolunargoN

noisufsnartteletalpdetimiL

rotcafgnitalumitshtworgetycolunargoN

ecivresyparehtoidaroN

)yregruscitsalp,yregrusoruenge(seitlaicepsbuslacigrusotsseccadetimiL

laicnanifdnalaicoS

slatipsohotstneitapfonoitatropsnarT

stnerapdnanerdlihcfonoitairtapeR

nerdlihcdaedfonoitairtapeR

noitarapesylimaF

troppuslacitcarpfokcaL

troppuslaicosfokcaL

pukaerbdnadrocsidlatiraM

latipsohnignivilfoesnepxE

latnemnorivnE

eneigyhytinummocrooP

noitavitcaer/noitcefnifoytilibissop–sisolucrebutfoecnelaverphgiH

ytinummocniesaesidsuoitcefnifoecnelaverphgiH

144


referral to centres outside the home provincefor children with cancer. Paediatricians arepractising in most of the major and provincialhospitals in the country and it may well bebetter to ensure the availability ofchemotherapeutic drugs and the appropriateregimens, and open channels ofcommunication to enable the paediatriciansto treat the children as near as possible totheir home and family support.

The child with cancer and family membersare frequently hospitalized for long periods.They have to cope with a disease processthat is hard for them to understand andtreatment that is unpleasant. We have littledetailed information about the stresses thatthis places on the child and the family. Wedo know, however, that prolongedhospitalization results in major financialproblems, marital discord and sometimesfamily breakdown. It is imperative, therefore,that our management should include social,cultural and spiritual support whenevernecessary.

Children, at an advanced stage of a cancerwhich would have been potentially treatablehad it been diagnosed earlier or with a typeof cancer that is realistically incurable presentthe paediatrician with difficult choices.Objectively the decision is straightforward –why waste money and resources on a childwith an incurable condition? In practice,however, the decision ‘not to treat’ is nevereasy. There may be considerable familypressure to ‘give treatment a try’. Decisionsshould be determined as far as possible byconsideration of the likely quality of the child’sremaining life. Management options andlikely side-effects should be fully discussedwith the child’s family. The decision not totreat with chemotherapy does not absolve thedoctor from responsibility for patient care. Alldoctors should be familiar with appropriatepalliative care – including adequate painrelief. This is an area which requiresconsiderable attention – there are nopaediatric formulations of morphine availablethrough the Health Department, adultformulations are often not available, and the

cost of opiate analgesia through the privatesector is prohibitive.

It is important that in approaching the topicof paediatric cancer we should not benihilistic. Our study is not all bad news. Itillustrates that some children with cancer canbe successfully treated even in resource-poorcountries such as PNG. Some cancers havea good prognosis provided they arediagnosed early and available treatmentinstituted quickly. These include Wilm’stumour, Burkitt’s lymphoma, retinoblastomaand probably L1 acute lymphoblasticleukaemia. For example, a study fromNamibia reported a projected 5-year survivalfor children with cancer of only 37%, but forWilm’s tumour the calculated 5-year survivalwas 76% (3). Health workers need to beconstantly reminded of the early signs of thepotentially curable cancers such as anabdominal mass (Wilm’s), swellings of thefacial bones (Burkitt’s), recent onset of squint(retinoblastoma) and unexplained weight lossand lymphadenopathy (leukaemia andlymphomas). They should be encouragedto refer children with such presentations asquickly as possible. The possibility ofpaediatric cancer should be considered inchildren failing to respond to antituberculoustreatment.

Our analysis has highlighted majorproblems with referral and treatment ofchildren with cancer. We believe that carefulconsideration should be given to developingprograms for treatment of children withcancer in the Papua New Guineansocioeconomic and cultural context.Centralizing cancer services – thedevelopment of a national referral unit –might result in some improvement instatistical outcome, but at present there isno good evidence from PNG to support sucha policy, which would certainly be associatedwith major problems for children and theirfamilies. A model in which children aretreated at their nearest hospital by theprovincial paediatrician with the assuranceof regular supplies of appropriate drugs andthe ready availability of advice from a

145


173.5 Usmani GN. Paediatric oncology in the third world.

Curr Opin Pediatr 2001;13:1-9.6 Levi F, La Vacchia C, Lucchini F, Negri E, Boyle

P. Patterns of childhood cancer mortality: America,Asia and Oceania. Eur J Cancer 1995;31A:771-782.

7 Wilmas JA, Ribiero RC. Paediatric haematology-oncology outreach for developing countries.Haematol Oncol Clin North Am 2001;15:775-787.

8 Gomez-Almaguer D, Ruiz-Arguelles GJ, Ponce-de-Leon S. Nutritional status and socio-economicconditions as prognostic factors in the outcome oftherapy in childhood acute lymphoblastic leukaemia.Int J Cancer Suppl 1998;11:52-55.

9 Tefuarani N, Vince JD, Murthy DP, SenGupta SK,White JC. Childhood malignant tumours in PapuaNew Guinea. Ann Trop Paediatr 1988;8:201-206.

10 Parkin DM, Kramárová E, Draper GJ, Masuyer E,Michaelis J, Neglia J, Qureshi S, Stiller CA, eds.International Incidence of Childhood Cancer, VolumeII. IARC Scientific Publication No 144. Lyons:International Agency for Research on Cancer, 1999.

11 Pearce MS, Parker L. Childhood cancerregistrations in the developing world: still more boysthan girls. Int J Cancer 2001;91:402-406.

12 Papua New Guinea Department of Health.National Health Plan 2001-2010. Volume 1. PortMoresby: Department of Health, 2000.

13 United Nations Children’s Fund. The State of theWorld’s Children 2001. New York: UNICEF, 2001:78-81.

14 Clezy JKA, Martin WMC. Cancer Management forthe Neoplasms in Papua New Guinea: A Handbookfor Doctors. Port Moresby: Department of Health,1990.

centralized source would seem, at least tous, to be preferable at this stage in PapuaNew Guinea’s development. Such a centralsource would also be responsible for detaileddata collection and analysis. This wouldallow comparison of outcomes for patientswith ‘treatable’ cancers (such as Wilm’s andBurkitt’s) managed at provincial hospitals withthose of children treated at PMGH, thusproviding evidence on which to base referralpolicies in the future. The existing paediatricsurveillance system has the capacity toprovide detailed information but is dependenton the input from paediatricians and othercontributing doctors. Such input can certainlybe improved.

REFERENCES

1 Mokela D. Annual Report of the Paediatric Unit.Port Moresby General Hospital, Papua New Guinea,2001.

2 Wagner HP, Antic V. The problem of pediatricmalignancies in the developing world. Ann NY AcadSci 1997;824:193-204.

3 Wessels G, Hesseling PB. Outcome of childrentreated for cancer in the Republic of Namibia. MedPediatr Oncol 1996;27:160-164.

4 Sankaranarayanan R, Black RJ, Swaminathan R,Parkin DM. An overview of cancer survival indeveloping countries. IARC Sci Publ 1998;145:135-

146


micronutrients such as iron, iodine andvitamin A in infant and maternal nutritionprograms in developing countries (2-6). It isnow known that lesser degrees of zincdeficiency are more common than wasappreciated and that the subclinicaldeficiency of zinc can contribute to increasedincidence and severity of infectious diseasesin a susceptible population (7-10).

Recent scientific evidence suggests thatzinc also deserves special attention, becausezinc deficiency has been associated withlower birthweight, poor growth in childhood,reduced immunocompetence, increasedsusceptibility to infectious diseases,diarrhoeal disease and abnormal motordevelopment in malnourished children (2,11-

Introduction

Childhood malnutrition is a major problemthat can lead to long-term deficits in growth,immune function, and cognitive and motordevelopment. It can also adversely affectchildren’s behaviour and intellectualperformance and presents an increased riskof morbidity and death from infectiousdiseases. In the past, these negativeconsequences of malnutrition were mainlyassociated with inadequate protein-energyintake (1). Malnutrition is, however, a morecomplex phenomenon, and often alsoinvolves deficiencies of one or moremicronutrients (2,3). There is enoughscientific evidence to justify the specialattention being given to specific

Zinc in human health

VICTOR J. TEMPLE1 AND ANDREW MASTA1,2

School of Medicine and Health Sciences, University of Papua New Guinea, PortMoresby

SUMMARY

Malnutrition is a contributing cause of about half of the 10 million deaths annuallyworldwide, and contributes to a substantial proportion of the infectious disease morbidityamong children in developing countries. Recent epidemiological and clinical evidencehas shown that in most developing countries deficiencies of specific micronutrientsare partly responsible for the severity of infectious disease morbidity and mortality inmalnourished children. Efforts to improve micronutrient status have focused on iron,vitamin A and iodine. Supplementation with iron and vitamin A significantly reduceschild mortality, while implementation of the universal salt iodization strategy reducesthe incidence of iodine deficiency disorders. These strategies are considered to beamong the most cost-effective health interventions in developing countries. A numberof recent zinc supplementation studies in developing countries suggest that greaterpriority should also be given to the correction of mild to moderate zinc deficiency inchildren, pregnant women and lactating mothers. Some of these studies showed thatzinc supplementation reduces the duration of malaria, and the severity of diarrhoea andrespiratory infections (including pneumonia), and improves immunocompetence insusceptible children. The results of these studies indicate that zinc may be anotherspecific micronutrient in which there is widespread deficiency in developing countriesand that great benefits can be achieved by its supplementation.

1 Discipline of Biochemistry and Molecular Biology, Division of Basic Medical Sciences, School of Medicine andHealth Sciences, University of Papua New Guinea, PO Box 5623, Boroko, NCD 111, Papua New Guinea


146-158

147


such as gustin – a Zn-dependent polypeptidethat is essential for normal development oftaste buds and normal taste acuity (17,19).Zinc functions as an antioxidant and helps tostabilize cell membranes (14,19,22).

Zinc affects cell replication, growth andmaturation because it influences the activityof multiple enzymes at the basic levels ofreplication and transcription (22). Theseinclude DNA polymerase (18,22), thymidinekinase (17,18), DNA-dependent RNApolymerase (18,22), terminaldeoxyribonucleotidyl transferase (14,22),aminoacyl transferase, RNA synthetase(14,22), and a family of transcriptionalregulators with a structure known as zinc-finger motif that are involved in sequence-specific DNA recognition and geneexpression (14,17,18,22). Zinc is requiredfor the expression of multiple genesregulating mitosis (14,18,22).

Thus, the requirement of zinc for normalcellular growth and differentiation mayunderlie the impairment of physical growththat is the hallmark of zinc deficiency inchildhood (14,22).

Zinc and vitamin A metabolism

Adequate zinc status is partly necessaryfor the absorption, transport, metabolism,hepatic release and tissue utilization ofvitamin A (22,23). Zinc is required for thebiosynthesis of retinol dehydrogenase thatcatalyzes the oxidative conversion of retinolto retinaldehyde in various tissues, includingthe retina, where this process is involved inthe visual cycle (22,23). Zinc regulatesvitamin A absorption and is required for thebiosynthesis of retinol-binding protein in theliver (22,23). This implies a regulatory rolefor zinc in mobilizing vitamin A within cellsand in the liver.

Circulating zinc and vitamin Aconcentrations appear unrelated in well-nourished states but tend to co-vary inmarginally nourished individuals withcoexisting zinc and vitamin A deficiencies

16).

This review highlights the metabolic roleof zinc, its bioavailability and implication indisease states with special focus on infantand maternal health.

Metabolic significance and biochemicalfunctions of zinc

Zinc is one of the most essential traceelements in human nutrition because it is acomponent of over 200 metalloproteins witha wide range of biochemical functions,involving enzymes, structural proteins andhormones (17-19). The ability of zinc toparticipate in an impressive range ofbiochemical functions is partly due to itsflexible and easily exchangeable ligand-binding with biomolecules (19). Zinc is arelatively safe element when compared withiron and copper, especially because of its lackof oxidant properties (17,18). This greatlyfacilitates the transport of zinc within the bodyand its metabolism in individual cells (19).

Zinc is absorbed by passive diffusionmainly in the jejunum. Metallothionein andother metalloproteins such as á2-macroglobulin, where zinc is notexchangeable, albumin and transferrin,where zinc is exchangeable, and otheroligopeptides play a regular role in theabsorption, transportation and distribution ofzinc (10,18,19). Muscle and bone have thehighest concentrations of zinc, followed byskin, gastrointestinal tract, hair, nails, eyesand prostate gland (19). The relative size ofthe blood zinc pool is about 1% of the totalbody pool (19).

The ligand-binding properties of zinc areutilized effectively at the catalytic site of awide range of enzymes (17,18). Zinc isinvolved in the regulation of blood sugar,thyroid and gonadal functions, adrenalhormone and prolactin production, acid-basebalance and calcium metabolism (17-22).Zinc has structural roles in biologicalmembranes, cell receptors (for hormonesincluding testosterone) and other proteins,

148


areas usually do not show obvious signs ofzinc deficiency, but have often been found tohave low plasma zinc levels compatible withsubclinical zinc deficiency (2,6,8,28).

The phytate content in plant-based staplefoodstuffs can be reduced by enzymatic andnon-enzymatic hydrolysis (22,26,29). Someof the methods that enhance phytase-induced enzymatic hydrolysis of phytateinclude fermentation, sprouting and soaking(26,29). Endogenous phytase, which can beactivated by soaking, is high in most cerealsand in the microflora on the surface of grains(22,26). Soaking is particularly effective innon-enzymatically reducing the amount ofwater-soluble phytate in legumes and cereals(26,29). Zinc bioavailability can also beenhanced by citric acid, lactic acid, formicacid, vitamin C, high dietary protein, highlevels of sulphur-containing amino acids(methionine, lysine and cysteine), lactose andpancreatic secretion (2,22,26,28). Some ofthese compounds enhance the absorption ofzinc by preventing the formation of insolublezinc-phytate complex in the gastrointestinaltract (22,27,28). The phytate-to-zinc molarratio of a diet can be used to predict theinhibitory effect of phytate on zinc absorption(27,28,30).

The phytate-to-zinc ratio >12 found insome plant-based diets that contain lowanimal protein is considered to be a majorfactor contributing to zinc deficiency indeveloping countries (9,28). It is thereforeimportant to reduce the phytate contentduring preparation of meals. Suboptimal zincdeficiency in humans has been associatedwith phytate-to-zinc molar ratios >15 (27, 28).

The bioavailability of zinc can be reducedby a high level of calcium in a diet that ishigh in phytate (28). The complex (Ca-Zn-phytate) formed in the intestine is less solublethan the Zn-phytate complex (28). Thebioavailability of zinc can also be reduced bychemical food colourings (22,31). Accordingto Ward (31), hyperactive children showed asignificant reduction in serum zinc levels andan increase in urinary zinc output following

(22,23). Thus, zinc deficiency could imposea secondary vitamin A deficiency onmalnourished infants (23). Zinc deficiencymight also limit the effectiveness of vitaminA supplementation programs.

Most studies regarding zinc-relatedvitamin A deficiency in humans areinconclusive (23,24). Data supporting theoccurrence of vitamin A-related zincdeficiency in humans are very sparse.However, based on available researchinformation, the metabolic functions ofvitamin A are greatly enhanced in thepresence of adequate zinc status (22-24).

Thus, given the critical role of both zincand vitamin A in maintaining normalmetabolic functions and resistance toinfection, and in view of the dire lack ofadequate information about the public healthimplications of a potential zinc–vitamin Ainteraction, there is an express need for moreresearch in this direction, particularly indeveloping countries.

Sources and bioavailability of dietaryzinc

Major sources of dietary zinc include milk,red meat, liver, poultry, fish, oysters, crabs,cereals, legumes, tubers, nuts andvegetables (2,19,22,25,26). Bioavailability ofzinc is higher in milk, liver, red meat, fish,oysters and crabs; it is low in cereals,legumes, tubers and vegetables(2,19,22,25,26). The low bioavailability ofzinc from plant sources is due to their contentof phytate, lignin and fibre, which forminsoluble complexes with the ingested zincimpairing intestinal absorption (2,22,25,26).

In most developing countries, because ofeconomic constraints, the consumption ofanimal protein foods, which are rich sourcesof readily available zinc, is usually low(26,27). Staple foods in these countries aremainly cereals, legumes, tubers andvegetables, thus the bioavailability of dietaryzinc is low because of the high content ofphytate in the diet (26-29). Children in these

149


and casein in cow’s milk, which impair theabsorption of zinc (32,35). The same is truefor soymilk because of its high phytatecontent (19).

Zinc supplementation (between 10 and 20mg Zn/day) enhances linear growth andsignificantly reduces the incidence ofanaemia (13,22,32,35). Stunted childrenbenefit more than non-stunted children;children of up to 24 months of age benefitmore than older children (19,32,35). Anotherbenefit of zinc supplementation reported ina recent study by Albert et al. (40) is thatsupplementation with zinc improvesseroconversion to vibriocidal antibody and,hence, has the potential to improve theefficacy of oral cholera vaccine in normal andmalnourished children.

These findings clearly indicate the needto develop strategies for improving the zincstatus of infants and children in developingcountries. An important factor is to ensureadequate bioavailability of zinc and othermicronutrients in the diet of infants. Thereare a number of commercially preparedadequately fortified complementary foodsthat can be used for this purpose. However,these products are too expensive forwidespread use in low-income communities.A more viable option there is to increaseaggressive advocacy of awareness of zincdeficiency and to educate people on thebenefits of fortification of the already availablelocally prepared complementary foods.Fortification of home-made complementaryfoods can be achieved by addition of locallyavailable foodstuffs that are rich in zinc, suchas powdered forms of whole grains, drybeans, nuts, adequately prepared andminced seafood, poultry, pork and red meat(36-38,41,42). Gibson et al. (26) have useda similar community-based dietaryintervention strategy in rural southern Malawiwith some success.

The safe upper limits of zinc intakeadjusted for body weight proposed by theWorld Health Organization (WHO)(3,15,28,34,43) are as follows: 13 mg/day for

the consumption of beverages containingeither the food colouring ‘tartrazine’ (code E102) or ‘sunset yellow’ (code E 110).

Zinc nutriture in infants

The first six months of life are a period ofrapid growth, and zinc intake varies with themode of feeding (22,32). The relatively highzinc requirements during this period can bemet satisfactorily by breastmilk alone for mosthealthy infants (19,22,32). The infant is ableto utilize zinc from hepatic zinc thionein forseveral weeks postpartum to supplementzinc derived from breastmilk (32,33). Thebioavailability of zinc in breastmilk is very high(80%) compared to whey-adjusted cow’s milk(35%), even though the concentration of zincin breastmilk is lower than in cow’s milk (19).This difference in bioavailability is due tohigher levels of citrate and the presence oflactoferrin in breastmilk, and high levels ofphytate, calcium and casein in cow’s milk(19,22).

Healthy infants, fed exclusively onbreastmilk for four to six months, usually donot develop zinc deficiency within this period(32,33). Several researchers (32,34,35)have documented the possibility ofsuboptimal zinc status in breastfed infantsafter six months of age. This is partly relatedto the fact that zinc content of breastmilk fallswith the duration of lactation (32).

From six months to two years of age,adequacy of zinc intake becomes highlydependent on the amount and bioavailabilityof zinc from complementary foods (35-39).Thus, prolonged breastfeeding withoutadequately prepared complementary foodsmay reduce an infant’s zinc intake, therebyincreasing the risk of zinc deficiency (35-39).

The breastfed, low-birthweight infant isusually at risk of developing zinc deficiencybecause of increased requirements,potentially lower intake and/or lowerabsorption efficiency (22,35). Zinc deficiencymay also occur in children fed with cow’s milk,because of the high levels of phytate, calcium

150


developing countries (45). In addition,research in developed countries relating tomaternal zinc status and birthweight ofneonates has not produced consistent results(45-47). However, several reports haveindicated that women are at increased riskof zinc deficiency during pregnancy, becauseof the high fetal requirement for zinc (45,46).Severe maternal zinc deficiency has beenassociated with spontaneous abortion andcongenital malformations (45). Mild tomoderate zinc deficiency has beenassociated with prolonged gestation,intrauterine growth retardation, pretermdelivery, complications of labour and delivery,and low birthweight (45,47).

According to Caulfield et al. (45), thequestion as to whether maternal zincdeficiency increases the likelihood of labourand delivery complications is critical fordeveloping countries, where women havelittle access to essential obstetric services.It is therefore necessary to provide zincsupplementation for pregnant women indeveloping countries (47).

Effect of zinc on immune function

Zinc plays a crucial role in the developmentand maintenance of the immune system(14,20,22,48). It is involved in multipleaspects of the immune system, from thebarrier of the skin to the regulation of geneswithin lymphocytes (11,14,15,22,48).

Zinc is important for the normal functionof cell-mediated immunity throughneutrophils, macrophages and natural killercells (14,22,48). It also affects thedevelopment of acquired immunity andimmunoglobulin production (14,22,48). Itplays a significant role in the biological activityof a thymic hormone (thymulin) and in thesynthesis and release of various cytokines,such as interleukins (IL-1 and IL-2)(10,22,48,49). Zinc deficiency can causeimpaired immune function and increasedsusceptibility to bacterial, viral and fungalinfections (11,14,22,48). Mild to moderatezinc deficiency affects the development and/

infants from 6 to 12 months of age; 23 mg/day for children from 1 to 6 years of age; 32mg/day for girls from 10 to 12 years of age;and 34 mg/day for boys from 10 to 12 yearsof age.

For young children, zinc supplementationshould be considered in cases ofmalnutrition, low dietary zinc bioavailability,severe stunting, low plasma zinc, persistentdiarrhoea and severe infections. Since zincis only one of several nutrients that arenecessary for adequate growth in infants andyoung children, the deficiency of othernutrients may negatively affect the lineargrowth response to zinc supplementation(41,42). Thus, it is better to use amicronutrient mix containing zinc in therehabilitation of such children. Zincsupplementation to improve stunted growthis more effective when zinc is the primarygrowth-limiting nutrient in the diet (41,42).

According to Gibson (42), caution must beobserved when selecting the dose for zincsupplements, especially for catch-up growthin children with severe protein-energymalnutrition and an already compromisedimmune system. High doses of zincsupplements without the required generalnutritional rehabilitation including othermicronutrients may cause increasedmorbidity and mortality in this group ofchildren (42,44).

Maternal zinc nutriture

The non-availability of well-definedbiochemical indicators of maternal zincstatus, and the lack of consensus onappropriate indicators of zinc status forpregnant women are partly responsible forthe inability of policy-makers to recognize thepublic health significance of maternal zincdeficiency and its consequences forpregnancy and neonatal survival and well-being (45,46).

There is paucity of data on the contributionof maternal zinc deficiency to maternal andinfant morbidity and mortality rates in

151


or function of most cells of the immuneresponse, T cells, B cells and macrophages(14,20,48). Zinc deficiency causesdepressed thymic hormone activity,decreased natural killer cell activity, a specificCD4+ T-cell population depression, anddecreased production of IgA, IgM and IgG(14,20,48,49).

According to Shankar and Prasad (14), theeffect of zinc on some major components ofthe immune system is due to the many rolesthat zinc plays in basic cellular functions, suchas DNA replication, RNA transcription, geneexpression, cell division, cell activation andmaintenance of cell membrane integrity.

Effect of zinc on HIV/AIDS

There is substantial research informationindicating the significance of micronutrientsupplementation in improving the nutritionalstatus of HIV+ patients. However, the effectof zinc supplementation on HIV/AIDSpatients is a controversial issue.

Zinc has been reported to act as anantiviral agent (20,50,51), as an antioxidantand as immune-modulator in HIV infection(20,49). Zinc is capable of enhancing andinhibiting the activity of the HIV proteaseenzyme (20). Integrase, a Zn-dependentenzyme that requires ‘zinc-finger protein’ foroptimal activity, catalyzes the integration ofviral DNA with host DNA (20). Zinc, therefore,has a dual role as a major component of viralreplication and an inhibitor of replication. Lowlevels of zinc in plasma and in serum havebeen reported in HIV+ patients in theasymptomatic state (20,52) as well as in AIDSpatients (53). Some studies (20,54) did notobserve any reduction in serum zinc level inHIV+ patients. Other studies (20,52,55)reported declining zinc levels in plasma andserum of patients as the disease progresses.Studies to evaluate the impact of zincsupplementation in AIDS patients show anapparent reduction in infectiouscomplications in zinc-supplemented patientscompared to placebo (20,56). Results ofsome studies (20,56) showed significant

increase in mean CD4+ cells and in theabsolute counts of CD3+ lymphocytes after10 weeks’ supplementation with 0.45 mg/kg/day of elemental zinc. Zinc supplementationstudies using oral zinc in HIV+ children havebeen documented (48,57). Significantincreases in the CD4+ count were recordedin some of the patients in the studies.However, a study by Tang et al. (58) showedan association between increasing zincintake and increasing disease progression.Yet other epidemiological and clinical studiesof zinc intake, serum zinc levels and theirrelationship to progression to AIDS found noassociation between increasing zinc intakeand increasing disease progression(20,52,59).

Zinc plays an important role in HIVreplication. Therefore, the effect of zincintake on AIDS patients is complicated andmay have more impact on the virus than onthe immune response of the patient (20, 48).While there is an urgent need for moreintensive controlled studies of zinc in HIV/AIDS, the adverse effect of zinc deficiencyon the immune system requires that normalplasma levels of zinc be maintained throughsupplementation.

Effect of zinc on diarrhoea in children

In developing countries, the use of bothoral rehydration solution (ORS) therapy tocorrect fluid and electrolyte abnormalities,and appropriate antimicrobial drugs to treatselected diarrhoeal diseases have led to adrastic reduction in mortality and morbidityamong children (43,60). However, the poornutritional status of children in mostdeveloping countries puts them at further riskof diarrhoea-associated morbidity andmortality, regardless of the ORS therapy(43,61). Appropriate strategies to improvethe nutritional status of affected children andto reduce the frequency of stool passagehave been the prime focus of researchstudies (10,62,63).

Several studies, including clinical trials,have shown that diarrhoea is associated with

152


an increased loss of zinc and that zincsupplementation holds substantial promiseas an adjunct to the treatment of diarrhoea(15,62-65). According to Castillo-Duran andco-workers (66), marked faecal zinc lossoccurs during diarrhoea, thus a low serumzinc level can be considered as a reflectionof acute zinc loss.

Zinc supplementation studies carried outin a number of developing countries reportedsignificant reduction in the duration andseverity of episodes of acute waterydiarrhoea in infants, especially those whoseinitial plasma zinc levels were low (64,67,68).Similar results have been reported forchildren with persistent diarrhoea (69,70).Zinc supplementation was also more usefulin cases of severe or prolonged diarrhoeaepisodes than in short-lasting or milderepisodes (15,65,67,68,70).

Studies by the Zinc Investigators’Collaborative Group (65,71) indicated thatzinc supplementation (10 to 20 mg elementalzinc), given for 14 days during and afterdiarrhoea, reduces the incidence of diarrhoeain the subsequent two to three monthswithout additional zinc supplementation. Adouble-blind placebo-controlled study byRahman et al. (72) showed that combinedzinc (20 mg elemental zinc daily for 14 days)and vitamin A (200000 IU orally on day 14)supplementation is more effective in reducingpersistent diarrhoea and dysentery thaneither vitamin A or zinc alone.

The benefit of zinc supplementation in allcases of diarrhoea was not limited to stuntedchildren, acutely malnourished children or tochildren in any particular age group(61,63,65,71,72). In addition, the benefit ofthe supplementation was not affected by theaetiology of the diarrhoea episodes in thechildren, as both virus-associated and virus-free diarrhoea episodes responded well totherapy (64,65).

The issue of dose regimen for zincsupplementation and the frequency withwhich it should be given is still contentious

(43). The effectiveness of the dose stronglydepends on the age and zinc status of eachindividual (15,43,63,68). However, clinicallyimportant benefits have been reported whensupplements containing 10 mg Zn/day assulphate (15), 20 mg Zn/day as methionate(15), and more than 20 mg elemental Zn/daywere administered to children under threeyears who presented with acute diarrhoea(43,63,68). Fuchs (43) has suggested that adose adjusted for body weight might benecessary to ensure the efficacious and safetherapeutic use of zinc in the treatment ofacute infectious diarrhoea.

Thus, current research reports stronglysuggest that zinc supplementation mayrepresent an important therapeutic advancein the treatment of acute and persistentdiarrhoea in developing countries. This newstrategy may successfully complement ORSand other therapies of proven benefit, suchas continued breastfeeding of infants andgeneral nutritional rehabilitation.

Effect of zinc on acute lower respiratorytract infections in children

According to Bahl et al. (15) the prevalenceof acute lower respiratory tract infections wasabout 3-fold higher in children (12 to 59months of age) with low plasma zinc (<8.4ìmol/l) compared to children with normalplasma zinc (>8.4 ìmol/l). In addition, boyswith low plasma zinc had a 4-fold higher riskof suffering from acute lower respiratory tractinfections than girls. The authors noted thepossible involvement of coexistingdeficiencies of other micronutrients in plasmathat were not measured in the study.

Scientific reports on the effect of zincsupplementation on acute lower respiratorytract infections in children in developingcountries are still unclear. In a double-blind,randomized controlled trial, Sazawal et al.(73) showed a 45% reduction in the incidenceof acute lower respiratory infections inchildren (6 to 36 months of age) receiving10 mg elemental zinc supplementation dailyfor six months.

153


The Zinc Investigators’ CollaborativeGroup (71) found that the incidence ofpneumonia in children was reduced by 41%after pooled analysis of 10 randomizedcontrolled zinc supplementation trials in ninedeveloping countries. These trials usedbetween 10 and 20 mg elemental zinc for atleast two weeks. The researchers (71)concluded that the substantial benefits ofeither short- or long-term zincsupplementation for the prevention ofpneumonia is an important means ofimproving child survival in developingcountries.

In another study (72), zinc was associatedwith a significant increase in acute lowerrespiratory tract infection in children, 12 to35 months of age, receiving 20 mg elementalzinc for 14 days. This adverse effect wasreduced when a vitamin A capsule (200000IU) was administered to a similar group ofchildren on day 14 of the trial (72). Theseresults indicate that zinc alone might increaselower respiratory illnesses, but interactionbetween zinc and vitamin A reduces theadverse effect (72). The authors, however,emphasize the need for further studies usinglower doses of elemental zinc. This findingfurther stresses the significance of combinedmicronutrient supplementation in combatingacute lower respiratory tract infections andother diseases in children in developingcountries.

In a double-masked, randomized placebo-controlled trial in an Indian urban slum (74),infants were given 10 mg and older children20 mg elemental zinc or placebo daily for fourmonths. On enrolment all the childrenreceived a single dose of vitamin A (100000IU was administered to infants and 200000IU to older children). Substantial reductionin the incidence of pneumonia was observedin both infants and older children receivingzinc supplementation compared to theplacebo, thus indicating that routine zincsupplementation in children (6 to 30 monthsof age) can reduce the incidence ofpneumonia. In addition, routineadministration of zinc was shown to be

effective in preventing severe (rather thanmild) respiratory infections (74).

In a recent double-blind, randomizedcontrolled trial, Mahalanabis et al. (75)evaluated the effect of zinc supplementationin children (9 months to 15 years of age) withclinically severe measles accompanied bypneumonia. Children were assignedrandomly to receive 20 mg elemental zincas acetate twice daily for 6 days, or a placebo.All the children received a single oral dose(100000 IU) of vitamin A together with a fullcourse of antibiotics (75). The time taken toachieve a clinically improved status was notdifferent between the two groups of children.In addition, after 6 days of treatment, theimprovement in zinc and retinolconcentrations in serum was similar in bothgroups. The results indicate that zincsupplementation showed no additionalbenefit to severely ill children with measlesaccompanied by pneumonia, treated withvitamin A and appropriate supportive therapy(75). In the opinion of the authors, long-termzinc supplementation can preventpneumonia in children, because of improvedimmune responses. However, in an acuteillness such as measles-associatedpneumonia, there probably is insufficient timefor mounting an adequate immune responseto favourably modify such acute illness.

Inconsistency in the reports on the effectof zinc supplementation in children with lowerrespiratory tract infections in developingcountries is a clear indication that furtherappropriately designed studies are urgentlywarranted in this area. However, given thefindings to date, there is the need to includeappropriate zinc supplementation in thetreatment of lower respiratory tract infectionsin infants and children with low plasma zinclevels.

Interactions of zinc with copper andiron

The uptake of both zinc and copper occursin the small intestine, where zinc can interferewith copper bioavailability as they compete

154


(14,81,83,84). A cross-sectional studyamong school-age children in Papua NewGuinea showed an inverse associationbetween zinc status and P. falciparumparasitaemia (14, 84). In another study inThe Gambia, children receiving zincsupplements had 32% fewer clinic visits formalaria due to P. falciparum than childrenreceiving placebo (83).

A placebo-controlled trial to assess theeffects of zinc supplementation on malariamorbidity among preschool children in amalaria-endemic region of Papua NewGuinea was carried out by Anuraj Shankaret al. (84). The results indicated that zincsupplementation reduced the frequency ofhealth centre attendance due to P. falciparummalaria by 38%. Moreover, a 69% reductionwas observed for malaria episodesaccompanied by high levels of parasitaemia(>100,000 parasites/µl), suggesting that zincmay preferentially protect against moresevere malaria episodes (84). These authorsreported that zinc supplementation had noeffect on densities of P. vivax parasitaemia.According to Shankar et al. (84), zinc doesnot reduce the incidence of malaria butreduces the number of parasites that causeillness, and thus its severity. This is becauseinsufficient zinc levels affect the developmentand functioning of most immune cells, andinhibit cell-mediated immune responses(78,84).

A randomized double-blind placebo-controlled trial designed to test the hypothesisthat zinc supplementation reduces morbidityfrom P. falciparum was carried out in Africanchildren in a malaria-endemic region ofBurkina Faso by Mûller et al. (81). The resultsof this study reported no evidence of zincsupplementation being effective against P.falciparum malaria in a population of Africanchildren with a high prevalence ofmalnutrition and zinc deficiency (81).Children receiving zinc supplements were nodifferent in the number of episodes of P.falciparum malaria, or any othermalariometric parameters, from the recipientsof placebo (81). The finding was for all age

for absorption via intestinal cellmetallothionein (19,22). Copper, however,does not affect the bioavailability of dietaryzinc (22,43). Metallothionein has a higheraffinity for copper than for zinc, but zinc, notcopper, induces synthesis of intestinalmetallothionein (19,22,43).

High intake of zinc increases thebiosynthesis of metallothionein inenterocytes, which then bind with copper.The copper-metallothionein complex isexcreted in the faeces after desquamationof the enterocyte (19,22,43). It is thereforepossible that zinc supplementation canaggravate marginal copper deficiency (22,43)in malnourished children. Thus, there is aneed to ensure that zinc supplementation isaccompanied by an adequate intake ofcopper and other micronutrients. However,clinical signs of copper deficiency have beenreported only with large daily doses of zinc(150 mg) given for prolonged periods (22,43).

Several studies have also shown that theefficiency of absorption of both zinc and irondepends on the ratio in which they arepresent in the diet or supplement(22,34,76,77). Negative effect of iron on zincabsorption in the gastrointestinal tract occursonly when the Fe:Zn ratio in the diet orsupplement is greater than 2:1 (34,76).

Effect of zinc on malaria

Morbidity and mortality in malaria-endemicregions in developing countries are very highamong malnourished children (78-81). Anumber of studies have associated vitaminA deficiency in children with Plasmodiumfalciparum parasitaemia (78,82). Zinc isknown to be essential for a variety oflymphocyte functions implicated in resistanceto malaria (78). These functions includesynthesis of IgG, interferon-gamma, tumornecrosis factor and microbicidal activity ofmacrophages (14,78).

Studies on the possible efficacy of zincsupplementation in reducing childhoodmorbidity from malaria are contradictory

155


Despite the limitations of using plasmazinc as an indicator of zinc status for anindividual, plasma zinc can serve as a usefulmethod for assessing zinc status inepidemiological studies. Zinc status can alsobe assessed by determining whether thegrowth rate of a child is affected byappropriate zinc supplementation (22,28).

Dietary zinc intake data can also be usedto assess zinc status of a population (28).This would include estimating the amountand sources of zinc, phytate and fibre in thediet (86,87) and calculating the dietZn:phytate molar ratio (28,29).

Conclusion

It is clear from the information presentedin this review that the public healthsignificance of zinc deficiency cannot beoveremphasized. Those in the high-riskgroup for zinc deficiency include low-birthweight infants, children with eitherfrequent or persistent diarrhoea,malnourished children, pregnant women andlactating mothers, particularly in developingcountries. The available data indicate thateither adequate dietary zinc intake or zincsupplementation, together with othermicronutrients and appropriate clinicalmanagement, enhances the recovery ofchildren suffering from diarrhoea, infectiousdiseases, malaria and malnutrition. Not onlycan mild to moderate zinc deficiency inchildren enhance the development of theseconditions, but these conditions can causean increase in the children’s requirement forzinc. It is therefore very important to improvethe zinc status of individuals in the high-riskgroup by integrating a zinc intervention intoongoing primary health care programs and/or existing nutrition and public healthprograms in developing countries.

REFERENCES

1 Alleyne GAO, Hay RW, Picou DI, Stanfield JP,Whitehead RG. Protein Energy Malnutrition.London: Arnold Publishers, 1977.

2 Black RE. Zinc for child health. Abstract in ChildHealth Research Project Special Report.Proceedings of the Conference on Zinc for Child

groups, and was consistently seen duringboth the longitudinal study and the cross-sectional surveys. The authors concludedthat their inability to show the effect of zincsupplementation on morbidity from malariamay provide evidence that cell-mediatedimmunity is less important in the case ofmalaria in humans. Such a hypothesis wouldbe supported by the overwhelming evidenceof malaria not behaving as an opportunisticinfection in African children with HIV infectionor AIDS (81,85).

The opposite views expressed by theseauthors (81,84) create an urgent need formore placebo-controlled studies to elucidatethe effect of zinc and other micronutrients onthe morbidity and mortality of malaria inendemic regions in developing countries.

Assessment of zinc status

The non-availability of acceptablesensitive, specific and low-cost indicators forassessing the zinc status of an individual ora population at risk of developing zincdeficiency is partly responsible for the slowimplementation of appropriate interventionprograms in developing countries (10,15,28).

Zinc has a dynamic metabolic state that isregulated by strong homeostatic control ofabsorption and excretion (10,19,22). Whenabsorption of zinc is low, the body reducesthe rate of excretion of zinc in an effort topreserve zinc store (10,19,22). However,results from several studies strongly supportthe use of mean plasma zinc concentrationas the most practical biochemical method forassessing zinc status of an individual and atthe population level (2,6,10,15,19).

Factors militating against using plasmazinc for assessing zinc status of an individualinclude hypoalbuminaemia, time of samplingin relation to meals, diurnal variations andnonspecific decrease in plasma zincconcentration as a result of an acute phaseresponse following infection in adults(2,10,15,19,22).

156


Prentice-Hall International, 1996:461-462.19 Kruse-Jarres JD. Basic principles of zinc

metabolism. In: Kruse-Jarres JD, Scholmerich J,eds. Zinc and Diseases of the Digestive Tract.Proceedings of the International Falk Workshop,Freiburg, Germany, 27 Oct 1996. Berlin: Springer-Verlag, 1997:3-15.

20 Patrick L. Nutrients and HIV: part 2 – vitamins Aand E, zinc, B-vitamins, and magnesium. Altern MedRev 2000;5:39-51.

21 Neve J. Clinical implications of trace elements inendocrinology. Biol Trace Elem Res 1992;32:173-185.

22 Shils ME, Olson JA, Shike M, Ross AC, eds.Modern Nutrition in Health and Disease, 9th edition.Sydney: Lippincott Williams & Wilkins, 1998:223-239,741-742.

23 Christian P, West KP Jr. Interactions between zincand vitamin A: an update. Am J Clin Nutr 1998;68(2Suppl):435S-441S.

24 Solomons NW, Russell RM. The interaction ofvitamin A and zinc: implications for human nutrition.Am J Clin Nutr 1980;33:2031-2040.

25 Hunt JR. Bioavailability of iron, zinc, and other traceminerals from vegetarian diets. Am J Clin Nutr2003;78(3 Suppl):633S-639S.

26 Gibson RS, Yeudall F, Drost N, Mtitimuni B,Cullinan T. Dietary interventions to prevent zincdeficiency. Am J Clin Nutr 1998;68(2 Suppl):484S-487S.

27 Gibson RS, Heywood A, Yaman C, Sohlstrom A,Thompson LU, Heywood P. Growth in childrenfrom the Wosera subdistrict, Papua New Guinea, inrelation to energy and protein intakes and zinc status.Am J Clin Nutr 1991;53:782-789.

28 Gibson RS, Ferguson EL. Assessment of dietaryzinc in a population. Am J Clin Nutr 1998;68(2Suppl):430S-434S.

29 Temple VJ, Odewumi L. Soy beans and soy bean-based diets. Abstract in Proceedings of the ThirdRegional Workshop on Rural Development, Jos,Plateau State, Nigeria, 31Jul-2 Aug 1991:45-50.

30 Oberleas D, Harland BF. Phytate content of foods:effect on dietary zinc bioavailability. J Am Diet Assoc1981;79:433-436.

31 Ward NI. Assessment of chemical factors in relationto child hyperactivity. J Nutr Environ Med (Abingdon)1997;7:333-342.

32 Krebs NF. Zinc supplementation during lactation.Am J Clin Nutr 1998;68(2 Suppl):509S-512S.

33 Krebs NF, Reidinger CJ, Miller LV, Hambidge KM.Zinc homeostasis in breast-fed infants. Pediatr Res1996;39:661-665.

34 Walravens PA, Chakar A, Mokni R, Denise J,Lemonnier D. Zinc supplements in breastfedinfants. Lancet 1992;340:683-685.

35 Allen LH. Zinc and micronutrient supplements forchildren. Am J Clin Nutr 1998;68(2 Suppl):495S-498S.

36 Temple VJ, Badamosi EJ, Ladeji O, Solomon M,Gowok HG. Proximate chemical composition ofthree locally formulated complementary foods. WestAfr J Biol Sci 1996;5:134-143.

37 Ibrahim LM, Temple VJ, Badamosi EJ, Obatomi

Health, Johns Hopkins School of Public Health,Baltimore, Nov 1996. Baltimore: Johns HopkinsSchool of Public Health, Jun 1997.

3 World Health Organization, Food and AgricultureOrganization of the United Nations, InternationalAtomic Energy Agency. Trace Elements in HumanNutrition and Health. Geneva: WHO, 1996.

4 Delange F, de Benoist B, Burgi H, InternationalCouncil for Control of Iodine DeficiencyDisorders Working Group. Determining medianurinary iodine concentration that indicates adequateiodine intake at population level. Bull World HealthOrgan 2002;80:633-636.

5 Temple VJ, Ighogboja IS, Okonji MC. Serumvitamin A and beta-carotene levels in pre-school agechildren. J Trop Pediatr 1994;40:374-375.

6 Temple VJ, Agye EO, Badamosi EJ, Olumu NI,Odewumi O. Serum levels of some biochemicalparameters in preschool age children. West Afr JBiol Sci 1996;5:119-127.

7 Shrimpton R. Zinc deficiency – is it widespreadbut under-recognized? In: Subcommittee on NutritionNews, Volume 9. Geneva: United NationsAdministrative Committee on Coordination, 1993:24-27.

8 Akinkugbe FM, Ette SI. Role of zinc, copper andascorbic acid in some common clinical pediatricproblems. J Trop Pediatr 1987;33:337-342.

9 Gibson RS. Zinc nutrition in developing countries.Nutr Res Rev 1994;7:151-173.

10 Brown KH. Effect of infections on plasma zincconcentration and implications for zinc statusassessment in low-income countries. Am J Clin Nutr1998;68(2 Suppl):425S-429S.

11 Lira PIC, Ashworth A, Morris SS. Effect of zincsupplementation on the morbidity, immune function,and growth of low-birthweight, full-term infants innortheast Brazil. Am J Clin Nutr 1998;68(2Suppl):418S-424S.

12 Gibson RS, Vanderkooy PD, MacDonald AC,Goldman A, Ryan BA, Berry M. A growth-limitingmild zinc deficiency syndrome in some southernOntario boys with low height percentiles. Am J ClinNutr 1989;49:1266-1273.

13 Umeta M, West CE, Haidar J, Deurenberg P,Hautvast JGAJ. Zinc supplementation and stuntedinfants in Ethiopia: a randomized controlled trial.Lancet 2000;355:2021-2026.

14 Shankar AH, Prasad AS. Zinc and immunefunction: the biological basis of altered resistanceto infection. Am J Clin Nutr 1998;68(2 Suppl):447S-463S.

15 Bahl R, Bhandari N, Hambidge KM, Bhan MK.Plasma zinc as a predictor of diarrheal andrespiratory morbidity in children in an urban slumsetting. Am J Clin Nutr 1998;68(2 Suppl):414S-417S.

16 Black MM. Zinc deficiency and child development.Am J Clin Nutr 1998;68(2 Suppl):464S-469S.

17 Delvin TM, ed. Textbook of Biochemistry withClinical Correlations, 4th edition. Brisbane: Wiley-Liss, 1993:116,153,919.

18 Murray RK, Granner DK, Mayes PA, Rodwell VW.Harper’s Biochemistry, 24th edition. London:

157


A, Greeley S. Nutrient composition of three locallyprepared weaning foods. Biosci Res Comm1996;8:71-76.

38 Badamosi EJ, Ibrahim LM, Temple VJ. Nutritionalevaluation of a locally formulated weaning food:JUTH-pap. West Afr J Biol Sci 1995;3:85-93.

39 Lind T, Lonnerdal B, Persson LA, Stenlund H,Tennefor SC, Hernell O. Effect of weaning cerealswith different phytate contents on hemoglobin, ironstores, and serum zinc: a randomized interventionin infants from 6 to 12 months of age. Am J ClinNutr 2003;78:168-175.

40 Albert MJ, Qadri F, Wahed MA, Ahmed T, RahmanAS, Ahmed F, Bhuiyan NA, Zaman K, Baqui AH,Clement JD, Black RE. Supplementation with zinc,but not vitamin A, improves seroconversion tovibriocidal antibody in children given an oral choleravaccine. J Infect Dis 2003;187:909-913.

41 Mortorell R. Benefits of zinc supplementation forchild growth. Am J Clin Nutr 2002;75:957-958.

42 Gibson RS. Zinc supplementation for infants.Lancet 2000;355:2008-2009.

43 Fuchs GJ. Possibilities of zinc in the treatment ofacute diarrhea. Am J Clin Nutr 1998;68(2Suppl):480S-483S.

44 Doherty CP, Sarkar MA, Shakur MS, Ling SC,Elton RA, Cutting WA. Zinc and rehabilitation fromsevere protein-energy malnutrition: higher-doseregimens are associated with increased mortality.Am J Clin Nutr 1998;68:742-748.

45 Caulfield LE, Zavaleta N, Shankar AH, MerialdiM. Potential contribution of maternal zincsupplementation during pregnancy to maternal andchild survival. Am J Clin Nutr 1998;68(2Suppl):499S-508S.

46 Tamura T, Goldenberg RL. Zinc nutriture andpregnancy outcome. Nutr Res 1996;16:139-181.

47 Goldenberg RL, Tamura T, Neggers Y, Copper RL,Johnston KE, DuBard MB, Hauth JC. The effectof zinc supplementation on pregnancy outcome.JAMA 1995;274:463-468.

48 Fabris N. Zinc, the immune and endocrine system,and aging. In: Kruse-Jarres JD, Scholmerich J, eds.Zinc and Diseases of the Digestive Tract.Proceedings of the International Falk Workshop,Freiburg, Germany, 27 Oct 1996. Berlin: Springer-Verlag, 1997:36-51.

49 Tanaka Y, Shiozawa S, Morimoto I, Fujita T. Roleof zinc in interleukin 2 (IL-2)-mediated T-cellactivation. Scand J Immunol 1990;31:547-552.

50 Favier A, Sappey C, Leclerc P, Faure P, MicoudM. Antioxidant status and lipid peroxidation inpatients infected with HIV. Chem Biol Interact1994;91:165-180.

51 Sprietsma JE. Cysteine, glutathione (GSH) and zincand copper ions together are effective, natural,intracellular inhibitors of (AIDS) viruses. MedHypotheses 1999;52:529-538.

52 Graham NM, Sorenson D, Odaka N, BrookmeyerR, Chan D, Willett WC, Morris JS. Relationship ofserum copper and zinc levels to HIV-1 seropositivityand progression to AIDS. J Acquir Immune DeficSyndr 1991;4:976-980.

53 Koch J, Neal EA, Schlott MJ, Garcia-Shelton YL,

Chan MF, Weaver KE, Cello JP. Zinc levels andinfections in hospitalized patients with AIDS. Nutrition1996;12:515-518.

54 Schumacher M, Peraire J, Domingo JL. Traceelements in patients with HIV-1 infection. Trace ElemElectrolytes 1994;11:130-134.

55 Allavena C, Dousset B, May T, Dubois F, CantonP, Belleville F. Relationship of trace elements,immunological markers, and HIV-1 progression. BiolTrace Elem Res 1995;47:133-138.

56 Isa L, Lucchini A, Lodi S, Giachetti M. Blood zincstatus and zinc treatment in humanimmunodeficiency virus-infected patients. Int J ClinLab Res 1992;22:45-47.

57 Reich EN, Church JA. Oral zinc supplementationin the treatment of HIV-infected children. PediatrAIDS HIV Infect 1994;5:357-360.

58 Tang AM, Graham NM, Kirby AJ, McCall LD,Willett WC, Saah AJ. Dietary micronutrient intakeand risk of progression to acquiredimmunodeficiency syndrome (AIDS) in humanimmunodeficiency virus type 1 (HIV-1)-infectedhomosexual men. Am J Epidemiol 1993;138:937-951.

59 Abrams B, Duncan D, Hertz-Picciotto I. Aprospective study of dietary intake and acquiredimmune deficiency syndrome in HIV-seropositivehomosexual men. J Acquir Immune Defic Syndr1993;6:949-958.

60 Claeson M, Merson MH. Global progress in thecontrol of diarrheal diseases. Pediatr Infect Dis J1990;9:345-355.

61 Bahijri SM. Serum zinc in infants and preschoolchildren in the Jeddah area: effect of diet anddiarrhea in relation to growth. Ann Saudi Med2001;21:324-329.

62 Bhan MK, Bhandari N. The role of zinc and vitaminA in persistent diarrhea among infants and youngchildren. J Pediatr Gastroenterol Nutr 1998;26:446-453.

63 Black RE. Therapeutic and preventive effects ofzinc on serious childhood infectious diseases indeveloping countries. Am J Clin Nutr 1998;68(2Suppl):476S-479S.

64 Al-Sonboli N, Gurgel RQ, Shenkin A, Hart CA,Cuevas LE. Zinc supplementation in Brazilianchildren with acute diarrhoea. Ann Trop Paediatr2003;23:3-8.

65 Baqui AH, Black RE, El Arifeen S, Yunus M,Chakraborty J, Ahmed S, Vaughan JP. Effect ofzinc supplementation started during diarrhoea onmorbidity and mortality in Bangladeshi children:community randomized trial. BMJ 2002;325:1059-1062.

66 Castillo-Duran C, Vial P, Uauy R. Trace mineralbalance during acute diarrhea in infants. J Pediatr1988;113:452-457.

67 Sazawal S, Black RE, Bahn MK, Bhandari N,Sinha A, Jalla S. Zinc supplementation in youngchildren with acute diarrhea in India. N Engl J Med1995;333:839-844.

68 Sazawal S, Black RE, Bahn MK, Jalla S, Sinha A,Bhandari N. Efficacy of zinc supplementation inreducing the incidence and prevalence of acute

158


diarrhea: a community-based, double-blind,controlled trial. Am J Clin Nutr 1997;66:413-418.

69 Sachdev HPS, Mittal NK, Yadav HS. Oral zincsupplementation in persistent diarrhoea in infants.Ann Trop Paediatr 1990;10:63-69.

70 Sazawal S, Black RE, Bhan MK, Jalla S, BhandariN, Sinha A, Majmudar S. Zinc supplementationreduces the incidence of persistent diarrhea anddysentery among low socioeconomic children inIndia. J Nutr 1996;126:443-450.

71 Bhutta ZA, Black RE, Brown KH, Gardner JM,Gore S, Hidayat A, Khatun F, Mortorell R, NinhNX, Penny ME, Rosado JL, Roy SK, Ruel M,Sazawal S, Shankar A. Prevention of diarrhea andpneumonia by zinc supplementation in children indeveloping countries: pooled analysis of randomizedcontrolled trials. Zinc Investigators’ CollaborativeGroup. J Pediatr 1999;135:689-697.

72 Rahman MM, Vermund SH, Wahed MA, FuchsGJ, Baqui AH, Alvarez JO. Simultaneous zinc andvitamin A supplementation in Bangladeshi children:randomized double blind controlled trial. BMJ2001;323:314-318.

73 Sazawal S, Black RE, Jalla S, Mazumdar S, SinhaA, Bhan MK. Zinc supplementation reduces theincidence of acute lower respiratory infections ininfants and preschool children: a double-blind,controlled trial. Pediatrics 1998;102:1-5.

74 Bhandari N, Bahl R, Taneja S, Strand T, MolbakK, Ulvik RJ, Sommerfelt H, Bhan MK. Effect ofroutine zinc supplementation on pneumonia inchildren aged 6 months to 3 years: randomizedcontrolled trial in an urban slum. BMJ2002;324:1358-1361.

75 Mahalanabis D, Chowdhury A, Jana S,Bhattacharya MK, Chakrabarti MK, Wahed MA,Khaled MA. Zinc supplementation as adjuncttherapy in children with measles accompanied bypneumonia: a double-blind, randomized controlledtrial. Am J Clin Nutr 2002;76:604-607.

76 Whittaker P. Iron and zinc interactions in humans.Am J Clin Nutr 1998;68(2 Suppl):442S-446S.

77 Fairweather-Tait SJ, Wharf SG, Fox TE. Zincabsorption in infants fed iron-fortified weaning foods.Am J Clin Nutr 1995;62:785-789.

78 Shankar AH. Nutritional modulation of malaria

morbidity and mortality. J Infect Dis 2000;182(Suppl1):S37-S53.

79 Dominguez-Vazquez A, Alzate-Sanchez A.[Nutritional status in children under 6 years of ageand its relation to malaria and intestinal parasitism.][Sp] Salud Publica Mex 1990;32:52-63.

80 Sharp PT, Harvey P. Malaria and growth stuntingin young children of the highlands of Papua NewGuinea. PNG Med J 1980;23:132-140.

81 Müller O, Becher H, van Zweeden AB, Ye Y, DialloDA, Konate AT, Gbangou A, Kouyate B, GarenneM. Effect of zinc supplementation on malaria andother causes of morbidity in west African children:randomized double blind placebo controlled trial.BMJ 2001;322:1567-1570.

82 Shankar AH, Genton B, Semba RD, Baisor M,Paino J, Tamja S, Adiguma T, Wu L, Rare L,Tielsch JM, Alpers MP and West KP Jr. Effect ofvitamin A supplementation on morbidity due toPlasmodium falciparum in young children in PapuaNew Guinea: a randomized trial. Lancet1999;354:203-209.

83 Bates CJ, Evans PH, Dardenne M, Prentice A,Lunn PG, Northrop-Clewes CA, Hoare S, Cole TJ,Horan SJ, Longman SC, MRC Dunn NutritionUnit, Cambridge. A trial of zinc supplementation inyoung rural Gambian children. Br J Nutr1993;69:243-255.

84 Shankar AH, Genton B, Baisor M, Paino J, TamjaS, Adiguma T, Wu L, Rare L, Bannon D, TielschJM, West KP Jr, Alpers MP. The influence of zincsupplementation on morbidity due to Plasmodiumfalciparum: a randomized trial in preschool childrenin Papua New Guinea. Am J Trop Med Hyg2000;62:663-669.

85 Chandramohan D, Greenwood BM. Is there aninteraction between human immunodeficiency virusand Plasmodium falciparum? Int J Epidemiol1998;27:296-301.

86 Temple VJ, Ojobe TO, Onobun CE. Chemicalcomposition of Livingstone potato tubers. J Sci FoodAgric 1991;56:215-217.

87 Temple VJ, Namsat WN. Chemical evaluation ofthe nutritive value of S. rotundifolius tubers. NigerianJ Biochem 1992;7:104-109.

159


The epidemiology of malaria in the Papua New Guinea highlands: 3.Simbu Province

IVO MUELLER1,2, JULIUS KUNDI3, STEVE BJORGE4, PIOTO NAMUIGI1, GERARD SALEU1, IAN D. RILEY2 AND

JOHN C. REEDER1

Papua New Guinea Institute of Medical Research, Goroka, Malaria Surveillance andControl Unit, Goroka, Papua New Guinea, Papua New Guinea Department of Health,

Port Moresby and Australian Centre for International and Tropical Health andNutrition, Brisbane, Australia

SUMMARY

Two very distinct malaria zones can be found within Simbu Province. The north ofthe province is characterized by the absence or very low level of local malariatransmission, but there is a considerable risk of epidemics prevalent in the lower-lyingparts. During non-epidemic periods, parasite prevalence was usually under 5%, withsimilar frequencies for Plasmodium falciparum and P. vivax (47% each), and malariawas an only minor source of febrile illness. During epidemics, however, 13-36% ofpeople were infected, predominantly with P. falciparum (64%), and high levels of severemorbidity were present. In south Simbu malaria is clearly endemic with an overallprevalence of 35%, combined with a strong age-dependence of infections, lowhaemoglobin levels, high rates of enlarged spleen and moderate to severe anaemia(haemoglobin level < 7.5 g/dl) in children. The malaria epidemiology in south Simbu isthus more similar to the lowlands than to other highlands areas. Epidemic prevention,surveillance and response in the north, and bednet distribution and strengthening ofcurative services in the south, are therefore the priorities for malaria control in SimbuProvince.

1 Papua New Guinea Institute of Medical Research, PO Box 60, Goroka, EHP 441, Papua New Guinea2 Tropical Health Program, Australian Centre for International and Tropical Health and Nutrition, University of

Queensland Medical School, Herston, Queensland 4006, Australia3 Malaria Surveillance and Control Unit, Department of Health, PO Box 778, Goroka, EHP 441, Papua New

Guinea4 World Health Organization Malaria Scientist, Department of Health, PO Box 5896, Boroko, NCD 111, Papua

New Guinea

Introduction

Simbu Province consists of twogeographically distinct areas. Themountainous north of the province (elevation1400-4500 m) is densely populated andlinked to the large intermontane Wahgi andAsaro valleys in the Western and EasternHighlands respectively. The southern partof the province is remote, low lying (elevation300 to >2000 m) and is linked to themalarious lowland areas in the Gulf Province.

While the south was probably alwaysendemic for malaria (1), malaria in thenorthern part was traditionally restricted tolow-lying valleys and among communitiestrading with the Bomai people (in southSimbu) for salt, or the Bundi area in the RamuValley (2). The opening up of the highlandsand subsequent extensive recruitment oflabour from Simbu for lowlands plantations,and the establishment of coffee plantationsin lower-lying valley areas, led to a significantincrease in malaria in the northern Simbu

159-173

160


valleys (2). As elsewhere in the highlands,overall malaria parasite rates of 5% to 10%were regularly observed in the early 1960s(3), with most of these infections probablydue to P. vivax.

Malaria control with indoor DDT sprayingwas initiated in the mid-1960s in bothnorthern and southern parts of the province.In addition mass drug administration (MDA)was used in the south. Residual housespraying was highly successful in controllingmalaria in the northern parts, and rapidlybrought parasite rates down to less than 1%.In south Simbu, however, control was lesssuccessful. In the Karimui/Daribi areaparasite rates remained at 5-10% even afterseveral years of control (4). Among the semi-nomadic people in the Bomai area DDTspraying had little effect and after cessationof monthly MDA malaria quickly reboundedto pre-control levels (4). By the early 1980s,when spraying had become irregular, malariaprevalence in Karimui/Daribi had reached30% (5), significantly exceeding the pre-control level of 19% (1). Subsequently, vectorcontrol was abandoned and treatment ofsuspected cases became the mainstay ofmalaria control in all of Simbu Province.

In the last two decades, malaria has alsorebounded in the northern parts of Simbu,with major epidemics observed during the1997 El Niño event (National Department ofHealth, unpublished reports). Bothepidemics were associated with severemorbidity and a rise in mortality.

In this paper, we report the findings of anextensive series of parasitological surveyscarried out in all parts of the Simbu Provinceduring 2001 and 2002. This work was partof a more extensive project for mappingmalaria risk in the Papua New Guinea (PNG)highlands (6,7), undertaken by the PNGInstitute of Medical Research (PNGIMR) inGoroka, in collaboration with the NationalDepartment of Health and with support of theWorld Health Organization/Western PacificRegional Office (WHO/WPRO). Theobjectives of these surveys are to: a)

determine the prevalence of malarialinfection in all potentially malarious areas ofthe PNG highlands; b) delineate areas of no,epidemic and endemic malaria; and c)propose appropriate malaria controlstrategies for the different areas.

Methodology

Selection of survey villages

In order to get a good estimate of malariaprevalence in Simbu Province, with a limitedamount of surveys, a geographicalinformation system (GIS) was set up thatused village locations from the PNG NationalMapping Bureau, altitude information fromthe tactical pilot charts (TCP) and informationon different landforms from the PNGResource Information System (8), for theselection of survey villages. Villages werestratified according to altitude and landform,and random lists of villages were selectedso that lower areas (ie, with higher malariatransmission potential) were over-sampledand all major landforms were included. Theselection was done separately for each of theendemic areas in south Simbu and the non-endemic area in the north. This randomsample contained more villages than couldfeasibly be surveyed and a final selection ofsurvey villages was based on further criteria,such as accessibility, the prevailing localsecurity situation and advice from local healthauthorities. If it was not possible to samplea selected village, then a neighbouring villagewas included instead.

Surveys in the northern part of the provincewere timed to coincide with the end of therainy season (ie, April-June), traditionally themonths of peak transmission in the highlands(9,10). Areas in the south were surveyedboth in the rainy period in December-May anda dry(ier) season in June-November.

Following the above approach, a total of 6villages in south Simbu and 10 villages inthe northern parts were selected (Figure 1).Surveys were conducted in the dry season(ie, July to September 2001) and in the rainy

161


season (ie, March to May 2002). Twovillages in south Simbu (Haia andYogurumaru) were surveyed both in the dryand wet season.

Survey methodology

In order to achieve a sample asrepresentative as possible of the entire villagepopulation, a household-based samplingstrategy was used. From each selected

household, every member who could bereached during the stay in the village, wasincluded in the survey. If the village had lessthan 200 inhabitants, complete sampling ofthe village was attempted.

From each individual of the household,demographic data were recorded, a thick andthin blood film was prepared, the spleen waspalpated in a lying position and the axillarytemperature was taken. Haemoglobin levels

Figure 1. Surveyed villages in Simbu Province. Classification according to time of survey and altitudinal strata.

162


were measured using the Hemocue system(HemoCue AB, Ångelholm, Sweden).Symptomatic individuals were tested withOptimal test kits (Diamed, Cressier,Switzerland) and those found positive weretreated with chloroquine plus Fansidar. Ashort questionnaire on current symptoms,past malaria episodes, treatment, use ofbednets and recent travel was administeredto each subject or their guardian.

Giemsa-stained blood films wereexamined under the microscope for 100thick-film fields under oil immersion beforebeing declared negative. The parasitespecies in positive films were identified anddensities recorded as the number ofparasites per 200 white blood cells (WBC).Densities were converted to the number ofparasites per ìl of blood assuming 8000WBC per ìl. The slides were read at thePNGIMR in Goroka and Madang byexperienced microscopists. A more detaileddescription of the survey methodology isfound in Mueller et al. (6).

Data entry was done at the PNGIMR inGoroka using a double entry system.Statistical analyses were done using STATA7.0 (Stata Corporation, College Station, TX)and SPlus (Insightful Corporation, Seattle,WA) statistical packages. Chi-squared testsand logistic regression analyses were usedfor categorical variables. Continuousvariables were investigated using Student’st-tests, linear regression and analysis ofvariance (ANOVA). Haemoglobin valueswere adjusted for age and gender effectsusing regression splines.

Results

A total of 2902 people were sampledduring the 18 surveys in 16 different villages(Figure 1). Among the participants 52% werefemale and 48% male, 6% were aged <2years, 11% 2-4 years, 14% 5-9 years, 18%10-19 years and 52% 20 years of age or over,with equal sex ratios in all age groups.

The prevalence of malaria infections

ranged from 24% to 42% in the endemicareas in the south of the province and 0.5%to 5% in non-endemic surveys elsewhere(Tables 1 and 2, Figure 2). No seasonaldifference in prevalence rates was found insouth Simbu villages (dry 35%, wet 35%, p>0.5). In 3 epidemics investigated in 2002,21% (range 13-37%) of people had malarialinfections. Further epidemics were observedin June 1999 in the Movi and in March-April2001 in the Kerowagi area (MalariaSurveillance and Control Unit, Goroka; seeFigure 2). There was a very strong agedependence in prevalence in the endemic(÷2= 135.6, df = 4, p <0.001) and epidemic(÷2= 13.7, df = 4, p = 0.01) surveys, withprevalence peaking in the 2-4 year age groupand 5-9 year age group respectively (Table3). In non-endemic surveys no agedependence among children andadolescents was seen (÷2= 2.6, df = 3, p =0.45), although adults (>20 years) have asignificantly lower prevalence (adults 2% vs5% in others, p = 0.01).

Three of the four human malaria specieswere detected (Table 4). The most commonspecies in endemic and epidemic surveyswas Plasmodium falciparum, followed by P.vivax and P. malariae. In non-endemicsurveys, however, P. vivax was as commonas P. falciparum (47%, Table 4). P. ovale wasnot found.

Most infections were of low density: 60%of infections were sparse (<500/ìl), 9% light(500-999/ìl), 24% moderate (1000-9999/ìl)and 7% were heavy (>10,000/ìl), with P.falciparum and mixed infection showing thehighest densities (Table 4). Geometric meandensities were highest in epidemic andlowest in non-endemic surveys (Table 5).There were highly significant differences inintensity of infection among age groups inendemic surveys (F4,468 = 27.9, p <0.001).Infections were heaviest in infants andchildren <5 years old (1180/ìl, CL95[859,1621]), intermediate in children 5-9 years(394/ìl, CL95[279, 555]) and lightest inadolescents (10-19 years) and adults (287/ìl, CL95[212, 388] and 131/ìl, CL95[105, 165],

163


1EL

BAT

SU

MM

AR

YO

FR

ES

ULT

SO

FP

AR

AS

ITO

LOG

YA

ND

MO

RB

IDIT

YIN

EN

DE

MIC

AR

EA

SIN

SO

UTH

SIM

BU

egalliVedutitlA

)m(

NecnelaverP

)%(

etaretisaraPseicepsmP/vP/fP

neelpS)

%(etar

pmeT

)%(

C°5.73>reveFsyad

3)

%(

)ld/g(b

Hnae

Mela

meFela

Mb

H)

%(ld/g5.7<

1002nosaes

yrD

*uramasio

B0001

4718.23

2/41/348.72

7.48.13

1.215.11

8.1

aiaH

027891

3.032/22/74

9.930.2

8.024.01

2.012.21

enanU

048102

8.141/51/47

2.660.5

0.828.01

2.014.8

*uramurugoY

0901002

5.635/81/45

7.926.7

7.520.21

8.016.3

2002nosaeste

W

*uboD

099851

0.8341/11/93

4.221.5

5.029.11

0.111.5

aiaH

027871

2.835/51/45

9.843.51

7.925.01

7.96.9

*obageN

021108

8.328/0/11

3.115.2

9.317.11

0.211.5

*uramurugoY

0901151

4.434/41/73

6.757.01

2.711.11

1.116.7

*uaetalpiu

miraK

nodetautis

segalliv =fP

urapiclafmuido

msalP

=v

P;m

xavivmuido

msalP

=m

P;eairala

mmuido

msalP

levelnibolgo

meah=

bH

164


2EL

BAT

SU

MM

AR

YO

FR

ES

ULT

SO

FP

AR

AS

ITO

LOG

YA

ND

MO

RB

IDIT

YIN

NO

N-E

ND

EM

ICA

RE

AS

INN

OR

THS

IMB

U

egalliVedutitlA

)m(

NecnelaverP

)%(

etarseiceps

etisaraPmP/vP/fP

etarneelpS)

%(p

meT 5.73>°

)%(

CreveF

)%(

syad3

)ld/g(b

Hnae

Mela

meFela

Mb

H)

%(ld/g5.7<

1002nosaeste

W

ivoM

anomanI

0271681

5.00/1/0

5.05.0

5.023.31

8.216.0

upakiliL0661

4911.2

0/3/15.0

5.05.01

5.412.31

0.0

2002nosaeste

W

aniSaniS/awaidnu

K

lgubmo

K0841

4312.2

0/3/10.0

0.05.4

3.315.21

3.2

lgokabmaY

0571441

7.00/1/0

0.00.0

3.119.31

1.310.0

puK/iga

woreK

obma

mlaG

0451061

4.40/3/4

5.26.2

1.46.31

5.210.0

aneG

0261801

0.313/1/11

8.29.2

1.726.31

5.119.1

olgusiadnaM

0061961

3.52/0/7

0.04.2

9.88.31

2.216.0

enimu

G

ialokm

O0661

0711.4

0/5/26.0

2.18.6

9.311.31

0.0

luS

0251041

7.530/51/93

1.23.4

1.711.31

1.212.2

evauhC

ugamia

M0061

7517.21

0/31/71.7

0.07.81

3.313.21

9.1

scimedipe

gnirudsyevrus

etacidniscilati

nistluser

dnase

maN

=fP

murapiclafmuido

msalP

=v

P;xaviv

muidomsal

P=

mP;

eairalam

muidomsal

Plevel

nibolgomeah

=b

H

165


Figure 2. Prevalence of malaria in Simbu Province.Circles: non-epidemic surveys. Triangles: surveys during epidemics.For locations in south Simbu with both dry and wet season surveys, average prevalence is displayed.

respectively).

Spleen rates in the different villagesranged from 0% to 66%, surpassing 10% onlyin the endemic southern areas (Tables 1 and2). In these areas, enlarged spleens weremost common in the 2-9 year olds (59%, CL95[54, 64]), while infants (33%, CL95 [24, 44])

and adults (27%, CL95 [23, 31]) had the lowestspleen rates. The average size of anenlarged spleen was 2.7 (Hackett’s grade)with grade 2 spleens (31%) found mostcommonly. An enlarged spleen was highlysignificantly associated with a concurrentinfection (in endemic surveys 53% withenlarged spleen among positive vs 33% in

166


3ELBAT

AGE-SPECIFIC PREVALENCE OF MALARIA INFECTIONS IN RELATION TO SURVEY TYPE

puorgegA cimednE cimedipE cimedne-noN latoT

N )%(RP N )%(RP N )%(RP N

sraey2< 99 4.63 21 3.8 15 9.5 261

sraey4-2 371 1.06 33 2.12 021 2.4 623

sraey9-5 122 4.35 35 9.53 721 4.2 104

sraey91-01 482 9.53 96 5.72 261 2.6 515

> sraey02 365 1.02 832 0.61 996 9.1 0051

latoT 0431 504 9511 4092

puorgegahcaeniselpmasforebmun=Netarecnelaverpairalamllarevo=RP

slide-negative cases; in other surveys 11%vs 0.8%, respectively, both p <0.001). Therewas no difference in this association amongthe different Plasmodium species.

The overall prevalence rates of bothmeasured and reported fever (Table 1) weresignificantly correlated with the prevalenceof malaria parasites in a population (r = 0.79,p <0.001 and r = 0.70, p = 0.003, respectively;see also Table 5). However, while 40% ofreported fevers were associated with acurrent infection in endemic and epidemicsurveys, only 6% of reported fevers had aconcurrent malaria infection during non-endemic surveys (Table 5). Among theparasite-positive cases only 10% had atemperature >37.5°C at the time of thesurvey, while another 20% reported fever inthe last 3 days. There was no difference inprevalence of reported fever among slide-positive people between endemic, epidemicand non-endemic surveys (Table 5).However, the occurrence of febrile symptomswas highly dependent upon the intensity ofinfection (Fisher exact test, p = 0.003). 41%(CL95 [26, 57]) of cases with a parasite density>10,000/ìl reported fevers compared to 5%(CL95 [3, 8]) of those with densities <1000/ìl.Malarial infections were a major cause of

febrile illness both in endemic and epidemicvillages and accounted for 43% (CL95 [37,48]) and 39% (CL95 [28, 50]) of reportedfevers, respectively. In non-endemicsurveys, however, only 6% (CL95 [3, 12]) ofpeople with reported fevers had concurrentmalaria infections.

Mean haemoglobin levels in a village(Tables 1 and 2) were highly negativelycorrelated with the prevalence of malarialinfections (r = –0.87, n = 16, p <0.001) andranged from 13.8 g/dl in Lilikapu with amalaria prevalence rate (PR) of 2% to 10.1g/dl in Haia (PR 49%). Within villages aconcurrent plasmodial infection wasassociated with a decrease of 0.7 g/dl(CL95[0.5, 0.9], p <0.001). The reduction wassimilar for all types of single infections, butwas significantly larger in endemic surveys(endemic surveys: 1.4, CL95[0.9, 1.8] vs othersurveys: 0.6, CL95[0.4, 0.7] g/dl, p <0.001).The (adjusted) decreases in haemoglobinare, however, strongly dependent on levelsof parasitaemia (<1000 parasites/ìl: 0.4 g/dl; >1000 parasites/ìl: 1.1 g/dl; F2,532 = 9.1, p<0.001).

Outside south Simbu moderate or severeanaemia (Hb <7.5 g/dl) was virtually absent

167


4ELBAT

SPECIES COMPOSITION AND GEOMETRIC MEAN DENSITIES OF MALARIAL INFECTIONS

seicepS N % )lµ/(ytisneD

naemcirtemoeG ]xaM,niM[

)0431=N(ubmiShtuos–syevruscimednE

murapiclaf.P 423 5.86 634 ]065,66,04[

xaviv.P 08 9.61 913 ]044,42,02[

eairalam.P 53 4.7 281 ]0613,04[

snoitcefnidexiM 43 2.7 305 ]065,51,04[

vP/fP 82 124 ]065,51,04[

mP/fP 5 879 ]0805,061[

mP/vP/fP 1 0252

)504=N(syevruscimedipE

murapiclaf.P 25 9.16 766 ]000,051,04[

xaviv.P 42 6.82 485 ]0465,02[

eairalam.P 3 6.3 755 ]0271,041[

snoitcefnidexiM 5 0.6 903 ]065,021[

vP/fP 5

)9511=N(syevruscimedne-noN

murapiclaf.P 61 1.74 762 ]000,61,04[

xaviv.P 61 1.74 522 ]0652,04[

eairalam.P 2 9.5 75 ]08,04[

=fP ;murapiclafmuidomsalP =vP ;xavivmuidomsalP =mP eairalammuidomsalP

(Table 2), but during epidemics anaemiarates are increased significantly from 0.4%in non-endemic areas to 2% in epidemicareas (Table 5) and anaemia was associatedwith concurrent infections (OR 3.9, p = 0.06).In the endemic south, anaemia is commonin infants/toddlers (33%, CL95[24, 43]) andchildren 5-9 years (16%, CL95[11, 23]), but

rare in adolescents and adults (>10 years:1.8%, CL95[1.0, 3.1]). Within southernvillages, concurrent parasitaemia was ahighly significant risk factor for anaemia(adjusted OR 2.6, CL95[1.7, 4.2], p <0.001),irrespective of type of infection. However,the risk of anaemia increased dramaticallywith increasing intensity of infection (<1000/

168


5EL

BAT

SU

MM

AR

YC

OM

PA

RIS

ON

OF

PA

RA

SIT

OLO

GIC

AL

FEA

TUR

ES

AN

DM

OR

BID

ITY

LEV

ELS

DU

RIN

GE

ND

EM

IC,E

PID

EM

ICA

ND

NO

N-E

ND

EM

ICS

UR

VE

YS

syevrusci

mednEsyevrus

cimedipE

syevrusci

medne-noN

eulavp

naeM

LC

59nae

ML

C59

naeM

LC

59

ygolotisaraP

etarecnelaver

P%3.53

]9.73,7.23[%7.02

]0.52,9.61[%9.2

]1.4,0.2[100.0<

%murapiclaf.P

1%5.07

]4.47,3.66[%0.46

]9.37,2.35[%1.74

]9.46,8.92[10.0

snoitcefnidexi

m%

%2.7]9.9,0.5[

%0.6]3.31,0.2[

%0.0]3.01,0.0[

62.0

ytisnednae

mcirte

moegllarevO

293]364,133[

906]229,304[

522]814,121[

20.0

slevelytidibro

mllarevO

Cº5.73>p

meT%8.6

]2.8,5.5[%2.3

]5.5,7.1[%9.0

]6.1,4.0[100.0<

revefdetrope

R%3.42

]7.62,0.22[%4.02

]7.42,6.61[%0.01

]8.11,3.8[100.0<

aimeatisarap

htiwrevef

detropeR

2%8.24

]4.84,3.73[%0.93

]4.05,4.82[%1.6

]2.21,5.2[100.0<

edosipeairala

mdetrope

R3

%5.03]0.33,0.82[

%2.64]2.15,2.14[

%0.41]1.61,0.21[

100.0<

neelpsdegraln

E%2.04

]9.24,6.73[%2.4

]7.6,5.2[%7.0

]4.1,3.0[100.0<

)ld/g()bH(level

nibolgomea

H9.01

]1.11,8.01[6.21

]8.21,4.21[3.31

]4.31,1.31[100.0<

ld/g5.7<

bH

%9.6]5.8,5.5[

%0.2]9.3,9.0[

%4.0]0.1,1.0[

100.0<

169


noitcefnitnerrucnochti

wdetaicossa

ytidibroM

Cº5.73>p

meT%9.9

]9.21,3.7[%8.01

]6.91,1.5[%1.3

]2.61,1.0[24.0

revefdetrope

R%6.92

]0.43,5.52[%1.83

]3.94,7.72[%9.12

]0.04,3.9[71.0

neelpsdegraln

E%8.25

]3.75,1.84[%1.31

]2.22,7.6[%3.6

]8.02,7.0[100.0<

)ld/g()bH(level

nibolgomea

H1.01

]3.01,9.9[4.11

]9.11,9.01[5.21

]2.31,7.11[100.0<

ld/g5.7<

bH

%1.11]4.41,3.8[

%8.4]7.11,3.1[

%0.0]9.01,0.0[

30.0

LC

59sti

milecnedifnoc

%59=

1snoitacifitnedi

seicepsllafo 2ai

meatisaraptnerrucnochti

wsesacrevef

detroperfonoitroporp 3

yevrusotroirp

skeew

owt

170


ìl: adjusted OR (AOR) = 1.5; >1000/ìl: AOR= 5.5; LR-test: ÷2= 34.5, df = 2, p <0.001).

Significantly more people in epidemicsurveys reported having had a ‘malaria’episode in the two weeks prior to the surveythan in endemic and non-endemic villages(Table 5). However, of all people reporting amalaria episode only 55% went to a healthfacility for treatment and 62% reported takingantimalarial drugs to treat the infection.Within surveys, a reported malaria episodein the prior two weeks was moderatelyassociated with an increased risk of infection(adjusted OR = 1.3, CL95 [1.0, 1.6], p = 0.03).Of all people that reported prior antimalarialdrug use, 27% had a positive blood slide atthe time of the survey.

Only a small number of people in eachvillage (6%) reported using bednets. Only inthe Kerowagi area, which was ‘hit’ by anepidemic in 2001, did bednet coverageexceed 10%. Travel outside the provincewas uncommon: 4% of people in villages inthe north had access to the highlandshighway, but only 0.2% of people in southSimbu reported travelling in the month priorto the survey.

Discussion

Simbu Province can be divided into twoareas each with very different malariaepidemiology. With an overall parasiteprevalence of 35%, south Simbu is one ofthe areas with the highest endemicity amongthe PNG highlands provinces surveyed todate. Similar rates are only found in the lowerJimi Valley (6) or in the Lake Kutubu area inSouthern Highlands (11). Consequently,malaria epidemiology in south Simbu is moresimilar to lowland areas than to otherhighlands areas. Although parasite rates aresomewhat lower than in the Madang area –40% in Cattani et al. (12) – or the Woseraarea, East Sepik Province – 60% in Gentonet al. (13) – strong age-dependence ofinfections (indicative of high levels ofacquired immunity), high spleen rates, lowhaemoglobin levels and high prevalence of

anaemia in children are all typical featuresof highly endemic areas (14). Malaria isclearly the predominant source of febrileillness and a major public health problem insouth Simbu.

Compared to the situation before and atthe end of the malaria control era, two strikingpatterns emerge. Firstly, the prevalence ofmalaria in Karimui today is significantly higherthan in the early 1960s (1), but onlymarginally higher than the level observed inthe early 1980s (5). Secondly, there has beena dramatic shift from P. vivax to P. falciparumdominance, with the proportion of P.falciparum increasing from 41% in 1982 (5)to >70% in 2002. The shift towards P.falciparum might be at least partly due toincreasing antimalarial drug resistance, asindicated by the high number of people withpositive blood slides despite reporting priorantimalarial drug use. A complete discussionof changes in malaria epidemiology and itsrelation to control in Karimui will be givenelsewhere (Mueller et al. in preparation).

The malaria situation in the rest of theprovince is comparable to the situation foundin neighbouring provinces surveyed to date(6,7). The malaria situation in the Kerowagiarea and along the Wahgi River in Gumineis similar to the one observed in the lowerWahgi Valley in Western Highlands Province(6), with parasite prevalence even in the wetseason usually below 5% but with asignificant risk of epidemics. The situationin the other areas is comparable with mostof Eastern Highlands Province, wheremalaria transmission is negligible outside ofepidemics, which have a tendency to occurat the end of the wet season (7). The presentsurveys show that malaria in Simbu has onceagain reached pre-control levels. However,P. falciparum is now at least as important asP. vivax in local transmission, and duringepidemics it is clearly the predominantspecies.

As in Eastern Highlands Province (7),epidemics are associated with high levels ofacute morbidity in all age groups, with a

171


significant proportion of reported feversattributable to malarial infections. However,malaria is not a major contributor to febrileillness in north Simbu outside epidemics.Despite this fact, most fevers get regularlytreated with antimalarials. A more detaileddiscussion of morbidity patterns in similarhighlands populations is given elsewhere(6,7).

Conclusion and recommendations

Based on overall parasite prevalence andoccurrence of epidemics Simbu Province canbe divided into 3 areas with differing malariatransmission potential (Figure 3).

Area with endemic malariatransmission

In all parts of south Simbu malaria is highlyendemic with the current prevalence ratessignificantly above the levels prior to control.There is no seasonality of transmission andP. falciparum is the clearly dominant species.Malaria is a major cause of febrile illness,haemoglobin levels are low and severeanaemia is common. Malaria is probably themost important public health problem in theseareas and interventions are clearlywarranted.

Due to lack of data, the malariological

Figure 3. Areas of distinct malaria transmission potential in Simbu Province.

172


status of the Mogiagi area remains unclear.However, the altitude of the area and thecloseness to the highly endemic area inBomai increase the likelihood for endemicmalaria, albeit at a lower level than elsewherein south Simbu.

Many populations in south Simbu are stillsemi-nomadic due to the importance of sago,bush foods, hunting and fishing for localsubsistence. The mobility of these peopleand the remoteness of many areas greatlyreduce the effectiveness of indoor residualspraying (IRS). Therefore, distribution ofinsecticide-treated mosquito nets (ITN)together with health education andstrengthening of curative services should bethe mainstay of malaria control in Karimui.

Areas with limited endemictransmission but high risk of epidemics

In non-epidemic times, parasiteprevalence in these areas is generally low(<5%) with infections due to P. falciparum andP. vivax equally important. During epidemics,however, P. falciparum clearly dominates andtransmission can reach high levels (PR>30%). Most epidemics occur late in therainy or early in the dry season (March-June).Unless properly controlled with earlydiagnosis, prompt treatment with appropriateantimalarial drugs and vector control,morbidity during epidemics can be severe.This zone includes all areas in north Simbualong major rivers below 1600-1650 m. Ofspecial importance are the areas along theWahgi River from Kerowagi to Gumine,which border areas of epidemic risk in theneighbouring Western Highlands Province(Mueller et al., in press). Seasonal verticalmigration to lower-lying food and coffeegardens, where people sleep in makeshiftgarden houses, may put villagers at risk ofepidemics, even though the main villages aresituated at higher altitudes wheretemperatures preclude malaria transmission(15).

Malaria control in these areas shouldconcentrate on prevention and control of

epidemics. Yearly indoor residual housespraying in villages below 1600-1650 m, eg,along the Wahgi River, should be considered.Where migration to low-lying gardens isimportant, mosquito nets for use in gardenhouses should supplement the spraying.Reliable reporting and efficient epidemicresponse is essential for the entire northernpart of Simbu.

Areas with no or very low local malariatransmission

These areas have parasite prevalencebelow 3% with little seasonal variation. Whilesome local P. vivax transmission in areasbelow 1700 m cannot be excluded, most ofthe infection is probably acquired duringtravels to malarious areas. Malaria is not asignificant source of morbidity. Below 1750m there is a potential for epidemics duringextraordinary climatic conditions such in theMovi area in 1998. Above 1700 m normalmean temperatures are too low for localmalaria transmission. This zone includes allthe areas of Simbu 1600-1650 m above sealevel. No vector control activities areindicated; however, strengthening ofdiagnostic and curative services to deal withimported malaria cases and theestablishment of an effective epidemicsurveillance and response system are highlyrecommended.

REFERENCES

1 Ford PG. Patrol Report – Project 21 A. Departmentof Public Health, Malaria Service, Port Moresby,1965.

2 Radford AJ, Van Leeuwen H, Christian SH. Socialaspects in the changing epidemiology of malaria inthe highlands of New Guinea. Ann Trop MedParasitol 1976;70:11-23.

3 Ewers WH, Jeffrey WT. Parasites of Man in Papua-New Guinea. Brisbane: Jacaranda Press, 1971.

4 McMahon JE. Malaria endemicity amongst thesemi-nomadic people of the Karimui area of PapuaNew Guinea. PNG Med J 1974;17:99-107.

5 Barker J, Harvey P, Hide R, Shield J, Tulloch J,Vrbova H. Nutrition, malaria, intestinal parasitosisand morbidity in Karimui. Report of the Simbu LandUse Project. Papua New Guinea Institute of MedicalResearch, Goroka, Papua New Guinea, 1989.

6 Mueller I, Taime J, Ivivi R, Yala S, Bjorge S, RileyID, Reeder JC. The epidemiology of malaria in thePapua New Guinea highlands: 1. Western Highlands

173


Province. PNG Med J 2003;46:16-31.7 Mueller I, Bjorge S, Poigeno G, Kundi J,

Tandrapah T, Riley ID, Reeder JC. Theepidemiology of malaria in the Papua New Guineahighlands: 2. Eastern Highlands Province. PNG MedJ 203;46:166-179.

8 Bellamy JA. Papua New Guinea Inventory ofNatural Resources, Population Distribution and LandUse Handbook. Canberra: Institute of BiologicalResources, Commonwealth Scientific and IndustrialResearch Organization (CSIRO), 1986.

9 Peters W, Christian SH. Studies on theepidemiology of malaria in New Guinea. V. Unstablehighland malaria – the entomological picture. TransR Soc Trop Med Hyg 1960;54:537-548.

10 Peters W, Christian SH. Studies on theepidemiology of malaria in New Guinea. IV. Unstablehighland malaria – the clinical picture. Trans R SocTrop Med Hyg 1960;54:529-536.

11 Hii J, Dyke T, Dagoro H, Sanders RC. Healthimpact assessments of malaria and Ross River virusinfection in the Southern Highlands Province ofPapua New Guinea. PNG Med J 1997;40:14-25.

12 Cattani JA, Tulloch JL, Vrbova H, Jolley D, GibsonFD, Moir JS, Heywood PF, Alpers MP, StevensonA, Clancy R. The epidemiology of malaria in apopulation surrounding Madang, Papua New Guinea.Am J Trop Med Hyg 1986;35:3-15.

13 Genton B, Al-Yaman F, Beck HP, Hii J, Mellor S,Narara A, Gibson N, Smith T, Alpers MP. Theepidemiology of malaria in the Wosera area, EastSepik Province, Papua New Guinea, in preparationfor vaccine trials. I. Malariometric indices andimmunity. Ann Trop Med Parasitol 1995;89:359-376.

14 Genton B, Al-Yaman F, Beck HP, Hii J, Mellor S,Rare L, Ginny M, Smith T, Alpers MP. Theepidemiology of malaria in the Wosera area, EastSepik Province, Papua New Guinea, in preparationfor vaccine trials. II. Mortality and morbidity. AnnTrop Med Parasitol 1995;89:377-390.

15 Mueller I, Kaiok J, Reeder JC, Cortés A. Thepopulation structure of Plasmodium falciparum andPlasmodium vivax during an epidemic of malaria inthe Eastern Highlands of Papua New Guinea. Am JTrop Med Hyg 2002;67:459-464.

174


A comparison of booked and unbooked mothers delivering atthe Port Moresby General Hospital: a case-control study

F. FAILING1, P. RIPA2,3, N. TEFUARANI2 AND J. VINCE2

Department of Paediatrics, Port Moresby General Hospital, Papua New Guinea andSchool of Medicine and Health Sciences, University of Papua New Guinea, Port

Moresby

SUMMARY

A case-control study of unbooked mothers delivering at the maternity unit of the PortMoresby General Hospital was undertaken over a period of 7 months. 48 mothers whohad no antenatal attendances during pregnancy were recruited with 96 booked controls.Reasons for non-attendance, understanding of the importance of antenatal clinics andsocioeconomic and demographic factors were recorded to assess likely risk factors fornon-attendance. Almost half the mothers cited financial difficulties as the reason fornon-attendance at antenatal clinics despite most of them knowing that it was importantto have antenatal care during pregnancy. The two most important risk factors for beingunbooked were mothers having no education (p <0.001) and the type of employment oftheir spouse (p <0.01). Unbooked mothers were more likely to have preterm babies (OR16.1; 95%CI 3.4-75.7) and all 6 perinatal deaths occurred in babies born to unbookedmothers. Remedial approaches would need to take into account maternal education,education of partners and the fact that despite free antenatal services in urban clinicsfinancial difficulties in terms of other costs involved still remain an obstacle to overcome.

1 Department of Paediatrics, Port Moresby General Hospital, Free Mail Bag, Boroko, NCD 111, Papua NewGuinea

2 Child Health Discipline, School of Medicine and Health Sciences, University of Papua New Guinea, PO Box5623, Boroko, NCD 111, Papua New Guinea


Introduction

That antenatal care improves bothperinatal and maternal outcomes is now wellestablished particularly for the moreunderprivileged segments of society (1-3).Papua New Guinea (PNG), like many otherdeveloping countries, has made antenatalservices one of its priorities and hasattempted to increase primary health servicedelivery to its largely rural populace and toincrease the numbers of pregnant womenattending antenatal clinics and delivering inhealth facilities.

There is, however, great variation in theuse of antenatal services throughout the

country. A countrywide survey in 1996reported that 78% of mothers received someform of antenatal care and 53% hadprofessionally supervised deliveries (4). PortMoresby in the National Capital District hasthe best antenatal coverage with only 5-6%of deliveries at the only public maternityhospital being from unbooked mothers.These deliveries, however, have a high riskfor perinatal mortality and morbidity (5). In2002 the total perinatal mortality rate fordeliveries at Port Moresby General Hospital(PMGH) was 25/1000; however, the perinatalmortality rate for unbooked mothers was108.9/1000 compared to 20.8/1000 forbooked pregnancies (6).

174-180

175


The purpose of this study was to ascertainwhy unbooked mothers do not use availableantenatal services in an urban setting whereaccess to services should be relatively easyand to assess the association of several riskfactors with poor use of antenatal services.A secondary outcome was to determine theperinatal outcome of the unbookedpregnancies in relation to the bookedmothers.

Subjects and methods

This case-control study was undertakenin the maternity unit of PMGH from 1February to 31 August 2001. Cases weremothers who had no recorded antenatal clinicattendance. For each case 2 bookedmothers who delivered just before or just afterthe case were used as controls.

Data collected were both quantitative andqualitative. Names, antenatal history, parityand residential area were collected from thelabour ward registry. Mothers wereinterviewed either in the labour ward or thepostnatal ward with a pretested questionnairethat recorded the following information:reasons for non-attendance at antenatalclinics; understanding of the importance ofantenatal clinic visits; type of housing;maternal education; maternal occupation;husband’s occupation; previous antenatalhistory; baby’s outcome; gestational age atbirth; baby’s birthweight; special care nursery(SCN) admission; and the outcome of theadmission. Babies were defined as pretermif they were less than 37 weeks’ gestation.The progress of babies admitted to SCN wasmonitored.

Qualitative data relating to reasons for notattending antenatal clinic were subsequentlyplaced into predetermined categories suchas financial or work-related difficulties.

Data were analyzed using Epi Info version6.0 and SPSS for windows version 10.Univariate analyses were performed onfactors likely to contribute to non-attendanceat clinics using ÷2 tests and Mann-Whitney U

tests. Univariate variables were entered intoa logistic regression model to predictindependent risk factors for non-attendanceat antenatal clinics. A p value of <0.05 wastaken as significant.

Results

The study consisted of 48 cases and 96controls. The characteristics of the twogroups are shown in Table 1.

Area of residence

The majority of the mothers were from PortMoresby city whilst the rest came fromperiurban as well as rural villages in theCentral Province. There was no significantdifference between cases and controls interms of residential status overall. However,when analyzed as subgroups a higherproportion of the unbooked mothers camefrom rural villages compared to controls (27%compared to 14%, p = 0.065).

Age distribution and parity

The median age and interquartile rangesof unbooked and booked mothers were 24(IQR 22-31.5) and 25 years (IQR 22-29)respectively. There was no significantdifference in the ages of cases and controlseven at both extremes of age. There wasalso no significant difference in the parity ofthe mothers between booked and unbookedmothers.

Education of mothers

Lack of any formal education (in 33% ofthe unbooked mothers) was stronglyassociated with unbooked status (p =0.0009). For those who had some formaleducation there was no significant differencein educational levels between the bookedand unbooked mothers.

Occupation of mothers

There was no significant difference in the

176


1ELBAT

DIFFERENCES BETWEEN BOOKED AND UNBOOKED MOTHERS

dekoobnU dekooB

srehtomfonoitubirtsidegA

)segnarelitrauqretni(naideM )5.13-22(42 )92-22(52

ecnediserfoaerA )%(rebmuN )%(rebmuN

nabrU )3.85(82 )7.66(46

nabruireP )6.41(7 )8.91(91

segallivlaruR )1.72(31 )5.31(31

sutatsnoitacudE

*noitacudelamrofoN )3.33(61 )3.7(7

3edargotpU )2.4(2 )1.2(2

6-3edarG )52(21 )6.04(93

01-7edarG )1.72(31 )5.73(63

21-11edarG 0 )0.1(1

noitutitsniyraitreT )3.8(4 )3.7(7

loohcslanoitacoV )1.2(1 )2.4(4

*noitapuccos'dnabsuH

deyolpmenU )5.73(81 )6.51(51

delliksimeS )2.92(41 )9.12(12

dellikS )52(21 )3.65(45

lanoisseforP )3.6(3 )2.4(4

tnedutS 0 )0.1(1

)atadgnissim(nwonknU )1.2(1 )0.1(1

sutatsgnisuoH

edam-fleS )7.61(8 )3.6(6

esuohtiK )3.33(61 )3.13(03

tnanevocwoL )2.92(41 )7.71(71

tnanevocdiM )5.21(6 )7.14(04

tnanevochgiH )3.6(3 )1.3(3

)atadgnissim(nwonknU )1.2(1 0

* 100.0<psisylanaetairavitlumnotnacifingis

177


2ELBAT

REASONS WHY MOTHERS WERE UNBOOKED

nosaeR rebmuN %

seitluciffidlaicnaniF 02 7.14

)ysuboot(detalerkroW 11 9.22

setadfonoitaluclacsiM 6 5.21

)lorneotdemahsa(deirramtoN 3 52.6

)llamsoot(nerdlihcrehtotuobadeirroW 3 52.6

)srekrowhtlaehfodiarfa(suorapitlumdnarG 2 2.4

rehtO 3 52.6

latoT 84

occupation of booked and unbookedmothers.

Husband’s occupation and level ofhousing

The type of job husbands held was asignificant influence on the attendance ofmothers at antenatal clinics.

A little over half the husbands of unbookedmothers were either unemployed or hadunskilled jobs whilst over half the husbandsof the booked mothers had skilledemployment.

Overall the husbands of booked mothershad better jobs than those of the cases (p =0.009). Housing was also shown to be asignificant factor on univariate analysis butwas not significant on multivariate analysis.

Multivariate analysis

Multivariate analysis using logisticregression analysis showed that the twosignificant independent risk factors for non-attendance at antenatal clinics were noeducation (OR 5.8; 95%CI 2.1-15.6; p =0.0009) and type of husband’s job (p =0.009).

Multivariate analysis excluding mothersfrom rural villages showed that for the urbanmothers no education and husband’s jobwere still significant risk factors (p = 0.006and p = 0.04 respectively).

Reasons for non-attendance andunderstanding of the importance ofattending clinics

The main reason for non-attendance atclinics was financial difficulties with almosthalf (42%) citing financially related difficultiesin getting to antenatal clinics (Table 2).

The next most frequent reason was workrelated: unbooked mothers were either toobusy or had other reasons such as returningfrom the home village after recreationalleave. This was followed by less frequentcauses for non-attendance which includedshame for single mothers and fear ofcastigation by health workers in those whowere multiparous.

All mothers knew it was important to attendantenatal clinics but 13% of them apparentlydid not know the reason why they shouldattend the clinics. The varied responses totheir understanding of the importance ofattending antenatal clinics are listed in Table

178


3.

Perinatal outcome

14 babies were premature of whom 12were born to unbooked mothers (OR 16.1;95%CI 3.4-75.7; p = 0.000). All 6 deathsoccurred in the babies of unbooked mothers(p = 0.0001): 3 were stillbirths and 3 died inthe special care nursery.

Discussion

The study shows that mothers who do notattend antenatal clinic are more likely thanbooked mothers to have had no formaleducation and are more likely to be from apoorer socioeconomic background asindicated by husband’s employment andhousing status. These findings are similarto those in a study by Klufio and Kariwigaover a decade ago, which found maternalage less than 20, social class, husband’seducation, previous antenatal attendanceand maternal education as significantlydifferent. In their discussion they made thepoint that in a male-dominated societyhusband’s education was a clearly significantfactor in antenatal attendance, which hadpreviously been unsuspected, and thatremedial approaches should includeeducation of husbands on the benefits ofantenatal care (5).

Our study did not include the level ofhusband’s educational status nor the levelof husband’s income and their effects on thebooking status of mothers. It did, however,show that there was a higher proportion ofhusbands in the booked group who wereemployed, especially in the skilled category,than in the unbooked group, in which two-thirds of the husbands were eitherunemployed or engaged in semiskilled jobs.It is likely, therefore, that the educationalstatus and income of husbands in the bookedgroup were higher.

Financial difficulties were cited by almosthalf of the mothers who did not attendantenatal clinic. In a study by Benjamin etal. on the effect of user fee charges onutilization of obstetric services in PortMoresby, household income was noted tohave a significant association (7).

Though perinatal outcome was not theprimary outcome measure in this study non-attendance at antenatal clinics wasassociated with detrimental outcomes. All 6perinatal deaths occurred in babies ofunbooked mothers. Babies of unbookedmothers were also more likely to be bornpreterm and be admitted to the neonatalward.

This is consistent with the annual PMGH

3ELBAT

MOTHERS' UNDERSTANDING OF PURPOSE FOR ANTENATAL ATTENDANCE

nosaeR rebmuN %

snoitacilpmocdiovA 83 0.72

ybabevilyhtlaehreviledoT 63 5.52

wonkt'noD 81 8.21

sixalyhporpdnapu-kcehcroF 13 0.22

yreviledysaednadrawruobalotsseccaysaE 7 0.5

oretuniybabfossergorpwonkoT 11 8.7

latoT 141

179


statistics (6) and other studies in PNG whichshow that poor or no antenatal care isassociated with increased perinatal mortality(8-9), increased incidence of prematurity andlow birthweight (10) and increased admissionto neonatal care nurseries.

Several studies have shown that theleading causes of perinatal deaths could bepreventable. Syphilis is cited as contributingtowards 10-22% of perinatal mortality(8,9,11). Malaria is a common cause of lowbirthweight (10), which in itself is a risk factorfor mortality (11). Antenatal care shouldreduce mortality and morbidity due to thoseconditions as well as identifying high-riskmothers.

Unfortunately those mothers most needingantenatal care are the ones who do not attendantenatal clinics. Mothers are more likely notto attend antenatal clinic if they are frompoorer socioeconomic backgrounds or haveno formal education. Financial and logisticdifficulties are cited as contributing factorsbut it is not clear how well prospectivemothers understand the need for antenatalreview.

Our questionnaire reveals that somemothers have some knowledge of whyantenatal care is important but theirknowledge seems inadequate. A morequalitative study may be needed to assessthe level of knowledge of mothers.

Limitations to this study

There are several limitations to this study.Firstly, the study looked at two groups ofwomen who delivered at the PMGH: thosewho attended antenatal clinics and those whodid not. Mothers who had home deliveries,whether or not they were antenatal clinicattenders, would have been missed in thisstudy. The number of women who deliver athome without presenting to hospital isunknown.

Secondly, the study included rural motherswho delivered at PMGH. This may impose

limitations on our study conclusions as thisintroduces heterogeneity into the studypopulation since rural mothers may havediffering reasons and risk factors for clinicattendance, especially accessibility. Analysisof data on the urban mothers, however,showed that no education and husband’semployment were still independent riskfactors; the sample of rural mothers is toosmall for any similar analysis.

Thirdly, the fact that perinatal deaths wereall in the unbooked mothers may suggestbias in choosing of controls. However, suchbias is unlikely since controls were chosenfrom booked mothers just before and justafter the index case. The observed perinatalmortality in the unbooked mothers mostprobably reflects the small sample size of 144mothers in the study against the backgroundof a total annual delivery of over 10,000mothers at PMGH. In that context it isconsistent with the high level of perinatalmortality in PMGH in unbooked mothers(108.9 per thousand in unbooked motherscompared to 20.8 per thousand in bookedmothers in 2002) (6).

Could there have been differences in theway staff treated uneducated motherscompared to educated mothers that mayhave resulted in reluctance to use antenatalservices? Though there is no evidence tosuggest that this may have been the case,this is an issue that needs to be evaluated.

The qualitative issues could have beenbetter designed and investigated to look atissues of why antenatal services are not usedby some mothers considering their level ofunderstanding of the purpose of antenatalcare.

Finally, there are limitations to conclusionsdrawn from a case-control study design.There may be other confounding factors thathave not been considered and the associatedrisk factors may not be causative. However,these associated risk factors can beimportant markers for interventional studiesand programs to address the issues of

180


perinatal and maternal mortality that are stillimportant health care issues in PNG andother third world countries.

Conclusion

Maternal education and socioeconomicstatus are important determinants of use ofantenatal care in Port Moresby.

There is a need for more studies intofactors involved in utilization of antenatal careand other health services. A betterunderstanding of how well prospective users/clients view the importance of such servicesis needed in order to plan remedialapproaches. There is also a need for similarstudies in other parts of the country as thepattern of antenatal clinic use could be verydifferent from metropolitan Port Moresby.

REFERENCES

1 Gortmaker S. The effects of prenatal care uponthe health of the newborn. Am J Public Health 1979;69:653-660.

2 Greenberg R. The impact of prenatal care indifferent social groups. Am J Obstet Gynecol1983;145:797-801.

3 Conde-Agudelo A, Belizan JM, Diaz-Rossello JL.Epidemiology of fetal death in Latin America. ActaObstet Gynecol Scand 2000;79:371-378.

4 National Statistical Office of Papua New Guinea.Demographic and Health Statistical Survey. PortMoresby: National Statistical Office, 1996.

5 Klufio CA, Kariwiga G. A comparison of unbookedmothers delivering at Port Moresby General Hospitalwith mothers seen antenatally: socioeconomic andreproductive characteristics. PNG Med J 1992;35:3-9.

6 Amoa AB. Department of Obstetrics andGynaecology Annual Report. Port Moresby GeneralHospital, Port Moresby, 2002.

7 Benjamin AL, Sapak P, Purai JK. User chargesand utilisation of obstetric services in the NationalCapital District, Papua New Guinea. Pac HealthDialog 2001;8:38-43.

8 Amoa AB, Klufio CA, Moro M, Kariwiga G, MolaG. A case-control study of stillbirths at the PortMoresby General Hospital. PNG Med J 1998;41:126-136.

9 Duke T, Michael A, Mgone J, Frank D, Wal T,Sehuko R. Etiology of child mortality in Goroka,Papua New Guinea: a prospective two-year study.Bull World Health Organ 2002;80:16-25.

10 Brabin BJ, Ginny M, Sapau J, Galme K, Paino J.Consequences of maternal anaemia on outcome ofpregnancy in a malaria endemic area in Papua NewGuinea. Ann Trop Med Parasitol 1990;84:11-24.

11 Frank D, Duke T. Congenital syphilis at GorokaBase Hospital: incidence, clinical features and riskfactors for mortality. PNG Med J 2000;43:121-126.

181


Strongyloides fuelleborni kellyi and other intestinal helminths inchildren from Papua New Guinea: associations with nutritional status

and socioeconomic factors

SARAH E. KING1,2 AND C. G. NICHOLAS MASCIE-TAYLOR1

Department of Biological Anthropology, University of Cambridge, United Kingdom

SUMMARY

This survey examined the prevalence and intensity of Strongyloides fuelleborni kellyiand other intestinal helminths in children 5 years of age or under living near Kanabea,Papua New Guinea. Of 179 samples, 27% of the children tested positive for Strongyloides,with 81% of these children being a year or less in age. Overall, 68% of the children hadone or more infections including Ascaris lumbricoides and hookworm (Necatoramericanus) as well as Strongyloides. Egg counts in the stools ranged from 100 to98,300 eggs/ml for Strongyloides, 100 to 59,200 eggs/ml for Ascaris and 100 to 3400eggs/ml for hookworm. There were significant associations between Strongyloidesintensity and weight for age and weight for height such that children with higherintensities had, on average, lower z-scores. Relationships between the prevalence ofhelminth infections and socioeconomic factors were also observed. Logistic regressionmodels showed that children living farther away from Kanabea (more than 2 hours’walking distance), in smaller households (5 or less people) and with uneducated mothersbest predict children with Strongyloides. Two of these variables also predicted thepresence of hookworm: maternal education and household size. However, in contrastto Strongyloides, a larger household size (6 or more people) was significantly associatedwith the presence of hookworm. House type was associated with the prevalence ofAscaris, with children living in houses with tin roofs being less likely to have Ascaristhan those living in traditional houses. In addition, maternal education was associatedwith Ascaris intensity in those children with infection, such that the mean intensitieswere greater in children of uneducated mothers.

1 Department of Biological Anthropology, University of Cambridge, Downing St, Cambridge CB2 3DZ, UnitedKingdom

2 Current address: Mother and Infant Research Unit, Department of Health Sciences, University of York, YorkYO10 5DD, United Kingdom

Introduction

In Papua New Guinea (PNG), there is anintestinal helminth that is unique to thecountry: Strongyloides fuelleborni kellyi.Although it has not yet been renamed, recentanalyses do not support a close relationshipwith S. fuelleborni fuelleborni, suggesting thatit likely originated from a local zoonoticsource (1). It has been found to have awidespread, although discontinuous,distribution (2,3), and has been associated

with a potentially fatal disease called‘swollen belly syndrome’. This syndromehas repeatedly occurred in two areas:Kanabea (Gulf Province) and Wanuma(Madang Province) (4-6), with onlysporadic cases reported elsewhere (3).Previous research has examinedrelationships between this helminth andanthropometric status, but the resultshave been inconsistent (G. Barnish,personal communication, 7,8). One studyhas examined S. fuelleborni kellyi in

181-191

182


relation to socioenvironmental variables(village and family size), but no significantrelationships were observed (9). The presentstudy describes the results from a smallparasitological survey conducted nearKanabea and examines relationships ofStrongyloides and other intestinal helminthswith anthropometric status and householdsocioeconomic variables. While no swollenbelly syndrome was observed during thesurvey, this study provides new informationregarding relationships betweenStrongyloides and the socioeconomicenvironment in PNG.

Materials and Methods

The study area and population

This study was carried out in severalvillages surrounding Kanabea MissionStation in the Kaintiba District of the GulfProvince. Although the Mission Station hassome modern developments including ahealth clinic (established in 1968), runningwater and hydroelectricity, the villages aroundKanabea remain largely traditional and mostfamilies rely on shifting horticulture. Thepeople in this area, the Kamea, have beenreported to suffer from poor nutrition andhealth (10) and local nurses note that malaria,respiratory infection, skin infections,diarrhoea and anaemia are commonlyobserved, as well as parasitic infections.

Most of the villages surveyed hadmountainside locations, ranging fromapproximately 1000 to 2500 metres abovesea level. The mean annual temperature inthe area is 20 to 26º C (11) and the annualrainfall is typically between 3556 and 5080millimetres (12). With respect to the physicalenvironment, the distributions ofStrongyloides, hookworm and Ascaris inPNG have been examined in association withrainfall, altitude, slope, population density,landform and rock type (13). None of thesevariables fully account for the distribution ofparasites, although Strongyloides was foundto be rare or absent in areas of limestoneand polygonal karst.

Data collection

The cross-sectional survey was conductedfrom September 1996 to January 1997.Villages were stratified based on size anddistance from Kanabea Mission Station, thenrandomly selected within each stratum. Ineach of the eleven villages selected, allfamilies that had at least one child 5 years ofage or under were invited to participate inthe study after obtaining permission from thevillage headman.

On the day of the survey, a labelled 60 mlscrew-topped plastic sample container andhalf a tongue depressor were given to theparents for each child. The parents wereasked to put a small amount of the child’sfaeces (the size of the end of a thumb) intothe container using the tongue depressor andreturn the container as soon as possible(within 24 hours). As soon as the sampleswere received, they were flooded in anexcess of 10% formalin and shaken to makea thick slurry. The samples were analyzedtwice (once in Kanabea and once in alaboratory in the United Kingdom) followingthe volumetric dilution method (3). Thismethod was chosen as it has been usedpreviously in the Kanabea area. It quantifiesegg counts as the number of eggs per packedml of faeces and approximates the eggs pergram of other methods (G. Barnish, personalcommunication). A list of the children whorequired treatment for parasitic infection wasprovided to the clinic and/or local aidpost.

Anthropometric measurements werecarried out following recommendedtechniques (14) with subjects wearing lightclothing and barefooted. Standing height andrecumbent length (on children less than 18months of age) were measured to the nearest0.1 centimetre, and weight to the nearest 0.5kilogram. The age of the children wasdetermined from clinic cards with the date ofbirth accurate to within two weeks. Nutritionalstatus was determined by comparing heightsand weights with the National Center forHealth Statistics (NCHS) references andproducing z-scores (15,16) using the Centers

183


for Disease Control/World HealthOrganization (CDC/WHO) program. Thetechnical error of measurement and reliabilitywere calculated, and found to be close to thelevel suggested by Ulijaszek and Kerr (17).With the help of a local translator,standardized interviews were conductedwhich provided information on thesocioeconomic characteristics of eachhousehold.

Results

Prevalence of helminths

Of 179 stool samples collected fromchildren 5 years of age or under, 27% testedpositive for Strongyloides fuelleborni kellyi,with 81% of these children being aged 1 yearor less. The distribution and prevalence ofsingle, double and multiple infections ofStrongyloides, Ascaris lumbricoides andhookworm (Necator americanus) arepresented in Table 1. No children were foundto be infected with Trichuris trichiura. Overall,68% of the children sampled were infectedby one or more helminths.

The youngest child infected withStrongyloides was only 2 months of age,

whereas the earliest cases of hookworm andAscaris were in children 5 and 6 months oldrespectively. Logistic regression analysisshowed that the prevalence of Ascaris andhookworm increased with age – Ascaris: age(p <0.001), age2 (p <0.005); hookworm: age(p <0.001), age2 (p <0.05). In contrast,Strongyloides prevalence decreased withage, with the highest prevalence between 3and 24 months – Stronglyloides: age (p<0.005). After controlling for age effects, nosignificant differences in prevalences wereobserved between the sexes for anyhelminth.

Intensity of helminth infections

Egg counts ranged from 100 to 98,300eggs/ml of faeces for Strongyloides, 100 to59,200 eggs/ml for Ascaris and 100 to 3400eggs/ml for hookworm. 19% of the childrenhad Strongyloides intensities over 9900 eggs/ml (Table 2), a value considered to berelatively low in comparison to the highintensities (ie, over 100,000) found in somechildren surveyed in PNG (2). The majorityof children with Ascaris (69%) had lightinfections – defined by the WHO (18) as aninfection of <5000 eggs/g of faeces. Noindividuals had a hookworm intensity of more

1ELBAT

DISTRIBUTION AND PREVALENCE OF SINGLE, DOUBLE AND MULTIPLE INFECTIONS

shtnimleH N %

sediolygnortS ylno 62 5.41

siracsA ylno 12 7.11

ylnomrowkooH 71 5.9

sediolygnortS dna siracsA 8 5.4

sediolygnortS mrowkoohdna 5 8.2

siracsA mrowkoohdna 63 1.02

sepyteerhtllA 9 0.5

shtnimlehoN 75 8.13

latoT 971 001

184


than 5000 eggs/ml, and almost all infectionswere less than 1000 eggs/ml (93%).

Table 3 gives the geometric mean intensityof helminthic infection by age group for thetotal sample, as well as the elevated meanswhen the uninfected children were excludedie, the positives only. Strongyloides intensitydecreased with age, whereas there was atrend of increasing intensity with age forAscaris and hookworm. When only thoseindividuals who tested positive wereconsidered, relationships with age were lessclear, with the exception of Ascaris. Nosignificant sex differences in mean egg

counts were found.

Helminth infections and nutritionalstatus

The overall mean z-scores for the 179children were –2.21 (standard deviation =1.47) for height for age (HAZ), –2.52 (SD =1.23) for weight for age (WAZ) and –1.56 (SD= 1.37) for weight for height (WHZ). The z-score data were normally distributed: HAZ(Kolmogorov-Smirnov = 0.05, df = 172, p =0.20); WAZ (K-S = 0.06, df = 179, p = 0.20);WHZ (K-S = 0.06, df = 167, p = 0.10). The z-scores suggest that both acute and chronic

2ELBAT

DISTRIBUTION AND PREVALENCE OF STRONGYLOIDES, ASCARIS ANDHOOKWORM BY INTENSITY (EGGS/ML FAECES)

lm/stnuocggE N %

sediolygnortS

009-001 92 4.06

0099-0001 01 8.02

00999-00001 9 8.81

> 000001 0 0.0

latoT 84 001

siracsA

009-001 61 6.12

0094-0001 53 3.74

00091-0005 41 9.81

> 00002 9 2.21

latoT 47 001

mrowkooH

009-001 26 5.29

0094-0001 5 5.7

> 0005 0 0.0

latoT 76 001

185


malnutrition were present. No sexdifferences were observed for the z-scoresalthough a negative relationship between ageand z-scores was found – HAZ: age (p<0.05); WAZ: age (p <0.001), age2 (p<0.001); WHZ: age (p <0.001), age2 (p<0.01).

Analyses of variance and regressionanalyses revealed no significant associationsbetween the prevalence of any helminth andz-scores (after controlling for the effects ofage). There were, however, significantnegative relationships between the intensityof Strongyloides infection and WAZ and WHZusing multiple regression tests (for positivesonly) (Table 4). While these relationshipswere not highly significant (p <0.05), themodels accounted for 22% and 27% of thevariance for WAZ and WHZ respectively. Nosignificant associations were observed forHAZ.

Helminth infections and socioeconomicvariables

Nine household socioeconomic variableswere examined including village distance toKanabea, maternal and paternal education,languages spoken by the parents, father’semployment, house type, household size andtoilet type.

Initially, chi-squared tests were conductedto examine associations between thesesocioeconomic variables and the presenceor absence of each type of parasite (Table5). For Strongyloides, children living invillages farther away from Kanabea werealmost three times more likely to be infectedthan children living a shorter distance,mothers with no education were five timesmore likely to have infected children, andchildren living in a house with five or lesspeople were almost twice as likely to be

3ELBAT

INTENSITY OF INFECTION (EGGS/ML) BY AGE GROUP (TOTAL SAMPLE AND POSITIVES ONLY)

egApuorg

)shtnom(

latoTelpmas

sediolygnortSnaemcirtemoeg

stnuocgge

siracsA cirtemoegstnuocggenaem

mrowkooHnaemcirtemoeg

stnuocgge

N llA sevitisoP llA sevitisoP llA sevitisoP

9.2-0 7 31.2 35.991 - - - -

9.5-0.3 91 03.62 78.0922 - - 92.1 00.001

9.11-0.6 91 19.22 08.7371 88.2 41.158 53.1 00.203

9.32-0.21 73 58.51 10.219 99.32 98.4851 28.2 78.322

9.53-0.42 73 88.2 40.752 59.45 78.1152 88.21 55.091

9.74-0.63 53 43.2 18.453 28.131 73.7643 09.83 40.752

9.95-0.84 81 42.3 20.2211 31.133 06.5206 17.54 30.903

0.46-0.06 7 98.3 48.30831 03.62 47.1402 76.43 45.441

latoT 971 52.6 38.629 03.62 91.0172 66.7 30.922

186


4ELBAT

WEIGHTED LEAST SQUARES REGRESSION OF STRONGYLOIDES INFECTION (POSITIVES ONLY) AS

PREDICTORS OF ZAW AND ZHW (FINAL MODELS)

tnednepeDelbairav

tnednepednIselbairav

ES±b ß p R2

ZAW egA 830.0±8711.0– 595.1– 300.0 22.0

egA 2 000.0±7100.0 162.1 610.0

goL sediolygnortS 181.0±8734.0– 233.0– 020.0

tnatsnoC 557.0±1170.0 - 529.0

ZHW egA 550.0±8802.0– 660.2– 500.0 72.0

egA 2 100.0±1300.0 167.1 200.0

goL sediolygnortS 352.0±2636.0– 943.0– 610.0

tnatsnoC 670.1±9467.2 - 410.0

egarofthgiew=ZAWthgiehrofthgiew=ZHW

rorredradnats=ES

infected than children living in a house withsix or more people. When these variableswere examined using logistic regressionanalysis, all three significantly predictedStrongyloides infection after controlling for thelinear effects of age (Table 6).

For Ascaris, mothers who spoke onlyKamea were more than twice as likely to haveinfected children than mothers who couldalso speak Pidgin and/or English (Table 5).In addition, children living in houses with tinroofs were less than half as likely to beinfected with Ascaris than children living intraditional houses. Using logistic regressionanalysis, the best predictor of Ascarisinfection was house type after controlling forthe effects of linear and quadratic age (Table6).

Two socioeconomic variables weresignificantly associated with hookworm:mother’s education and household size.Similar to Strongyloides, mothers who hadno education were more likely to have

children infected with hookworm thanmothers with some education (Table 5). Incontrast to Strongyloides, however, childrenliving in larger households were more ofteninfected with hookworm than children livingin smaller households. These two variablesremained significant in logistic regressionanalysis after controlling for age and age2

(Table 6).

Given the small number of children withhigh egg counts for any type of parasite,intensity of infection (high vs low) was notexamined using chi-squared tests or logisticregression analyses. In infected children (ie,positives only), the egg counts (eggs/mlfaeces) could be normalized using log10 (x+1)transformations for Ascaris only: log Ascaris(Kolmogorov-Smirnov = 0.08, df = 74, p =0.20). Regression analysis shows thatmother’s education was found to be the bestpredictor of Ascaris intensity, such that meanintensities were lower in children whosemothers had any education (Table 7). Thismodel accounted for 15% of the variance for

187


5ELBAT

PRESENCE OF STRONGYLOIDES, ASCARIS AND HOOKWORM BY SOCIOECONOMIC VARIABLES

elbairaV N ediolygnortS s%

siracsA%

mrowkooH%

aebanaKotecnatsidegalliV

ecnatsidgniklaw'sruoh2nahtsseL 73 8.01* 8.73 7.92

ecnatsidgniklaw'sruoh2nahteroM 241 0.13 3.24 4.93

noitacudes'rehtoM

noitacudeoN 851 7.92* 0.34 5.04*

noitacudeemoS 81 6.5 2.22 7.61

noitacudes'rehtaF

noitacudeoN 69 2.03 9.64 6.93

noitacudeemoS 36 2.22 3.33 3.33

egaugnals'rehtoM

ylnoaemaK 151 8.92 4.44* 4.04

hsilgnE/nigdiP 52 0.21 0.02 0.42

egaugnals'rehtaF

ylnoaemaK 26 3.23 9.14 7.83

hsilgnE/nigdiP 89 5.32 8.04 7.53

tnemyolpmes'rehtaF

deyolpmereveN 131 5.03 9.83 2.83

tnemyolpmeynA 74 0.71 8.64 2.63

epytesuoH

lanoitidarT 851 5.82 3.44* 6.83

foorniT 12 3.41 0.91 6.82

ezisdlohesuoH

sselro5 57 3.73** 7.43 7.62*

eromro6 201 6.91 1.64 1.54

epyttelioT

teliothsuB 93 5.83 6.34 7.84

enirtaltiP 831 9.32 6.04 1.43

* 50.0<p** 10.0<p

188


6ELBAT

LOGISTIC REGRESSION ANALYSES FOR PREDICTING CHILDREN WITH STRONGYLOIDES, ASCARIS AND

HOOKWORM USING SOCIOECONOMIC VARIABLES (FINAL MODELS)

tnednepeDelbairav


ß ÷22222 p

sediolygnortS egA 630.0– 863.9 200.0 :tcerroctnecreP

ecnatsidegalliV 684.1– 003.6 210.0 %4.53:seY

noitacudes'rehtoM 082.2 234.4 530.0 %9.29:oN

ezisdlohesuoH 440.1 066.7 600.0 %0.77:llarevO

tnatsnoC 594.2– 641.5 ÷2 148.22:100.0<p

siracsA egA 671.0 651.81 100.0< :tcerroctnecreP

egA 2 200.0– 748.01 100.0 %8.07:seY

epytesuoH 933.1 716.4 230.0 %2.96:oN

tnatsnoC 203.4– 699.32 %9.96:llarevO÷2 292.5:

50.0<p

mrowkooH egA 171.0 555.21 100.0< :tcerroctnecreP

egA 2 200.0– 919.4 720.0 %2.56:seY

noitacudes'rehtoM 774.1 978.3 940.0 %9.67:oN

ezisdlohesuoH 309.0– 902.5 220.0 %4.27:llarevO

tnatsnoC 167.4– 919.12 ÷2 449.8:50.0<p

Ascaris intensity.

Discussion

The prevalence and intensity of helminthinfections in children from the Kanabea areaare of significant concern given theirassociations with hookworm anaemia (19),Strongyloides-related swollen bellysyndrome (6) and other health consequences(20,21). It is estimated that children withhookworm may experience blood lossranging from 0.16 ml to 5.4 ml per day.

Pawlowski et al. (22) have estimated that acount of roughly 1000 Necator americanuseggs/g faeces is indicative of approximately32 worms – N. americanus is the speciesprevalent in Papua New Guinea (23). OneN. americanus worm can result in a loss of0.03 to 0.05 ml of blood per day (19,21).

A high percentage (81%) of children agedfrom 0 to 12 months were infected withStrongyloides. In a previous surveyconducted in the Kanabea area (4), 86% ofchildren in this age group were infected with

189


7ELBAT

REGRESSION OF MATERNAL EDUCATION AS A PREDICTOR OF ASCARIS INTENSITY (FINAL MODEL)

tnednepeDelbairav


ES±b ß p R2

goL siracsA egA 500.0±6731.0 992.0 900.0 51.0

)ylnosevitisop( noitacudes'rehtoM 503.0±648.0– 703.0– 700.0

tnatsnoC 391.0±370.3 - 000.0

rorredradnats=ES

Strongyloides – a result that suggests thatno reduction in Strongyloides prevalence (inyoung children) has occurred in almost 20years (difference of proportions, z = 0.52,which is not significantly different). Amongthe Kamea, small babies spend most of theirtime in bilums (string bags) lined with driedbanana leaves or cloth. Eggs, larvae andfree-living adult Strongyloides have beenfound in leaves and cloth from the bilums ofinfected infants (24). It has been suggestedthat keeping infants in soiled bilums mayincrease the intensity of Strongyloidesinfection as a result of external auto-infection(3,6).

Data from the earlier survey for childrenless than a year in age (the group whereresults could be compared) also reveals that57% had hookworm and none had Ascaris(4). In comparison, the present study had asignificantly lower prevalence of hookworm(10%) (z = 3.59, p <0.001), and a significantlyhigher prevalence of Ascaris (14%) (z = 2.54,p <0.05) in this age group.

When the intensities of infection arecompared with those from a study conductedin Western PNG, including Kamea children(2), the percentage of children (5 years ofage or under) with heavy and light infectionsdoes not significantly differ (difference ofproportions: Ascaris z = 1.56, hookworm z =1.66, Strongyloides z = 1.21). Similar to thepresent study, Ashford et al. (2) observedheavy Strongyloides infections in infants. Inthe Ashford et al. survey, 28% of children 5

years of age or less had Strongyloidesintensities over 9900 eggs/ml faeces.

At least three other studies have examinedrelationships between anthropometric statusand Strongyloides fuelleborni kellyi. Onefound a negative relationship betweenweight-for-age and weight-for-height z-scores and egg counts in children aged from0 to 10 years living in isolated ruralcommunities (G. Barnish, personalcommunication). Barnish and Harari (8) alsofound that Strongyloides egg load wasnegatively associated with weight for age, butthe sample size was small. The final studyfailed to find a relationship betweenStrongyloides and nutritional status (7).While a relationship between Strongyloidesintensity and anthropometric status wasobserved in the present study, the result isinconclusive due to the small sample size.However, the relatively high amount ofvariance as demonstrated in the modelsprovides a strong basis for this relationshipto be further tested using a cross-overrandomized controlled trial.

Regarding socioeconomic variables, onlyone other study in PNG has examined theassociation between Strongyloides andfamily size, but no significant relationship wasobserved (9). While it is evident from thepresent study that relationships exist betweensocioeconomic variables and parasiticinfections among the Kamea, they are notstraightforward, nor are all of the relationshipshighly significant. Nevertheless, all threetypes of parasite were significantly

190


associated with maternal education. As hasbeen presumed elsewhere, maternaleducation is likely to be a reflection ofmaternal technology, defined by Mata (25)to include appropriate handling of food andwater, care of children during illness, disposalof children’s stools, personal and householdhygiene, and a positive attitude towardshealth care. Among the mothers surveyed,only 10% had any education, suggesting that,whatever the mediating factor is, thisrelationship should be given furtherconsideration given its potential impact onchild health. Indeed, maternal education isalso associated with nutritional status amongthe Kamea (26).

It is possible that the socioeconomicvariables associated with each type ofhelminth may differ depending on the modeof transmission (Strongyloides andhookworm can be transmitted through theskin, while Ascaris is transmitted faeco-orally). House type could be a proxy forsanitation. However, among the Kamea,traditional houses are typically made ofbamboo and thatched with pandanus leaf,but are similar in structure to houses with atin roof, which are also made of bamboo. Itis possible that for Ascaris, improvedsocioeconomic status may be an importantmediating variable, whereas forStrongyloides and hookworm, otherenvironmental factors may be moreimportant.

In conclusion, this study has demonstratedthat children 5 years of age and under in theKanabea area are at risk of helminthinfection, with both sexes equally affected.In addition, the prevalence of helminths inyoung children has not improved incomparison to previous surveys in the area.While conclusions based on cross-sectionaldata should be made with caution, this studysuggests that socioeconomic variables at thehousehold level are associated with theprevalence of helminths. While causativerelationships cannot be determined from thepresent study, the data have identified someuseful directions for further research.

ACKNOWLEDGEMENTS

Funding for the study was made possibleby the Cambridge Commonwealth Trust,Smuts Memorial Fund and Selwyn CollegeCott Fund. A small grant to write this paperwas kindly provided by the Wenner-GrenFoundation for Anthropological Research.Thanks also to Dr Stanley Ulijaszek, Dr GuyBarnish and Dr Dick Ashford who providedhelpful suggestions on collecting andidentifying parasites, to Dr James Barrett whoread a copy of the manuscript, to ananonymous reviewer, to everyone atKanabea Mission Station who made theresearch project possible and, finally, to allthe families who participated in the study.

REFERENCES

1 Dorris M, Viney ME, Blaxter ML. Molecularphylogenetic analysis of the genus Strongyloides andrelated nematodes. Int J Parasitol 2002;32:1507-1517.

2 Ashford RW, Hall AJ, Babona D. Distribution andabundance of intestinal helminths in man in westernPapua New Guinea with special reference toStrongyloides. Ann Trop Med Parasitol 1981;75:269-279.

3 Ashford RW, Barnish G, Viney ME. Strongyloidesfuelleborni kellyi – infection and disease in PapuaNew Guinea. Parasitol Today 1992;8:314-318.

4 Ashford RW, Vince JD, Gratten MJ, Bana-KoiriJ. Strongyloides infection in a mid-mountain PapuaNew Guinea community. PNG Med J 1979;22:128-135.

5 Vince JD, Ashford RW, Gratten MJ, Bana-KoiriJ. Strongyloides species infestation in young infantsof Papua New Guinea: association with generalisedoedema. PNG Med J 1979;22:120-127.

6 Ashford RW, Barnish G. Strongyloides fuelleborniand similar parasites in animals and man. In: GroveDI, ed. Strongyloidiasis: A Major RoundwormInfection of Man. London: Taylor and Francis,1989:271-286.

7 Barnish G, Barker J. An intervention study usingthiabendazole suspension against Strongyloidesfuelleborni-like infections in Papua New Guinea.Trans R Soc Trop Med Hyg 1987;81:60-63.

8 Barnish G, Harari M. Possible effects ofStrongyloides fuelleborni-like infections on childrenin the Karimui area of Simbu Province. PNG Med J1989;32:51-54.

9 Barnish G, Ashford RW. Occasional parasiticinfections of man in Papua New Guinea and IrianJaya (New Guinea). Ann Trop Med Parasitol1989;83:121-135.

10 Saweri W. Report of Kaintiba Nutrition and HealthSurvey, 1992-1993. Port Moresby: SmallholdersMarket and Food Supply Project, 1995.

191


11 Spenceley AP. Rainfall and temperature. In: KingD, Ranck S, eds. Papua New Guinea Atlas: A Nationin Transition. Port Moresby: Robert Brown andAssociates and The University of Papua NewGuinea, 1983:94-95.

12 Hart D. Rainfall. In: Ward RG, Lea DAM, eds. AnAtlas of Papua New Guinea. Glasgow: University ofPapua New Guinea and Collins, Longman, 1970:42-45.

13 Barnish G, Ashford RW. Strongyloides cf.fuelleborni and other intestinal helminths in PapuaNew Guinea: distribution according to environmentalfactors. Parassitologia 1990;32:245-263.

14 Lohman TG, Roche AF, Martorell R.Anthropometric Standardization Reference Manual.Champaign: Human Kinetics Books, 1988.

15 Hamill PVV, Drizd TA, Johnson CL, Reed RB,Roche AF. NCHS Growth Curves for Children Birth-18 Years. Vital and Health Statistics, Series II. No.165. DHEW Publication 78-1650. Washington:Department of Health, Education and Welfare, 1977.

16 Hamill PVV, Drizd TA, Johnson CL, Reed RB,Roche AF, Moore WM. Physical growth: NationalCenter for Health Statistics percentiles. Am J ClinNutr 1979;32:607-629.

17 Ulijaszek SJ, Kerr DA. Anthropometricmeasurement error and the assessment of nutritionalstatus. Br J Nutr 1999;82:165-177.

18 World Health Organization. Prevention and controlof intestinal parasitic diseases. WHO Technical

Report Series No 749. Geneva: WHO, 1987.19 Bell DR. Tropical Medicine. Oxford: Blackwell

Science Ltd, 1995.20 Crompton DWT, Stephenson LS. Hookworm

infection, nutritional status and productivity. In: SchadGA, Warren KS, eds. Hookworm Disease: CurrentStatus and New Directions. London: Taylor andFrancis, 1990:231-264.

21 Zaman V, Loh Ah Keong. Handbook of MedicalParasitology, 2nd edition. Singapore: KC AngPublishing, 1990.

22 Pawlowski ZS, Schad GA, Scott GJ. HookwormInfection and Anaemia. Geneva: World HealthOrganization, 1991.

23 Barnish G, Ashford RW. Strongyloides cffuelleborni in Papua New Guinea: epidemiology inan isolated community, and results of an interventionstudy. Ann Trop Med Parasitol 1989;83:499-506.

24 Adams MF. Fatal strongyloidiasis in young infantsin an isolated rural population in Papua New Guinea.Unpublished report. University of California,Berkeley, 1982.

25 Mata L. Sociocultural factors in the control andprevention of parasitic diseases. Rev Infect Dis1982;4:871-879.

26 King SE, Mascie-Taylor CGN. Nutritional status ofchildren from Papua New Guinea: associations withsocioeconomic factors. Am J Hum Biol 2002;14:659-668.

192


How the poor die in the settlements of Port Moresby, 2003-2004

P. SIMS1

Division of Public Health, School of Medicine and Health Sciences, University ofPapua New Guinea, Port Moresby

SUMMARY

A study of death and dying was carried out in the settlements of Port Moresby in 2003and 2004. This descriptive study of 53 patients indicated that they are dying from chronicillnesses, particularly tuberculosis and HIV (human immunodeficiency virus) infection.They are dying slowly over several weeks with minimum care from the health services.The study also showed the problem of violent deaths by guns, by stabbing, by beatingsand by police in pursuit of criminals. The lack of a well-coordinated ambulance serviceleads to preventable deaths in road accidents and in childbirth.

1 Division of Public Health, School of Medicine and Health Sciences, University of Papua New Guinea, PO Box5623, Boroko, NCD 111, Papua New Guinea

Introduction

There is interest around the world in theproblems of the dying and in terminal care,particularly pain relief, spiritual succour andmaterial aid. 56 million people die every year(1) and if each death affects say 5 otherpeople, then 300 million people or 5% of theworld’s population are affected by deathevery year (2). The problems, both for thecarers and for the dying person, needconsideration, as do the consequences forthe carer and family after death. Country,culture and religious beliefs need to be takeninto account when considering their materialsupport, ie, food, housing, education bills,their ability to cope with the legal andbureaucratic process and the time andimportance of grieving. The care of the dyingis recognized in Europe and the Americasas a considerable and sometimes specialistskill, requiring further postgraduate trainingand learning a new paradigm of care. Thewhole ethos of the ‘hospice movement’ is oneof maximum palliation and minimumintervention (3,4).

Developing countries account for 85% ofthe world’s deaths (2,5). Many children andadults at an age well below their potential

lifespan die and do so quite quickly frominfectious disease in developing countries.They either get better or die within a shortperiod. However, the global demographictransition (6) means that worldwide morecountries have an increasingly elderlypopulation, dying from vascular and heartdisease, strokes and cancers, particularly ofbowel, breast and bronchus. The final illnessis often drawn out, over weeks and months.The distressing symptoms of pain, cough,constipation or diarrhoea, restlessness,anorexia, weakness, wasting and fear arenow also seen more often in developingcountries. The AIDS (acquired immunedeficiency syndrome) epidemic has hastenedthis process, as patients waste away withdiarrhoea, opportunistic infections andimmune failure, while relatives may even tryto conceal their dying from fear and shameof the HIV (human immunodeficiency virus)stigma (7).

Papua New Guinea (PNG) is facing anAIDS epidemic and the HIV seroprevalencein pregnant women is around 1% (8). It isanticipated that in the next 5-10 years AIDSwill increasingly dominate medical practice.Already tuberculosis and AIDS account for70% of the bed occupancy on the hospital

192-201

193


medical wards and together account for over30% of all medical deaths (9) in Port MoresbyGeneral Hospital. It is likely that over the next10 years there will be an increased need toprovide some form of palliative care, as nearas possible to people’s homes, if the hospitalsare not to be overwhelmed by people withHIV/AIDS (N. Muirden, personalcommunication).

There has been little research into the areaof death and dying in resource-poorcountries, although the subject was firstdiscussed in Papua New Guinea more than20 years ago (10). It is important to attemptto estimate the size of the problem in PNGand to discuss its best management (11). Itmay then be possible to consider how bestsuch needs may be met. In PNG, highlyactive anti-retroviral therapy (HAART) hasonly become available in 2004 and there isstill no well-defined system of palliative care.The National AIDS Council Secretariat(NACS) is working with local church groupsto develop a home-based care andcounselling service and a holistic responseto the AIDS problem countrywide.

The settlements and shanty towns arepeculiarly difficult to investigate. Law andorder is a major problem in PNG and thesettlements are particularly unruly so that fewhealth workers venture into them. Thesettlements of Port Moresby accommodatean impoverished and immigrant population,who have come into contact for the first timewith countrymen who differ in language,culture and even appearance. Violenceeasily erupts between people from differentparts of the country living together in squalidconditions.

It is very difficult to know what is going onin the settlements. Many stories may beembellished in the telling and there is mediadistortion. There is no self-policing and littlecommunity spirit. Many among the criminalclasses skulk in the settlements, while thereare sometimes brutal police raids to attemptto capture young men suspected of crime andknown as ‘raskols’.

Methods

This study attempted to describe thepattern of death and dying in the poorsettlements of Port Moresby in 2003-2004.We collected data on adult deaths (aged 15to 50 years) occurring between July 2003 andMarch 2004 – a 32-week period – in 9-mileand surrounding settlements of Port Moresby,Papua New Guinea. The University of PapuaNew Guinea’s School of Medicine and HealthSciences’ Ethics Committee gave permissionfor this study.

There is no regular system of birth anddeath registration in PNG and only if a personis to be buried in a town cemetery will thedeath be registered. In the poor settlementsof the city, deaths may go unrecorded, aperson may be buried locally or the body maybe transported many miles to the homevillage.

The tuberculosis (TB) contact workers arerespected and known in the settlements; theyare usually young men who can operatesafely there. They also know about thedeaths in the settlement and have an informalsystem of communication. The TB workerin this study was a qualified nurse withseveral years’ experience in clinic andhospital work. On some occasions he hadaccess to the patient’s record card or theclinic TB register. Two TB contact workerswere recruited for this study; in carrying it out,one worker did all but one of the interviews.

A questionnaire was prepared whichexamined medical, social and communityissues around death and dying. It attemptedto collect demographic data on age, sex,occupation and marital status and informationabout the last illness, probable cause of deathand the circumstances of dying. Data weregathered about the domestic and socialsituation and the support available to thedying person and their family during and afterthe terminal illness and about the problemsexperienced.

An attempt was made to collect data

194


around symptoms, problems and helpprovided. We devised a scoring system from++ to ––, on the basis of presence orabsence, and then marked any plus scoreas 1 for each of ten symptoms. Thus a patientwho achieved a ++ or + for four differentsymptoms would have an overall score of 4out of 10.

For problems of provision and helpprovided by different sources, ++ (a lot) wasscored 2, + (some) as 1, again out of amaximum of 10, in an attempt to obtain anumerical reflection of the burdens of illnessand approaching death. Plainly, this onlydetermined the problems from the carers’perspective and the patient may have hadvery different feelings and perceptions.

The questionnaire required the interviewerin effect to conduct a verbal autopsy andmake the best estimate of the real cause ofdeath. The questionnaire design wasdiscussed with the interviewers and refinedin practice, and the data that were recordedwere further discussed with the workers aftercollection, to try to obtain a consensus aroundthe probable cause of death. Every effort wasmade to reach a ‘working conclusion’ on theprobable cause of death by this iterativeprocess ie, the interviewer recorded theinformation and made a diagnosis, the datawere examined by the researcher who thendiscussed the recorded information with theinterviewer until both were satisfied that thecause of death, as determined, was asaccurate as possible.

A small pilot study was successful and themain study began in July 2003 and ended inMarch 2004 although the main data collectionperiod was from September to November2003 and in February 2004. Details of deathswere collected in settlements aroundMoresby at 6-mile, 9-mile, 12-mile and 18-mile and at Sogeri. The denominatorpopulation is uncertain, somewhere between20,000 and 30,000, and the aim was to collectdata on all deaths outside hospital betweenthe ages of 15 and 50 (there was someuncertainty about age) occurring in that

population during this period of eight months.Some deaths may not have been reportedto the contact worker and there was a steadystream of inward and outward migrationwhich might have affected death rates.

The questionnaire data were entered intoEpi-Info, checked, cleaned and analyzed. Itwas checked for internal consistency and anyunusual findings and reviewed with theinterviewers.

Results

Demography

In essence the study group is a poorpopulation and nearly half were unemployed.Of the deaths, 64% were in males. The menwere probably less likely to be married thanthe general population and not to have areligious belief (Table 1).

Health, housing and sanitation were asettlement problem. Living conditions werepoor by urban standards in that only 12respondents had an inside lavatory (23%)and only 19 (36%) had electricity in the home.Most people (70%) managed with a tapoutside or possibly used rainwater or a river.Most (>70%) used a pit latrine, shared oftenwith another family. Cooking was over anopen fire or kerosene stove and lighting waswith kerosene lamps and candles.

Cause of death

The predominance of deaths from chronicinfections like TB and AIDS is important giventhe consequences for patterns of death anddying. Violent deaths and road accidentstogether accounted for nearly a third ofdeaths in this group (Table 2).

Duration of illness

In more than a third of the sample the timeto die, from diagnosis to death, was morethan 4 weeks, while for the AIDS/TB deathsthe duration of dying was much greater (Table3, Figure 1). The mean was 6.3 weeks,

195


1ELBAT

SOCIODEMOGRAPHIC ASPECTS OF THE DEATHS INVESTIGATED BY VERBAL AUTOPSY INPORT MORESBY SETTLEMENTS

rebmuN %

shtaedforebmuN 35

xeS

elaM 43 46

elameF 91 63

sutatslatiraM

deirraM 73 07

elgniS 61 03

noigileR

naitsirhC 24 97

cilohtaCnamoR 41 62

demrofeR 31 52

tsitnevdAyaDhtneveS 7 31

rehtO 8 51

naitsirhC-noN 11 12

tnemyolpmE

deyolpmE 82 35

launam-noN 7 31

launaM 9 71

efiwesuoH 9 71

rehtO 3 6

deyolpmenU 52 74

)sraey(egA

naeM 2.63

naideM 5.53

*tekcarbegA

sraey03< 91 73

sraey03> 33 36

* elbailersyawlatondnaetelpmocniatad

196


2ELBAT

CAUSES OF DEATH AND ITS CHARACTERISTICS

htaedfoesuaC )%(oN egA)naem()sraey(

elameF:elaM noitarudssenllI)skeew()naem(

emoH%htaed

noitcefnicinorhC a )3.82(51 5.83 2:3 5.11 39

esaesidralucsaV )5.7(4 55 elamllA 7.1 57

*ecneloiV )0.71(9 1.03 1:2 1< 44

tnediccaciffartdaoR )1.51(8 2.23 1:3 1< 0

amhtsA )7.5(3

recnaC )9.1(1

ycnangerP )7.5(3

)diohpyt(noitcefnietucA )8.3(2

ediciuS )9.1(1

yrecroS )9.1(1

egadlo/nwonknu/rehtO )3.11(6

latoT 35

a erewnoitcefnimorfshtaedcinorhcehtllatahtylekilsitidnaylluferacdekcehcerewshtaedsisolucrebut/SDIAehTnoitanibmocsihtmorf

* tnediccaroepar,redruM

median 3.5 weeks and the standard deviation(SD) 9.5 weeks. This distribution wasskewed because of the number of acutedeaths from violence.

Care requirements

Death is now a long drawn out andincreasingly difficult process and there arerequirements for nursing, medical andspiritual help (Figure 2).

Symptoms

The patients, dying slowly, suffered withpain (32%), diarrhoea (32%) and cough(40%). They were often so weak (60%) thatthey had to be helped to wash, eat and

defaecate. The burden of care, often inassociation with the stigma of HIV disease,in those dying from chronic infections wasgreater. In AIDS deaths these frequencieswere markedly increased so that 66% hadfaecal incontinence and/or diarrhoea and87% cough, while 40% had significant pain.

Place of death

The place of death was most often athome. In this population, 35 (66%) of peopledied at home and 18 (34%) elsewhere, forexample, following a stabbing in the street,or at the roadside after a road traffic accident.For this reason hospital statistics would notcapture these data and it is unlikely thatroutine death certificate data in the city would

197


3ELBAT

ILLNESS DURATION FOR 51 /SDIA TUBERCULOSIS DEATHS

noitaruD

skeew21-5 %06

skeew21> %04

naeM skeew5.11

naideM skeew9

capture it either (Figure 3).

Relationship of informant

The questionnaire informant was a spouse(11%), a child (11%), a parent (8%), a sibling(17%), a more distant relative (13%), a friend(25%) or somebody else in the community(15%).

Caregivers

Most terminal care was provided by thefamily (85%), who struggled to give nursingcare. Health services provided some care(29%) and the church provided care in 59%of cases. Friends and the community alsoassisted (Figure 4). The family in particularprovided help with money for medicines andfood while the person lived and, even moreimportant, contributed to the funeral in cashor kind. Indeed, it was the family whoprovided help with the funeral expenses andother financial needs around the time ofdeath – virtually 100% of the time.

Quantitative analysis

An attempt was made to quantify:

a. Problems of suffering – counting 1 foreach positive symptom to a maximum of10, the mean score was 2.2 (SD = 2.0).

b. Problems of provision – scoring 0, 1 or 2,with 0 = little or none, 1 = some and 2 =much or a lot, for 5 different types of careto a maximum of 10. The mean score

was 2.9 (SD = 2.4).

c. Sources of help – scoring 0, 1 or 2, with 0= little, 1 = some, 2 = much or a lot, for 5main sources of help to a maximum of10. The mean score was 3.3 (SD = 1.9)but nearly all this support was from thefamily.

These data are difficult to interpret andplainly the distress actually suffered by thedying person may be quite different from theimpressions of friends or family. It says atleast that symptom control is oftenproblematic, provision is limited and theburden falls largely upon the family.

Discussion

This paper is a qualitative description ofthe deaths in the settlements of Port Moresbyin an 8-month period in 2003-2004. It is notan epidemiological study, for there is noinformation on denominator andascertainment of deaths was almost certainlyincomplete.

The accuracy of recall may be problematic,when given other than by a family member;but as far as possible information wascollected within 3 months of the death. It isa retrospective and observational study in anattempt to examine a very difficult area.

At present a study of deaths in thesettlements, matched by age, sex anddiagnosis with deaths outside as controls, isnot possible in Port Moresby. There is no

198


Figure 1. Duration of last illness for all deaths.

Figure 2. Care requirements in the dying. The data were obtained from a questionnaire.

information about settlement deaths routinelyavailable; there is no way of tracing peopleaccurately by their address; while there arehuge problems of access and safety for staffin the settlements.

The questionnaire seemed to work welland one person collected most of theinformation. The data showed internalconsistency: for example, young men, asexpected, were more likely to suffer violent

deaths and they were less likely to die athome. While the HIV death rate indicatedthat men were dying younger than women,ie aged 36 years versus 42 years, thisdifference is not significantly different fromthe national figures (8), where women areinfected younger and die earlier.

The settlements probably have a higherdeath rate than the rest of the city as thepeople who live there are poor and lack work,

199


Figure 3. Place of death.

Figure 4. Sources of caregiving support.

and their standard of living is low. Basicservices like water, sanitation, safe cookingand lighting are lacking. The difficulties inthe long-term care of a terminal patient withHIV disease with pain, wasting and chronicdiarrhoea are substantial, particularly in theabsence of any community nursing service.

The settlements are violent and dangerousplaces – beatings, muggings, rapes, holdupsand shootings are commonplace. They areoften unreported and in the absence of anymedical care people quite often die from theiruntreated injuries. There are major problemsin providing 24-hour law and order and health

care in such communities, for there is littlecommunity spirit or ‘self-policing’, while thepolice are distrusted and seen as violent andcorrupt.

This study indicates that the ‘demographictransition’ is underway in PNG with deathsfrom cancer, vascular disease and asthmaincreasing. In PNG there is still a strong beliefin sorcery and traditional medicines. Thus,the traditional healer will often be consultedboth before and after the western doctor. TBand HIV infection are major problems andare causing much suffering in thesecommunities; the findings described here

200


reflect those of other studies (12). Indeed,these problems will certainly increase withtime and need to be urgently planned for.

There is a dearth of accurate data forstandard epidemiological studies in PNG, andthe statistics in the National Health Plan arebuilt around hospital and health clinic data ofarguable quality (13). Diagnosis and deathare recorded incompletely and, for example,the very common diagnosis of malaria isoften not supported by laboratoryconfirmation. There are few data on deathand illness outside hospitals and none for theurban settlements.

In the long term the cost and effort toimprove birth and death registration wouldbe repaid in better planning, developmentand use of services. Meanwhile, repeatingstudies similar to this would greatly assist inmonitoring a developing situation.

People are now dying slowly and havemultiple needs, which are not met by presenthealth and social care services, and relativesplainly struggle to cope. The poor housingconditions make this much more difficult – apatient with terminal AIDS and cryptosporidialdiarrhoea may have 20 or more bowelevacuations in 24 hours (7,14). In theabsence of an inside lavatory, no runningwater and no electric light, the nursingrequired is daunting. The law and orderproblems in the settlements are such thathealth staff are unwilling to provide care inpeople’s homes, as staff have been attackedand female staff sexually abused. TheCatholic Church continues to provide somecare and support and other religiouscommunities continue to visit where they can.

Although a single hospice in Port Moresbywould be plainly insufficient, it would help tochange thinking among the professionalcarers and it would motivate charities andenhance professional skills. An inpatient unitand an outpatient/day-care service couldbegin to serve the population, provided it wasgiven outside funding and could call upon theservices of a religious order to ensure

staffing, eg, The Sisters of Mother Theresaof Calcutta. The proper and committed careof the dying is now an essential part of globalmedicine (1,3,4).

The road traffic accident problem (15)affects the whole of the community – theroads are often in poor condition, drivingstandards variable and cars and buses notserviced. A crowded PMV or open-back truckpacked with people is a major hazard and inthe event of an accident there will be manypeople injured and killed. The road trafficaccident deaths reported in this study are theresult of a single bus crash on the PortMoresby to Sogeri road. Data on roadaccidents are unavailable, but theEmergency Unit at Port Moresby GeneralHospital has a constant daily stream of minorinjuries and usually one major accidentmonthly with two or more deaths every week.There is no reliable emergency ambulanceservice so patients cannot be rapidlytransferred to a hospital. They perish by theroadside from conditions which are eminentlytreatable with standard procedures.

In this study three young women died inchildbirth. The problem again involved lackof rapid transport to the Obstetric Unit at PortMoresby General Hospital and the failure toprovide an efficient and effective 24-hourtransport system.

The problems of living and dying in thesettlements of Port Moresby are highlightedby this study. There is an urgent need towork with the people in the communities, totry to listen to them and work with them. Inthis way the possibility of developing acommunity spirit, real leadership and self-policing is possible. It is on the guarantee oflaw and order and of staff safety that adomiciliary service may in time be built.Without that assurance the poor will continueto suffer and to die and their state will worsen.

ACKNOWLEDGEMENTS

I thank Dr John Millan of the National AIDSCouncil Secretariat, Peter Painakali, who wasan interviewer and a TB contact tracer, Dr

201


Adolf Saweri and Dr Nell Muirden. Theresearch was funded by the National AIDSCouncil of PNG and the National HIV/AIDSSupport Project.

REFERENCES

1 Singer PA, Bowman KW. Quality care at the endof life. BMJ 2002;324:1291-1292.

2 World Health Organization. The World HealthReport 2001. Mental Health: New Understanding,New Hope. Annex Table 2: Deaths by cause, sexand mortality stratum in WHO Regions, estimatesfor 2000. Geneva: World Health Organization, 2001.

3 Billings JA. Recent advances: palliative care. BMJ2000;321:555-558.

4 Twycroft R. Introducing Palliative Care, 3rd edition.Oxford: Radcliffe Medical Press, 1999.

5 World Health Organization. The World HealthReport 1999. Making a Difference. Annex Table 2:Mortality by sex, cause and WHO Region, estimatesfor 1998. Geneva: World Health Organization, 1999.

6 King MH. Demographic disentrapment. http://www.leeds.ac.uk/demographic.disentrapment[accessed on 30 Aug 2004].

7 Adler M, ed. ABC of AIDS, 5th edition. London:

BMJ Books, 2001.8 National AIDS Council Secretariat and

Department of Health. HIV/AIDS Quarterly Report,March 2003. Port Moresby: National AIDS CouncilSecretariat, 2003.

9 Port Moresby General Hospital. PhysiciansReport. Port Moresby General Hospital, Papua NewGuinea, 2003.

10 Hamilton DR. Sori tumas: the decision not to treat.PNG Med J 1982; 25:268-272.

11 Muirden N. Palliative care in Papua New Guinea.Report of the International Association of Hospiceand Palliative Care Traveling Fellowship. J PainPalliat Care Pharmacother 2003;17:191-198.

12 Curry C. HIV prevalence in patients attending thePort Moresby General Hospital Emergency MedicineDepartment. Unpublished report. Port MoresbyGeneral Hospital, Papua New Guinea, 2003.

13 Papua New Guinea Department of Health. PapuaNew Guinea National Health Plan 2001-2010. PortMoresby: Department of Health, Aug 2000.

14 Fallon M, O’Neill B, eds. ABC of Palliative Care.London: BMJ Books, 2001.

15 National Road Safety Council. Slow down – speedkills. Unpublished Report. National Road SafetyCouncil, Papua New Guinea, 2004.

202


Antenatal care in Goroka: issues and perceptions

GAIL L. LARSEN1, SEBEYA LUPIWA2, HELEN PAITO KAVE 2, SUE GILLIEATT1 AND MICHAEL P. ALPERS1

Curtin University of Technology, Perth, Australia and Papua New Guinea Institute ofMedical Research, Goroka

SUMMARY

The high maternal mortality rate in Papua New Guinea indicates an urgent need foraction. One area for examination is antenatal care. From April 2002 to August 2002 aqualitative study was undertaken in order to identify perceptions, beliefs, barriers andstrengths relevant to the utilization of antenatal care by women in the urban, periurbanand rural communities of Goroka, Papua New Guinea. Interview data about antenatalcare utilization were collected from 20 pregnant or parous women and 4 antenatal healthcare workers and relevant statistics were reviewed. This information was analyzed inorder to identify the constraints faced by the users of antenatal care and health careworkers providing such services and to make recommendations aimed to improve theutilization and delivery of antenatal care in Goroka. Multiple encouragers and barriersto using antenatal care were identified within the three categories of physical barriers/encouragers, cultural issues and health care system characteristics. The attitude ofhealth care workers and their perceived ill-mannered treatment of women was one ofthe most significant concerns raised by the women. Nevertheless, all of the womenexpressed overall satisfaction with the care given. All of the health care workers statedthat antenatal care is very important for the health of both the baby and the mother andexpressed a desire to improve the level of care. The major constraints faced were staffshortages, limited supplies and broken equipment. There were four key areas of strength:the broad level of coverage, the high regularity of attendance, the women’s commitmentto antenatal care and the willingness of health care workers to overcome resourcedifficulties in the provision of care. Recommendations to improve the delivery of antenatalcare services and their utilization by women addressed the situation of women and theinteractions between women and health care providers, and proposed innovations inthe health care system.

1 Centre for International Health, Division of Health Sciences, Curtin University of Technology, GPO BoxU1987, Perth, Western Australia 6845, Australia

2 Papua New Guinea Institute of Medical Research, PO Box 60, Goroka, EHP 441, Papua New Guinea

Introduction

A review of available literature revealedthat multiple barriers to antenatal care andhealth care facility utilization have beenidentified in Papua New Guinea (PNG);however, few studies have focused onunderstanding the total antenatal careexperience for women. Gillett (1) in hercomprehensive 1990 review of the health ofwomen in PNG estimated that during theaverage woman’s lifetime she has a 1 in 26

chance that pregnancy or childbirth will causeher death. By contrast, the lifetime maternaldeath risk is 1 in 1500 for the developed world(2). The maternal mortality rate for PNG in2000 remained at around 120 per 10,000 livebirths (3). This situation is dire anddemonstrates an urgent need for improvedmaternal care in PNG. A review of literaturerelevant to PNG identified a number ofbarriers to health care utilization: health caresystem factors, psychosocial and culturalvariables, financial situation, social support

202-214

203


systems and the personal experiences ofwomen with antenatal care (4-6).

The present study was designed to identifyperceptions, beliefs, barriers and strengthsrelevant to the utilization of antenatal careby women in the urban, periurban and ruralcommunities of Goroka in the EasternHighlands Province (EHP) of PNG. Datawere analyzed in order to identify theencouragers and constraints faced by boththe health care workers and the users of theservices. Recommendations were proposedto help improve the utilization and deliveryof antenatal care in Goroka, in both the shortand long term.

Methods

As this study aimed to increaseunderstanding of antenatal care utilizationthrough the analysis of beliefs andperceptions of women and health careworkers, it used a combination ofpredominantly qualitative and somequantitative methods. It was designed as across-sectional descriptive study and wasgenerally retrospective in nature in that itprimarily investigated prior experiences,beliefs and perceptions of women. It wasnot a large-scale or nationally representativestudy but an exploration of beliefs within thechosen community. The informal and openstyle of the interviews aimed to reduce thepossibility that the women would giveinformation that was perceived to be corrector more acceptable to the researchers.

The research design employed a socio-ecological approach which permitted aqualitative focus where data collectioninvolved semi-structured interviews withwomen and health care workers (7). Thisapproach allowed the users’ perspectivesand total situation to be explored within asocial, cultural, psychological andorganizational framework.

Ethical approval was obtained from thePNG Medical Research Advisory Committeeas well as the Curtin University of Technology

Human Research Ethics Committee.Informed consent was obtained from allwomen, health care providers and healthcare facilities prior to data collection. Nopersonal identifying data were recorded andgreat care was taken to ensure confidentialityof the study participants. Interviews wereconducted in tok pisin (Melanesian Pidgin),recorded and translated into English.

Purposive sampling was used to identifythree communities which were known to bereceptive to research projects, andconvenience sampling was employed toidentify women within those communitieswho were willing to participate. Thecommunities were chosen to enablecomparison between urban, periurban andrural communities. Over the four-monthperiod (mid-April to mid-August) in 2002, thefield team conducted semi-structuredinterviews with 20 pregnant or parous womenfrom the selected communities, facilitated 2focus groups with pregnant women,observed interactions of health care workersand their antenatal care clients, interviewed4 antenatal health care workers from threeclinics, and gathered relevant statistics fromhealth care services.

Results

According to the Papua New GuineaDepartment of Health (3), in 1990 in the EHPthe estimated number of live births was10174, the antenatal coverage was 61.5%,slightly lower than the national coverage(64.2%), and the annual number of newantenatal attenders was 6255 women in atotal attendance of 17098 (37%). In 2000the antenatal coverage for EHP (61%) hadnot changed (8). According to attendancerecords at Goroka Base Antenatal Clinic, thetotal attendances at Goroka Base AntenatalClinic have decreased over the last threeyears. In 2000 the total attendance was 2712women, in 2001 it was 2425 and in 2002,from January to June, there had been 1158women (which gives an estimate, based onprevious years, of 2326 for the year).According to records at Goroka Hospital

204


there were 2855 live births at Goroka Hospitalin 2001. Maternal deaths for EHP were 12in 1990 (3), 24 in 1998, 16 in 1999 and 17 in2000 (8).

The data revealed four key areas ofinterest: the women’s reasons for attendingantenatal care, the women’s perceivedbarriers and encouragers to attending, theirexperiences of antenatal care and the healthcare workers’ beliefs and perceptions aboutantenatal care services.

Reasons for attending antenatal care

The reasons that women gave forattending antenatal care fell into threecategories: to seek out care for themselves,to seek out care for the unborn child and toreceive education or expected healthtreatment. 11 of the 20 women expressedmultiple reasons for attending antenatal care,whereas the others expressed only onereason.

In general, the women put a high value onattending antenatal care. The most commonreason given for attending antenatal care wasto receive information regarding the state ofthe unborn child (13/20). This was expressedin terms of finding out the lie of the baby andif there was anything wrong with the babyand ensuring that the baby would be bornhealthy. Another common reason (12/20)given for seeking out antenatal care was toreceive medication and nutritionalsupplements. Equally common (12/20) wasthe reason that the women wanted to receivegeneral antenatal care and to discover anysickness in themselves. It wasacknowledged among the women that therewas a possibility of complications duringpregnancy and that going to antenatal carewas a way of finding out whether or not theywere healthy in their pregnancy.

Two reasons for attending antenatal caregiven by a smaller number of women were adesire to secure a place to give birth at thehospital, which they believed would befacilitated by attending for antenatal care, and

to receive health education. Education wasmostly stated in a general sense and not forthe receipt of specific information.

The average time of first visit for womenon their first pregnancy was the fifth monthwith 4 of the 20 women going for antenatalcare as soon as they missed a period orwithin the first month. However, some errorin recall of this information is likely as somewomen had had their first pregnancy morethan ten years previously. In subsequentpregnancies, there was a pattern ofpostponing the first visit to the seventh oreighth month with only 3 mothers stating thatthey had gone earlier than the fifth month fortheir initial visit. Most women stated that insubsequent pregnancies they had gone fortheir first visit closer to the time of delivery.

All of the women attended regularly oncethey had an initial visit, following therecommended return date given to them bythe health care worker. The only exceptionto this was when women encountered anobstacle that delayed their visit. No womanreported not continuing to attend antenatalcare once she had started.

Barriers and encouragers to attendingantenatal care

Barriers

The most common barriers to utilizingantenatal care fell into the three categoriesof physical barriers, cultural issues and healthcare system characteristics.

“It is hard to come here. There are manythings that make it hard like findingtransportation, finding money to pay for thebus, making sure my children are lookedafter, and leaving more work for others athome.”

Of the 20 women, 9 stated that they hadno difficulties or felt no barriers to attendingantenatal care, while some women describednumerous barriers. All of the women whostated that they felt no barriers to attending

205


antenatal care lived within the urban andperiurban areas. It is interesting to note that,despite the women perceiving that there wereno obstacles for themselves, they wereaware of and concerned for the women inother areas who they knew did encounterbarriers.

“When I was pregnant it was easy for meto go for antenatal care because I live closeto the hospital. But for my sister who livesoutside of Goroka it is hard for her to findtransportation and to get the money fromher husband to buy her way to antenatalcare. She also does not have water like Ido, that is easy to wash with, or the timeto spend the whole day away from herfamily and garden. I am sorry for thewomen who live outside of Goroka.”

The primary physical barriers to thewomen were distances that needed to betravelled, which were difficult because of thelack of public transportation, lack of moneyto pay for transportation or physical weaknessthat made it difficult to walk the lengthydistance to the clinic. Women in the ruralarea in particular found that distance was abarrier to getting to antenatal clinics. Thewomen of the periurban area saw distanceor transportation as less of a barrier whilstnone of the women from the urban area hadany problem with distance or transportation.All of the 10 women interviewed in the ruralareas described that they had to walk all theway to the clinic, or walk some of the wayand then find public transportation to get intotown, followed by more walking or gettingother public transportation from the bus stopto the hospital. The women in the periurbanand urban areas stated that they walked tothe hospital because it was not far away andonly used public transportation when it wasraining or they were not feeling well. Moneywas a physical barrier in that the women didnot always have the money needed to payfor transportation and were not able to obtainmoney easily.

The expected roles of women within theculture presented some obstacles since

women found it more difficult to take theirother children with them to antenatal care orto find others to look after their children.Additionally, at the time of a relative’s deathor fighting within their village, it was notculturally acceptable for the women to leave.Another cultural issue raised by one womanwas that she missed one of her visits becauseher relatives were angry with her and hadput a ‘curse’ on her. When her relatives wereno longer angry with her she was again ableto attend antenatal care.

The health care system barriers primarilytook the form of long waiting times, negativeattitudes of health care workers, occasionaland unannounced closures of clinics, noantenatal health care worker present and lackof nutritional supplements or medications.

Encouragers

The most common encouragers fell intothe same three categories of physicalencouragers, cultural issues and health caresystem characteristics. All the women namedat least one thing that they felt encouragedthem or made it easier for them to attendantenatal care.

The main physical encourager to thewomen in the urban and periurban areas wasthat the clinic was close enough that theywere able to walk to it and did not need torely on having money or transportation. Allof the women of the periurban and urbanareas made similar statements about the wayin which they got themselves to antenatalcare.

The women who indicated that theywalked to the clinic also talked about howthe exercise strengthened them and helpedthem be healthy; and in one woman’s caseshe believed that it helped her have an easierlabour.

Another physical encourager as well as acultural encourager described by severalwomen in the rural areas was that they hadfamily to look after their other children. The

206


close kinship system of PNG is a strength inthat it assists women to receive support andassistance from their families. Most womendid report that their husbands, relatives orextended family were able to and did assistthem to receive antenatal care while theywere pregnant. The provision of socialsupport in much of PNG is a culturally feltobligation and is heavily depended upon. 15of the women stated that their husband wassupportive of them during their pregnancyand that he helped them with such things ashousehold duties, money, food, clothing andchildcare. 6 women said that their in-lawswere supportive in similar ways and 6 womenstated that their own families were supportive.

It was interesting to find that the womenin the periurban settlements stated that theyreceived less help from family and in-laws.This is probably because the families that areliving in settlements have moved there froma village and away from their extendedfamilies. All of the women who stated thatthey did not have enough support or found itdifficult to find childcare were living in theperiurban areas.

As all of the women in the study receivedantenatal care during one or more of theirpregnancies, each of them had access to aclinic which provided antenatal care. Theprovision of antenatal care in clinics was anencourager to women being able to receivethe care. The women also perceived thehealth care providers’ work as important andbeneficial and sought out their care.

Perceptions about antenatal care

All of the women expressed overallsatisfaction with the care given. Most oftenthey expressed satisfaction in terms of havingreceived the services that they felt wereimportant as well as reassurance that theirpregnancies were without complication.Some of the women expressed satisfactionin having the opportunity to receive antenatalcare because of its importance in pregnancyoutcomes.

“I thought about if I would survive or if mybaby would survive. If something wentwrong, it would be easy for one of us todie. I think that the nurses do good work,and important work. They help bychecking the baby, checking me, givingmedicines. Even though they get crosssometimes it is good work that they do.”

Despite the high level of satisfaction withservices received, it is interesting to note that14 women also brought up the issue of beingdissatisfied with the attitudes of or treatmentby the health care workers. This was themost common concern that the womenraised regarding the care that they received.

“If a nurse is angry at me she should talkto me quietly inside the room on my own.This practice of theirs to get angry at mewhen I am outside with all the other womenis not a good one.”

The health staff were described asshaming them and getting angry with themfor not having washed before attending fortheir visit, for not wearing convenient clothing,for having too many children and/or formissing a visit. Shaming or condescendingattitudes of health care workers were outlinedas a concern in health education talks, advicegiving and the collection of patientinformation.

Additionally, women reported beingconcerned with the long waiting time betweenthe time that they arrived and the time ofcompleting their visit as well as theinconvenient waiting area (when it was hot,rainy, cold, wet etc). Other issues that wereraised by women were lack of easycommunication with health care workers,high fees (20 kina) for delivery at the hospital,that nurses in training were performingantenatal care assessment activities withoutdirect supervision of a nurse and that malenurses in training were performing antenatalcare.

In addition to the information on antenatal

207


care provision, the women expressedopinions regarding the provision of labourand delivery care and some heartfelt issuesregarding the care that they received. Fromthe discussion on labour and deliverypractices and the women’s beliefs andperceptions, it was clear that there is a needfor further research and understanding ofwhat women experience in labour anddelivery.

Health care workers’ beliefs andperceptions

The four health care workers reportedproviding similar services in antenatal careand holding similar beliefs about antenatalcare and concerns regarding theinadequacies of antenatal care. All the healthcare workers said that antenatal care wasvery important for the health of both the babyand the mother. The barriers to providingcare were outlined by all four health careworkers as staff shortages, limited supplies(of medicines, water and educationalmaterials) and broken equipment. All the fourhealth care workers believed that womenshould be seen regardless of when theycome or how they come so as to provide carewhen it is most convenient to women. Thisis in conflict with some of the women’scomments of being turned away because ofcoming too early in pregnancy, not comingon their scheduled visit, or because they hadfailed to wash themselves before coming.

Another barrier nominated by the healthcare workers was the effect of inadequatefacilities for providing good quality care.These inadequacies included the lack of toiletfacilities, limited or no privacy and inadequatewaiting areas for the women. Additionally,the health care workers indicated that theyfelt that there are too few clinics providingantenatal care and therefore the patient loadon them was too high. Having a shortage ofhealth educational materials was alsodescribed as a barrier as was the limitedamount of time available to provide healtheducation. A shortage of midwives to provideantenatal care and to supervise nurses and

community health workers, as well as asevere shortage of female doctors, were alsoidentified as important problems. The fourhealth care workers desired more inservicetraining in order to further their own educationand ultimately benefit their patients.

Despite the challenges to providingantenatal care, all the health care workersexpressed a desire to improve the level ofcare and a personal belief in the importanceof quality antenatal care.

Discussion

Though Papua New Guinea is very diversein languages, cultures and the livedexperiences of women, there are commonthemes in the interaction of people with thehealth services. Therefore, despite the smallsample size drawn from a geographic areacontained within 15 kilometres of Goroka, webelieve that the results and conclusions ofthe study do have wider implications.Nevertheless, they need to be generalizedwith caution.

All the women interviewed in the study hadused antenatal clinic services. The rate ofutilization of antenatal services in the GorokaDistrict is high, which makes it an appropriateplace to conduct such a study and we believeit has identified important information.However, the attitudes of non-users and theadded difficulties of more remote andmarginalized women have not beeninvestigated. Though the factors identifiedhere are likely to be widely applicable, theirrelative weight may be quite different in lessfavoured areas.

Strengths of antenatal care

Numerous strengths of antenatal carewere identified by the four health careworkers and the users of antenatal care. Thefirst of four strengths was that all of thewomen interviewed had received someantenatal care during pregnancy and all ofthem stated a positive level of satisfactionwith or placed high value on antenatal care.

208


However, this was a small sample of self-selecting volunteers and their responses mayhave been biased because the researcherswere perceived as connected with the healthcare system. Nevertheless, women puttingvalue on receiving antenatal care and makingan effort to receive it demonstrated that theantenatal programs were effective in drawingwomen to them. In addition, the lack of feesfor antenatal care ensures that there is notan additional financial burden in attending theantenatal clinic.

Secondly, it was commendable that oncethe women started attending for antenatalcare they continued to attend on a regularbasis. The women did not have a pattern ofgoing only once or occasionally. The overallopinion of the women was that it wasimportant to attend according to the schedulegiven to them and they attempted to returnon the given date and to overcome anybarriers to attending.

Similarly, the third strength is that of theingenuity employed by women in overcomingbarriers in order to seek effective care andto assist others. The commitment of womento obtaining antenatal care was shown by onewoman who sold a pig in order to have moneyfor antenatal care and delivery. Anotherexample is illustrated by the following story.

“When I was nine months pregnant myvillage was fighting with the next village.They had blocked the roads and no buseswere coming. I had to go to the clinic formy antenatal visit, and as you know it is along walk from here. But with no buses Ihad to walk. It was not easy to leave thevillage with fighting; I was scared becauseI thought that with all the walking I mighthave my baby alone in a ditch. I did get tothe clinic and everything was okay but itwould have been easier for me to go to aclinic that was nearer if there was one.”

The fourth strength was that health careworkers, in spite of providing antenatal careto large numbers of women, were able tomaintain the level of care given the limited

resources and shortage of staff. Despite thetime constraints, staff shortages and limitedsupplies it was observed that the activitiesprovided during antenatal care wereconsistent with the standard of minimum careadopted by the World Health Organization(WHO) Technical Working Group onAntenatal Care (9). These observations werefurther supported through the interviews withthe women and health care providers.

Barriers analyzed

One of the key findings was that thewomen, particularly in the rural area,encountered multiple barriers in theirendeavour to receive antenatal care. Oneof the main sociodemographic barriersaffecting the rural women in this study wastheir low income and very limited financialresources. This was complicated by the highnumber of visits that the women wereexpected to make. The recommendedschedule was for monthly visits until 28weeks’ gestation, then every two weeks from30 to 36 weeks and weekly thereafter. Asmost women in this study did not startattending antenatal care until after thesixteenth week, they very soon had toundergo return visits on a fortnightly basisfollowed closely by the weekly visits. Thispresented a financial barrier for some of thewomen and took a lot of their time.

Another sociodemographic barrier washaving a low level of education. To movecloser to town and health care facilities wouldmean having some kind of formal income,which was more likely if a woman or herhusband had more formal education. Thisview is supported by the findings of Klufioand Kariwiga (4) in that the women in theirstudy who came from a lower socioeconomicclass and whose husbands were lesseducated received less antenatal care. Theyalso found that the women with higher paritytended to receive less antenatal care.

A situational factor that affected many ofthe women in the rural area was that ofaccess to antenatal care. The limited

209


availability of transport and the longerdistances to travel to health care facilitieswere barriers to attendance for rural women.This was particularly true of the high-riskwomen who were unable to get the care thatthey required at closer clinics and had totravel into Goroka on a regular basis.Compounding this barrier were psychosocialbarriers such as physical tiredness duringpregnancy that makes walking long distancesmore challenging and the high workload ofthe women in gardening, childcare andactivities of daily living. Embedded in thewomen’s workload were their expected rolesto support extended family and theircommunities in all customary obligations andactivities.

The National Sex and ReproductionResearch Team and Carol Jenkins (10) foundthat traditional beliefs about pregnancy,delivery complications and illnesscontradicted western medical beliefs. Thefindings of the present study did not revealmany instances where the traditional beliefswere contradicting western-style antenatalcare. However, there were beliefs regardingfertility and delivery which were contrary towestern medical beliefs. Despitecontradicting beliefs, the women still soughtand received antenatal care and health care.Social events with strong cultural elementssuch as a death in the community, clanwarfare and social unrest also createdbarriers for the women. The main reasonsthat women gave for attending were thesame as those found in the national study(10). Women in both studies attended mostcommonly in order to find out about the statusof their baby. This demonstrated that thewomen recognize the benefits of finding outwhat the western medical model has to sayabout the health status of their baby, whichin turn acknowledges the role that the healthcare system has played in informing peopleabout the benefits of antenatal care.

Attitudes and actions of health careworkers

The attitude of health care workers and

their perceived ill-mannered andunsympathetic treatment of women was oneof the most significant concerns raised by thewomen. Garner et al. (5) similarly found intheir study on infant mortality and antenatalattendance in a rural area of PNG that theattitudes of health care professionals asperceived by clients were a barrier toutilization. In developing and developednations, the quality of health care services isaffected by the unprofessional andunsympathetic attitudes of health careworkers (11), which in turn can negativelyaffect the desire of people to use healthservices.

Analysis of observations and interviews inthis study revealed why women perceived theattitudes of health care workers negatively.Health care workers were seldom observedsmiling, asking open-ended questions,making eye contact or delivering healtheducation in a friendly conversationalmanner. Most observations of therelationships between the health careworkers and the women revealed that thewomen did not ask questions, but did listento what was said to them, and that healthcare workers had little time for relationshipdevelopment. Even though the observationsand reports of women made it apparent thatthis is clearly an area requiring change,health care professionals consistentlyreported their desire to provide good care forwomen, and women repeatedly reportedsatisfaction with the overall care in spite ofthe perceived negative attitudes of healthcare workers.

We speculate that these perceivednegative attitudes are ‘something to be livedand dealt with’ or ‘expected’ and theiracceptance has its roots deep in the mentalityof accepting the directions and advice froma ‘superior’ or a ‘more educated person’,particularly in relation to public or governmentservices. This is complicated by the culturalrequirement to respect others who areviewed as more knowledgeable, which in thiscase is expressed through listening, and adesire to avoid being shamed; therefore, the

210


women remain silent.

However, this view is also complicated bythe importance of relationships and thedependency of people on one another inPNG. Repeatedly it was observed thatpeople go to great lengths not to offend eachother. Nevertheless, given this high valuesystem, it is surprising that health careworkers did not give more attention to thedevelopment of relationships with theirclients. Even though there is often no timein a busy clinic for relationships to bedeveloped, a smile, friendly body languageand reassuring remarks could improve thesituation dramatically. However, the teachinggiven to health care workers about a positiveattitude and other improved ways of relatingto clients and patients is rarely reinforced inclinical practice. As there are no ‘marketforces’ operating in PNG there is very littlechoice about where to attend for antenatalcare. If the attitude of a health care workeris not liked, in most cases there is no choiceabout going elsewhere. We suspect that thissubtly alters the attitudes of both health careworkers and clients. Another explanation asto why health care workers shame or demeanwomen is that it is part of an endeavour toincrease compliance on the assumption thatwomen will then be more keen to comply withrequests so as to avoid public shaming. Thiscould be a result of the perceived powerstatus of health care workers and theexercise of this power over female clients.

Health care system

The utilization rates of the three health carefacilities studied were higher than the nationalaverage reported by Gillett (1) in her 1990review. This is in part because an area withrelatively high antenatal care coverage waschosen for the study and perhaps alsobecause of changes in health care over thelast ten years. Access to antenatal care inand around Goroka was not as difficult toachieve as it would be in more rural areas ofPNG.

We believe that the high utilization rates

at the clinics in this study add to thefrustrations of the overworked staff as wellas further deplete very limited resources.Failures in the health care system in theareas of staff shortage and the limitedavailability of resources have been variouslyattributed to the process of decentralizationand recentralization (12), to the restructuringof provincial and hospital administration aswell as to detrimental policy (6,13-15) and tothe poor training of health care workers (5).

In spite of the barriers to improving thequality of health care, we believe that theantenatal care delivery system has survivedas well as it has because of the commitmentof health care workers and PNG’s previoushistory of good health care services. Thelevel of antenatal care in this study was withinthe acceptable range as laid out by the WHOTechnical Working Group on Antenatal Care(9). The areas and issues identified forimprovement in the delivery of antenatal carewere the lack of options for safe home births,the minimal levels of health promotion, thelack of maternal counselling on the dangersigns and symptoms of pregnancycomplications, the limited counselling onsmoking, alcohol and betelnut cessation andthe failure to carry out urinalysis.

Previously, there have been attempts atimplementing the use of village-trained birthattendants and village health workers.However, these have met with variedsuccess and have encountered multiplebarriers such as the lack of receptivenessby communities, cultural beliefs and taboos,issues of payment and supervision, and lackof monitoring and evaluation (16-18). Therehave also been positive aspects andfavourable outcomes in some of theseprograms which may assist in increasing thesuccess rate of village birth attendant andhealth worker programs if they were to betried out again.

The difficulties that health care workersface in providing sufficient health promotionactivities of good quality due to the limitedavailability of time, staff shortages,

211


inadequate training and lack of resources areclearly understood. Health promotioneducation can also prove challenging in thatthe messages are difficult to express cross-culturally, they contradict traditional beliefsor they are hard to simplify for generalunderstanding. Additionally, healthpromotion messages are met with resistancefrom women when they are delivered in anegative and demeaning way.

Clearly there are costs of both providingand receiving health care, but given thereviving strengths of the health care system,the beliefs and perceptions of women andthe desire to provide good care, there is roomfor much hope. We believe that the followingproposed recommendations are feasibleand, if implemented, will have a positiveeffect on the delivery and utilization ofantenatal care.

Recommendations

Our recommendations for improving theutilization and delivery of antenatal careservices in Goroka District fall into threecategories:

• recommendations applicable towomen

• recommendations at the relationshiplevel between women and health careproviders

• recommendations at the health caresystem level.

Improving women’s situation

One of the critical components ofimproving the reception of health care bywomen in developing countries is toempower women to advocate forthemselves. The situation for women in PNGis similar to that in many developing countriesin that they are less educated than men, havea disadvantaged social position, lack thefinancial resources for autonomy and have

a general sense of powerlessness (1,11).One basic and fundamental way ofempowering women is to increase theavailability of education and ease theiraccess to it. In a paper regarding educationand women in PNG presented at the WaiganiSeminar in 1997 (19), Lady Carol Kidustressed that education is a major means offreeing women from poverty and oppressionwhile raising their socioeconomic standards.Likewise, Gillett (1) points to education asbeing one of the most important factorsaffecting the health of women and theirchildren in PNG. The education andempowerment of women can then bringabout more community involvement andownership of better health and well-being atthe local level, resulting in a stronger demandfor health care facilities to improve.

Interaction with health care providers

We suggest that implementing villagehealth worker programs will bring health carecloser to the villages, provide culturallyacceptable care and give communities asense of ownership. These programs,however, need to be supported by strongcommitment from the community and byhealth care workers and government healthofficers to ensure that the programs aresustainable. A well-trained village healthworker, if properly supported, can providebasic health coverage for the entirecommunity (not just women) as well as be avaluable asset to health promotion.

An important area of education is thecontinued education of nurses, healthextension officers and doctors about effectiveinterpersonal communication, positive patientrelations and skills to deliver effective healthpromotion messages. There are manyresources available to assist health careworkers in learning to help communities,communicating effectively with others anddeveloping creative ways of deliveringeffective health promotion. One such text is‘Helping Health Workers Learn’ by DavidWerner and Bill Bower (20). Another is‘Training for Transformation: A Handbook for

212


Community Workers’ by Anne Hope andSally Timmel (21). Continuing education ofhealth care workers will help them to learnthat building relationships is a process. In acountry where the majority of the populationlive in rural areas, it is critical that health careworkers are able to lead and train others tolead communities and women through theprocess of looking at their resources,identifying their problems and learning to askquestions. Creative use of health promotionresources beyond those requiring literacy(posters, billboards, brochures) such asdrama, role-playing, songs, dance and story-telling are invaluable. Additionally, it is criticalfor more research to be done into theeducational programs of nurses, healthextension officers and doctors to ensure thatthey are being taught a curriculum that isrelevant to the situation of the people thatthey will be serving.

Health care system level

It is generally acknowledged that anincrease in the number of health careworkers, particularly for the aid posts thatserve as the primary entry point to the healthcare system, is necessary. This could beaddressed through effective lobbying in PortMoresby by community representatives,increasing funding to the mobile units,training of village health workers,implementing community health teams, andthe active involvement of schools.

There is also a need for an increase in theallocation of funds from provincialgovernment, national government and aidprojects to the public health sector to ensuresupplies and support for programs. This maybe difficult in the current financial climate, butthe importance of health to development isbecoming more widely recognized, evenamong traditional aid donors. Specificfunding to kick-start, or restart, community-based health programs should be sought,either by the government or by non-government organizations, includingcommunity-based organizations. Successfulcommunity-based programs should be given

more acknowledgement and support andwidely promoted as good models to befollowed.

These two recommendations are veryimportant but may take time to implement inthe face of current resource constraints.However, since the high maternal mortalityrate indicates an urgent need for action, whatcan be done now in the short term, withpresent resources, to improve antenatal careservices for women?

The first such recommendation, whichcould be implemented immediately, is toreduce the standard number of antenatalcare visits to four (for normal pregnancies)while encouraging starting antenatal careearly in pregnancy according to therecommendations of the WHO TechnicalWorking Group (9) for the minimal level ofcare for a normal pregnancy. However, it iscritical that with the reduction in the numberof visits a reduction in services does notfollow. Efforts must be made to ensure thatthe essential services are provided and it isequally important that the psychosocial andmedical needs of the women are addressed.This would allow health care workers toprovide any necessary interventions earlier,have more time for health education, reducethe number of barriers to women, decreasethe daily patient load on staff and give healthcare workers the opportunity to make the bestuse of the longer periods of time spent witheach client or patient. Increasing healtheducation during antenatal care and withinthe general community will promote generalknowledge about pregnancy and healthissues.

In addition to reducing the number ofantenatal care visits it is important toindividualize care. Individualized care andcounselling have been shown to empowermothers and their families to make informeddecisions regarding their health care (22).Individualized care plans are often used inwestern countries to enable health careworkers to communicate effectively with eachother and to enable the clients to be involved

213


in the care that they receive. Areas of needare mutually identified by the health careworker and client and agreed upon solutionsare proposed. This provides for continuity incare and the setting of goals. Care plansand individualized care also aid health careworkers to look at the total situation of theclients and not just at their physical healthstatus. This type of case management inantenatal care can have a positive influenceon how mothers view antenatal care as wellas a positive effect on changing lifestylebehaviours (23).

The final recommendation is the need toaddress the conditions and attitudes of healthcare workers. Through local action andleadership it may be possible to improve theconditions, morale and attitudes of healthcare workers by the creation of a local healthombudsman who would handle complaintsand reward praiseworthy performance. Inaddition, the ombudsman could provideindividual incentive for health workersdemonstrating responsibility to the public.The ombudsman would be available forclients to share their negative experienceswith as well as praise the performance ofhealth workers. The concept of anombudsman is well established in PapuaNew Guinea and the National Ombudsmanis a highly respected and politicallyindependent figure, whose actions are widelyand sympathetically reported in the media.It should be stressed that it is important thatthe local health ombudsman be createdthrough local action and leadership and notas a government tool, though the support inprinciple of the health authorities at provincialand national levels will have to be won andwill be essential to its success.

ACKNOWLEDGMENTS

We are very grateful to the women andthe health care providers who willingly gaveof their time and openly shared theirexperiences of antenatal care. We thank thePapua New Guinea Institute of MedicalResearch for assistance and for the supportgiven by its staff. In particular, we thankProfessor Charles Mgone for his advice and

his willingness to make resources available.

REFERENCES

1 Gillett JE. The Health of Women in Papua NewGuinea. Papua New Guinea Institute of MedicalResearch Monograph No 9. Goroka: Papua NewGuinea Institute of Medical Research, 1990.

2 Mahler H. The safe motherhood initiative: a call toaction. Lancet 1987;1:668-670.

3 Papua New Guinea Department of Health. PapuaNew Guinea National Health Plan 2001-2010. PortMoresby: Department of Health, Aug 2000.

4 Klufio CA, Kariwiga G. Booked and unbookedmothers delivered at Port Moresby General Hospital:a randomised case control study of theirsociodemographic and reproductive characteristics.In: Taufa T, Bass C, eds. Population, Family Healthand Development. Proceedings of the NineteenthWaigani Seminar, Port Moresby, 16-22 Jun 1991,Volume 2. Port Moresby: University of Papua NewGuinea Press, 1993:188-193.

5 Garner P, Heywood P, Baea M, Lai D, Smith T.Infant mortality in a deprived area of Papua NewGuinea: priorities for antenatal services and healtheducation. PNG Med J 1996;39:6-11.

6 Duke T. Decline in child health in rural Papua NewGuinea. Lancet 1999;354:1291-1294.

7 Sword W. A socio-ecological approach tounderstanding barriers to prenatal care for womenof low income. J Adv Nurs 1999;29:1170-1177.

8 Eastern Highlands Provincial AdministrationOffice of Health. Annual Report for 1998, 1999 and2000 of the Eastern Highlands ProvincialAdministration Community Health ServicesProgram. Goroka: Eastern Highlands ProvincialAdministration, Jun 2001:23.

9 World Health Organization. Antenatal Care.Report of a Technical Working Group, Geneva, 31Oct-4 Nov 1994. WHO Document No WHO/FRH/MSM/96.8. World Health Organization, Geneva,1996.http://www.who.int/reproductivehealth/publications/MSM_96_8/MSM_96_8_table_of_contents.en.html[accessed Feb 2002].

10 National Sex and Reproduction Research Team,Jenkins C. National Study of Sexual andReproductive Knowledge and Behaviour in PapuaNew Guinea. Papua New Guinea Institute of MedicalResearch Monograph No 10. Goroka: Papua NewGuinea Institute of Medical Research, 1994.

11 Basch PF. Textbook of International Health, 2ndedition. Oxford: Oxford University Press, 1999.

12 Campos-Outcalt D, Kewa K, Thomason J.Decentralization of health services in WesternHighlands Province, Papua New Guinea: an attemptto administer health service at the subdistrict level.Soc Sci Med 1995;40:1091-1098.

13 Kakazo M, Lehmann D, Coakley K, Gratten H,Saleu G, Taime J, Riley ID, Alpers MP. Mortalityrates and the utilization of health services duringterminal illness in the Asaro Valley, EasternHighlands Province, Papua New Guinea. PNG MedJ 1999;42:13-26.

214


14 Thomason J, Mulou N, Bass C. User charges forrural health services in Papua New Guinea. SocSci Med 1994;39:1105-1115.

15 Thomason JA. The implications of uncontrolledpopulation growth for the provision of rural healthservices. In: Taufa T, Bass C, eds. Population,Family Health and Development. Proceedings of theNineteenth Waigani Seminar, Port Moresby, 16-22Jun 1991, Volume 1. Port Moresby: University ofPapua New Guinea Press, 1993:287-300.

16 Albu R, Alto W. Training of village midwives in theSouthern Highlands Province. PNG Med J1989;32:89-95.

17 Garner PA. Voluntary village health workers inPapua New Guinea. PNG Med J 1989;32:55-60.

18 Wells MM. Midwifery services in Madang Province,Papua New Guinea: a proposal. PNG Med J

1985;28:147-153.19 Kidu C. Information and women in Papua New

Guinea. Paper presented at The Waigani Seminar,Port Moresby, 27 Aug-3 Sep 1997. http://www.pngbuai.com/600technology/information/waigani/ info-women/WS97-sec12-kidu.html[accessed Oct 2002].

20 Werner D, Bower B. Helping Health Workers Learn.Palo Alto, CA: Hesperian Foundation, 1982.

21 Hope A, Timmel S. Training for Transformation.London: ITDG Publishing, 1995.

22 Jahn A, Dar Iang M, Shah U, Diesfeld HJ.Maternity care in rural Nepal: a health serviceanalysis. Trop Med Int Health 2000;5:657-665.

23 Issel LM. Women’s perceptions of outcomes ofprenatal case management. Birth 2000;27:120-126.

215


Rehabilitation services in Papua New Guinea

AMELIA SHAW1

Angau Memorial General Hospital, Lae, Papua New Guinea

SUMMARY

It is now accepted that in developing countries community-based rehabilitation (CBR)is the most effective way to meet the needs of the disabled. The proportion of thepopulation of Papua New Guinea (PNG) having access to CBR is not known. The purposeof this project was to clarify the extent of rehabilitation services in PNG. It was hopedthat by establishing the extent of services, communication and cooperation betweenthem would increase, leading to more efficient and effective use of the limited resources(human and otherwise) available for rehabilitation in PNG. A questionnaire was sent toall known existing rehabilitation services, all provincial health departments, provincialhospitals and church health services. A 47% response rate was achieved. Resultsshowed that most provinces have some form of rehabilitation available but rehabilitationis not spread equally throughout PNG. Most of the services are based in the urbancentres and the rural population is badly served except in the Highlands Region and theSepik provinces, which appear to be more comprehensively served by CBR. The servicesthat do exist are hampered by lack of human and material resources and difficultyaccessing clients due to transport difficulties. There needs to be a greater movement ofrehabilitation into the community with government backing. The greatest effort is beingmade by Callan Services for Disabled Persons based in Wewak with its group of SpecialEducation Resource Centres that also carry out CBR. CBR would appear to be anappropriate way to address the needs of the disabled PNG population, but in order to besuccessful it requires greater backing and more trained personnel.

1 Angau Memorial General Hospital, PO Box 457, Lae, Morobe Province 411, Papua New Guinea

Present address: 83 Wheatfield Way, Cranbrook, Kent TN17 3NE, United Kingdom

Introduction

Although rehabilitation has beenestablished in Papua New Guinea (PNG) forsome years it has not been a priority for analready overstretched Health Departmentthat has looked more at curative health thanpreventive health. Church and non-governmental organizations (NGOs) have todate carried out most of the rehabilitation.The National Health Plan for 2001 to 2010states that rehabilitation services should beexpanded in collaboration with other sectors(1). There are currently a number ofrehabilitation services throughout PNG, butwithout governmental regulation andmonitoring the establishment of these

programs has been haphazard and has oftenoccurred without the government’sknowledge. As a consequence there areseveral organizations striving to improve therights of the disabled in PNG working inisolation or in ignorance of each other.

In recent years disability has progressedfrom being seen purely as a medical problemto more of a problem of society and its lackof acceptance of disabled persons. Inrecognition of this the World HealthOrganization (WHO) has published theInternational Classification of Functioning,Disability and Health (2). This documentmoves away from classification based onconsequence of disease to classification

215-227

216


based on components of health. Thedocument takes a neutral stand on aetiologyand includes environmental factors that canaffect the way in which a person lives.

Rehabilitation can be described as therestoration to maximal functional potential ofan individual following illness, injury oroccurrence of impairment. This can lead toimproved quality of life for the individual, lesspressure on family and social services (if theperson is able to do more for themself) andupholds the right to equal opportunities.

To meet the needs of disabled peopleworldwide, and to reflect the change ofdisability from a medical to a social problem,rehabilitation has also moved from aninstitution-based to a community-basedconcept. This has been successful in manycountries around the world. Community-based rehabilitation (CBR) is defined by theWHO, International Labour Organization(ILO) and United Nations Educational,Scientific and Cultural Organization(UNESCO) as:

“A strategy within community developmentfor the rehabilitation, equalisation ofopportunities and social integration of allpeople with disabilities. CBR isimplemented through the combined effortsof disabled people themselves, theirfamilies and communities, and theappropriate health, education, vocationaland social services.” (3)

Research into the effectiveness of CBR islimited (4) so that the success of CBRprograms is not often acknowledged. CBRis now, however, accepted as the best wayforward for rehabilitation in developingcountries as it addresses both the physicaland social needs of those living with adisability.

The social role of CBR is also relevant tothe United Nations Standard Rules on theEqualization of Opportunities for Personswith a Disability adopted in 1993 (5). Thisshows that people are beginning to

understand disability as a human rights issue.This resolution also states that rehabilitationshould be available to all disabled personswho require it.

CBR is in its infancy in PNG. The NationalHeath Plan 2001–2010 indicates that CBRis the direction to be taken in PNG with publichospitals providing relevant consultativeservices. It also states that staff will be trainedaccordingly (1). The Education Departmentpromotes ‘inclusive education’ for childrenwith disabilities so it is also supporting acommunity response to disability (6). CBRalso has a role in promoting health andpreventing disability, leading to reducedpressures on health and social services.

Currently most rehabilitation in PNGappears to be institution based. Indeed ithas recently been recommended by oneauthor that due to the lack of CBR in PNGand the difficulty following patients up asoutpatients, patients benefiting from availablerehabilitation should remain in hospital untiltheir needs have been met (7). A study fromTari that was published in 1993 concludedthat large-scale CBR projects were notrequired in the highlands (8). However, thisresearch was based on only a smallpercentage of the disabled population with a‘significant’ disability and failed toacknowledge the extended role of CBR insocial integration, disability prevention andimproved functional abilities for the disabledperson, rather than just as a means ofproviding care. In contrast a 1988 study ofthree years’ experience of the East NewBritain Project for the Disabled recorded 899persons with disability, the authorsconcluding that “functional assessment,rehabilitation and care of the disabled are notonly feasible in the community, butpreferable” (9).

So if CBR is a possible way to addressdisability in PNG how extensive is it and,more importantly, how much is it needed?Whilst it is difficult to compare prevalencerates of disability due to the many factorsknown to influence disability and the

217


problems of heterogeneity of data collection,WHO estimates that 10% of the world’spopulation is disabled (10). Two recentstudies in PNG both found a prevalence ofdisability of 3.2% in their study population.The Melanesian Institute reported a study ofthe prevalence of disability in the Woseraarea of East Sepik Province (11). The 3.2%prevalence included people with medium tosevere disability, excluding those disableddue to old age. It must be noted that theseverity of the disability was judgedsubjectively by the field workers. Disabilityprevalence ranged from 0.8% to 11.5% forthe villages studied. A study of disability inthe Middle Ramu of Madang Province alsorecorded a prevalence of 3.2% (12). In thiscase the disability was classified accordingto the International Classification ofFunctioning, Disability and Health (ICF) leveltwo classification. Disability prevalence forindividual villages ranged from 0.9% to11.1%. The study from the Tari region of theSouthern Highlands found a prevalence ofonly 0.46% (8), but as previously noted thissurvey looked only at severe disability andexcluded those with mental or intellectualdisabilities. The East New Britain studymentioned above (9) clearly showed a higherprevalence, though this was not calculated.

Little is known about the extent ofrehabilitation services in PNG. It is knownthat there are 11 Special Education ResourceCentres within PNG and that they runinformal rehabilitation programs for childrenand early intervention programs (13). TheDepartment of Health in 2001 admitted thatrehabilitation services are lacking, that thegovernment services are based in hospitalsand that the majority of consumers cannotaccess the services that do exist (1). Thereare currently only two Papua New Guineanphysiotherapists although this number shouldimprove with the recent opening of theDiploma course at Divine Word University(DWU) in Madang. Until thesephysiotherapists are qualified mostrehabilitation will continue to be carried outby overseas therapists, who tend to comefor a short time, and a dedicated group of

PNG rehabilitation workers who have hadtraining from these expatriate therapists.

The aims of this study were to establishwhat rehabilitation services are available inPNG, what areas have access torehabilitation services, what form theseservices take and which disabled personsthey cater for.

Objectives of the study were:

1 To establish the number of organizationsin PNG that are offering rehabilitation tothe disabled.

2 To collect data on the area covered, typeof disabilities addressed and mode ofdelivery of service of these organizations.

3 To establish the extent to which theserehabilitation services are knownthroughout PNG.

A secondary objective was to improvecommunication and networking betweenrehabilitation services in PNG, potentiallyleading to a more efficient and effective useof resources available for rehabilitation inPNG.

Method

A piloted self-administered close-endedquestionnaire was designed to obtain thedata required, most of which was of aquantitative nature. The questionnaire wassent with an explanatory letter to allrehabilitation services that I knew about(Table 1), all provincial health departments,the physiotherapists or Director of MedicalServices of all provincial capital hospitals,health secretaries for all the church healthservices and other organizations that mayhave been aware of rehabilitation services.Stamped addressed envelopes wereprovided for charitable organizations andsmall or rural rehabilitation projects. Thoseorganizations not responding within threemonths were sent another copy of thequestionnaire and a further letter asking for

218


TABLE 1

REHABILITATION SERVICES INITIALLY CONTACTED

Central Province and National Capital DistrictCheshire Homes, Port MoresbyFriends of the Disabled Association, Port MoresbyNational Board for Disabled Persons, Port MoresbyPNG Rehabilitation Centre, Port MoresbyPort Moresby Sheltered WorkshopSt John Association for the Blind, Port Moresby

East New Britain ProvinceCallan Services, RabaulHelen Goh, Occupational Therapist, Palmalmal, Pomio District

East Sepik ProvinceCallan Services, Wewak

Eastern Highlands ProvinceMount Sion, Goroka (Callan Services)

Enga ProvinceKarina Mills, Occupational Therapist, Kompian District

Madang ProvinceCreative Self-Help Centre, MadangLeprosy Mission, Madang

Morobe ProvinceMorobe Special Education Resource Centre, LaeNational Assembly for Disabled Persons, LaeNational Orthotic and Prosthetic Service, LaeSr B. Don, Physiotherapy Assistant, Braun Health Centre, Finschhafen

North Solomons ProvinceTaree Brearly, Occupational Therapist, Arawa Health Centre

Sandaun ProvinceFather Antonine Services, AitapeSenta Bilong Helpim, Vanimo

Simbu ProvinceSt Theresa’s Services for the Disabled, Kundiawa

Western ProvinceCallan Services, Kiunga

Western Highlands ProvinceBrothers of Charity, BanzCallan Services, Mt HagenCBR Worker, Aip Catholic Church, St Paul’s Parish, Mt HagenCBR Worker, Bukapena Health Centre, Mt HagenCBR Worker, Kenan Trangu Centre (Kuli Parish), Mt HagenCBR Worker, Ketepam Aid Post, Mt HagenCBR Worker, Kindeng Health Centre, Mt HagenCBR Worker, Kumdi Catholic Church and Mission, Mt HagenCBR Worker, St John of God Association for the Disabled, Tupa, Mt HagenCBR Worker, Togaba SDA Church, Mt Hagen

219


it to be returned. The author then collatedresults by hand. When addresses wereprovided questionnaires were sent to theadditional services that had not originallybeen contacted but which were mentionedby the initial respondents. Hard or electroniccopies of the results were sent to allparticipants.

Results

A total of 135 questionnaires were sentinitially and questionnaires weresubsequently sent to 10 of the 21 additionalservices that were mentioned by the initialrespondents. After the first mailing, 57 replieswere received. At the three-month follow-upa further 6 replies were received giving a totalof 63 questionnaires returned from whichinformation could be gained – a responserate of 47%. Of the questionnaires returned24 respondents (38%) stated that they didnot carry out any rehabilitation. These weremainly provincial and church health servicesand organizations involved with placingexpatriate rehabilitation workers. 39respondents (62%) stated that they wereinvolved in some form of rehabilitation. Thegeographical spread and names of the

rehabilitation services that replied are shownon the map (Figure 1), and the types ofrehabilitation and service provision in Table2.

Types of rehabilitation

Physical rehabilitation was carried out by38 respondents (97%). Other forms ofdisability were less well represented (Table2). 8 respondents (21%) stated other formsof rehabilitation such as income-generatingprojects, counselling and early intervention,and for remand prisoners, people withlearning disabilities, children withdevelopmental delay and older people.

Service provision

Of the 39 organizations identified as beingengaged in rehabilitation, 2 (5%) providedservices only to children. The other 37rehabilitation services provided support toboth children and adults. It was common forrespondents to offer more than one type ofservice. Other services provided includedincome-generating projects, orthotic andprosthetic fitting, eye and ear clinics, disabilityawareness, public education of disability and

2ELBAT

TYPES OF REHABILITATION AND SERVICES PROVIDED BY 93 ORGANIZATIONS IN GNP

foepyTnoitatilibaher

forebmuN)%(stnednopser

ecivreSnoisivorp

forebmuN)%(stnednopser

lacisyhP )79(83 gnittesetucA )65(22

latneM )65(22 lortaP/hcaertuO )27(82

gniraeH )65(22 stisivemoH )27(82

lausiV )26(42 desab-ertneC )45(12

hceepS )64(81 desab-loohcS )13(21

laicoS )14(61 RBC )46(52

noitacudelaicepS )14(61

rehtO )12(8

noitatilibaherdesab-ytinummoc=RBC

220


Figure 1. Rehabilitation services in PNG identified in the survey from the 39 positive responses to thequestionnaire.

1 Aip St Paul’s Parish CBR 21 Leprosy Mission2 Alotau Hospital 22 Madang Creative Self-Help Centre3 Angau Hospital 23 Mendi Hospital4 Arawa Health Centre 24 Mingende Rural Hospital5 Boram Hospital 25 Modilon General Hospital6 Braun Hospital, Finschhafen 26 Mt Hagen Hospital7 Brothers of Charity, Banz 27 Mt Sion Institute for the Blind8 Callan Services, Wewak 28 Nonga Base Hospital9 Callan Services, Kiunga 29 NOPS – Lae10 Callan Services, Mendi 30 Papul Community CBR11 Cheshire Homes 31 PNG Rehabilitation Centre12 Enga Disabled Association 32 Popondetta General Hospital13 Father Antonine Services, Aitape 33 Salvation Army14 Goroka Base Hospital 34 Senta Bilong Helpim15 Kenan CBR 35 St John Association for the Blind16 Kimbe Hospital 36 St John of God Association for the17 Kindeng Community CBR Disabled18 Kiripia CBR 37 St Theresa’s Services CBR19 Kompian District Hospital 38 St Vincent’s Special Education20 Kundiawa District Hospital Resource Centre

Note: One respondent location was unknown. CBR = community-based rehabilitation.NOPS = National Orthotic and Prosthetic Service.

training volunteers to carry out rehabilitation.Of those services based in hospitals, clinicsor centres 30 (77%) expected clients toprovide their own transport to attendrehabilitation. Only 9 (23%) of the servicesprovided transport for the clients to come torehabilitation. For those that carried outcommunity work (CBR, home visits,outreach) 9 (23%) supplied their ownpersonal transport to get to the clients. 22(56%) were able to use organization transport

to access their clients and 22 (56%) accessedtheir clients on foot. Other forms of transportused to reach community clients includedPMVs, bikes, hired car and plane.

Client referral and service information

29 organizations (74%) received referralsfrom local health institutions, 11 (28%) fromlocal schools, and 22 (56%) from otherrehabilitation workers. 31 (79%) received

221


and accepted self-referrals. The principalways in which rehabilitation servicesinformed potential clients of their services areshown in Table 3. Other forms of informingthe public of the services provided and otherways in which clients were made known torehabilitation included:

• Local doctors

• Word-of-mouth

• The courts

• Church leaders

• Community screening

• Disability awareness meetings

• Submitting reports

• Using existing health structures

• Networking meetings

• Use of the media and booklets

Awareness of other rehabilitationservices

Some 30 respondents (77%) who offered

rehabilitation services stated that they knewof other rehabilitation services in PNG. 7(18%) said they knew of no otherrehabilitation services and 2 respondents(5%) did not answer this question. Amongthe 24 respondents from groups who did notprovide rehabilitation services 13 (54%) knewof the existence of at least one namedrehabilitation service; 2 respondents (8%)failed to answer this question and 9 (38%)stated they knew of no rehabilitation services.Rehabilitation services most commonlyknown to exist by both respondents who didand those who did not carry out rehabilitationwere:

• Hospital services

• Callan services (one or all of theirbases)

• Creative Self-Help Centre

• Senta Bilong Helpim

• Brothers of Charity at Fatima

• The highlands CBR network

Less well known were:

• Leprosy Mission

3ELBAT

TYPES OF ADVERTISING USED BY RESPONDENTS TO INFORM POTENTIAL CLIENTS OF THEIRSERVICES

secivresgnisitrevdA )%(stnednopserforebmuN

htuom-fo-droW )78(43

shpargotohprosretsoP )13(21

snoitutitsninoitacudelacoL )13(21

snoitutitsniytinummoclacoL )44(71

snoitutitsnihtlaehlacoL )44(71

gnisitrevdalamroF )31(5

222


• National Board for Disabled Persons

• Father Antonine Services

• Morobe Special Education ResourceCentre

• Red Cross Special EducationResource Centre

• St John Association for the Blind

• Cheshire Homes

• PNG Rehabilitation Centre

Some other associations or organizationswere mentioned, but each by only one of therespondents.

Discussion

There are several limitations to this study.A questionnaire is not the most reliable wayof obtaining information and there is noguarantee that all rehabilitation projects inPNG have been identified. Specific problemswith the questionnaire include interpretationof such ideas as ‘social disability’. This wasmeant to identify those who were poorlyintegrated into society, but may have beeninterpreted differently. There may also havebeen overlap between the description of aservice as CBR or home visit. The study wasnot designed to evaluate the quality andeffectiveness of the projects that areoperating. Whilst it is easy to report that aprovince-wide service is provided, it is muchmore difficult to assess its standard andeffect.

The response rate of 47%, thoughdisappointing, was considered high enoughto enable some general conclusions to bedrawn. Perhaps a greater response ratewould have been achieved if stampedaddressed envelopes had been sent to allpotential respondents. It was interesting tonote that the greatest number of replies wasreceived from organizations that carried out

rehabilitation. Possibly those organizationsinvolved in rehabilitation saw responding asan opportunity to improve their own area ofinterest, whilst those who offer norehabilitation saw no gain from replying.

Location of services

Superficially it appears that rehabilitationservices are spread widely throughout PNG,but more detailed examination is required todetermine what is really happening. Therewere no reports of any rehabilitation servicesin operation in Manus, New Ireland and GulfProvinces. Oro Province and Milne BayProvince have hospital physiotherapy unitsonly, in Popondetta and Alotau. Whilst Alotauoffers outreach services people from the ruralareas of these provinces have very limitedaccess to rehabilitation services.

Madang, Morobe and Central (includingthe National Capital District [NCD]) Provincesalso have physiotherapy departments in themajor hospitals (Modilon, Angau and PortMoresby General Hospital). Modilon andAngau both offer acute services to those inthe hospital and home visits to those in urbanareas. They also try to incorporate some ruralwork when capacity and resources allow. InMadang, the Leprosy Mission offers servicesall over PNG. The work of the hospitalphysiotherapy staff is complemented by theCreative Self-Help Centre, who offer CBR,centre-based rehabilitation, home visits andspecial education to the province as a whole,although they are based in Madang town. Itwould appear that Madang Province is ableto offer some form of rehabilitation to at leastpart of its rural population. In Moroberehabilitation is offered in the Finschhafenarea, including acute, outreach and homevisits, and there is a Special EducationResource Centre (no response, but alreadyknown about) that offers special educationand carries out CBR within the Lae urbanarea. The National Orthotic and ProstheticService (NOPS), which services the wholecountry, and the National Assembly forDisabled Persons (NADP – a self-help group)are also based in Lae. The majority of rural

223


Morobe has no, or limited access torehabilitation services on a day-to-day basis.

In Central Province and NCD therehabilitation services are congregatedaround Port Moresby. As well as PortMoresby General Hospital (PMGH), there isSt John Association for the Blind, CheshireHomes, the Salvation Army and the PNGRehabilitation Centre. The National Boardfor Disabled Persons (NBPD), Friends of theDisabled Association (FODA) and PortMoresby Sheltered Workshop also exist, butthese are support services not rehabilitationservices. Of the respondents based in PortMoresby all offer centre-based rehabilitation.Two offer home visits (St John and SalvationArmy), two offer CBR (St John and PNGRehabilitation Centre) and St John also offersspecial education. It appears that CentralProvince/NCD offers a very limitedrehabilitation service to its rural population.

Western Province, North SolomonsProvince, East New Britain Province andWest New Britain Province all offer someform of CBR. I knew that Callan Servicesfor the Disabled run Special EducationResource Centres in Kimbe, Buka andRabaul, but no information was received fromthese centres. There are also physiotherapydepartments at Kimbe and Nonga BaseHospitals, both offering acute rehabilitationservices. Nonga also offers outreach andhome visits. At the time of the research therewere also two occupational therapists (OT)in these provinces, one based in Pomio inEast New Britain and one at Arawa in NorthSolomons. One did not respond to thequestionnaire; the other was offering a limitedservice to Arawa Health Centre and a fewvisits to the surrounding area. WesternProvince is offered services by CallanServices based in Kiunga. They offer hospitalservices, home visits, centre-basedrehabilitation, special education and CBR.Again, these provinces offer some form ofCBR but it does not cover the wholepopulation and those in some rural areaswould find accessing these services difficult.

Areas best served by rehabilitationservices are the two Sepik provinces and theHighlands Region. East Sepik offers acutehospital services in Wewak and a province-wide CBR service through Callan Services.Sandaun Province offers centre-basedrehabilitation in Vanimo and Aitape (SentaBilong Helpim and Father Antonine Servicesrespectively). Both these centres also offerhome visits and outreach work. FatherAntonine also offers a province-wide CBRprogram and there is physiotherapy at RaihuHospital in Aitape.

The highlands are served by acuterehabilitation services at Mt Hagen, Goroka,Mendi and Kundiawa Hospitals. Thesehospitals also offer home visits and outreachwork. These acute services are backed upby Callan Services in Mendi, Goroka (whichis also home to Mt Sion Institute for the Blind)and Mount Hagen, which offer specialeducation, centre-based and home-basedservices and CBR. This is further backed upby a network of village-based CBR projects.

Kundiawa is served by St Vincent’s SpecialEducation Resource Centre, offering home,community and centre-based rehabilitationas well as special education, and MingendeRural Hospital is served by St Theresa’sRehabilitation Services, offering acute,outreach, centre, home and CBR services.Banz is home to the Brothers of Charity, whooffer rehabilitation in the form of a centre andCBR to surrounding villages. Finally, in EngaProvince there is an OT at Kompian DistrictHospital offering acute services and trying tostart some outreach services and the EngaDisabled Association offering CBR, homeand centre-based rehabilitation province-wide.

This gives the overall impression of acountry in which the majority of the rurallybased population has very limited access torehabilitation services on a local level,although most could find some rehabilitationavailable within their province. The ruralservices that do exist seem to be

224


concentrated in the Highlands Region andSepik provinces. Although these areas aredensely populated there is an inequity ofservice provision. Chances of goodrehabilitation should not depend on thevillage location.

Disabilities provided for

What rehabilitation services that do existseemed to be biased towards physicaldisabilities. It would appear that there is lesscomprehensive rehabilitation for visual,hearing, speech and mental disabilities.Services for disabilities other than physicalseem to be fairly well (albeit thinly!) spreadthroughout the country.

Client group

Most rehabilitation services provideservices to adults and children. There is aslight bias towards children, highlighting theimportance of early intervention. This is onecomponent of CBR that gives support tochildren with a disability, through identificationand diagnosis of the problem, interventionto assist the disabled child and awareness-raising on the prevention of disabilities(14:139).

Type of service

Types of services provided were fairlyevenly spread. Whilst it was encouraging tonote that a high percentage of rehabilitationservices offered some form of home,community or outreach service theirfrequency and the area covered were notclear. CBR is the rehabilitation serviceemphasized by the WHO as well as the PNGNational Health Plan 2000-2010 (1). It shouldbe an integral part of communitydevelopment and should be led by thecommunity with the back-up of outsideagencies. The National Health Plan statesthat CBR in PNG needs to be backed byNGOs. This seems to be the case as thegovernment institutions are less involved indelivering CBR than the NGOs (although thephysiotherapists in the hospitals are often

used as resource personnel to assist theCBR workers).

Transport

Most disabled people will have difficultywith transport, but this is especially so in PNGwhere there is little consideration for disabledpersons when planning public transport. Thishas implications for the rehabilitationservices. The majority of services expectedtheir disabled clients to make their own wayto their rehabilitation. Transport difficultiesinevitably lead to many patients being unableto access services. Consideration needs tobe given to ways in which public transportmay be made more accessible for disabledpersons. For the severely disabled therehabilitation workers need to be able totravel to the patients. For those services thatdo visit their clients most are able to get totheir clients using organization transport oron foot. But a noticeable percentage oforganization personnel were using their ownpersonal transport to access patients andpresumably covering the costs themselves.This shows great commitment, but is notsustainable and issues around transport fordisabled persons and rehabilitation workersneed to be addressed.

Referrals

Self-referral appears to be the main sourceof clients for the majority of organizations.Referrals from local health institutions werealso a major source of referrals. Relativelyfew organizations reported acceptingreferrals from schools, but whether this wasbecause schools did not know how or whereto refer, because the rehabilitationorganizations refused referrals from schools(unlikely considering the high percentage thataccepted self-referral) or because disabledchildren do not make it to school in order tobe referred is unknown. This is an area thatneeds further investigation, as disabilities areoften detected at school. There need to befurther awareness campaigns so that schools

225


know about their local rehabilitation services.

Only one respondent mentionedcommunity screening as a way of identifyingclients. This would seem to be a practicalway to identify the disabled in the rural areasif enough personnel were available to do thescreening. Screening might best becombined with other projects such as childimmunizations, mother and child clinics, andschool immunizations.

Awareness of services

Rehabilitation services can only besuccessful if people know that they exist. Inorder to promote themselves and their work,rehabilitation services need to be proactivein advertising their services, especially in acountry where the benefits of rehabilitationare not generally known by the localpopulation. Some community awarenessmeetings were run by some of the NGOs aspart of their CBR program. The number oforganizations that rely on word-of-mouth forinforming clients of their existence issurprisingly high. Whilst word-of-mouth maybe an effective way of spreading informationit may not be the most reliable. Rehabilitationservices need to work together to improvethe effective dissemination of what they do.

Perhaps unsurprisingly most rehabilitationorganizations were aware of at least oneother rehabilitation program or project withinPNG. Interestingly, and encouragingly, morethan half the respondents who did not carryout rehabilitation were aware of at least onerehabilitation service in PNG. Theorganizations that were most often cited byrehabilitation workers were:

• Callan Services

• Hospital services



Among those who did not providerehabilitation the most commonly citedrehabilitation organizations were:

• Callan Services



These results are not surprising sincethese are the best resourced organizationsin the country, running large CBR programsand being highly involved in the needs of thedisabled in Papua New Guinea.

Conclusions

Strengths

• Rehabilitation services exist in PNG.

• There are efforts being madeespecially in the Sepik provinces andthe Highlands Region to takerehabilitation to the communitiesthrough CBR programs. Theseprograms are expanding and reachingother provinces also.

• Both children and adults seem to haveaccess to services.

• Even those services based in theacute setting are trying to get out intothe community to some extent.

Weaknesses

• The majority of services are based inurban areas so are biased towards thispopulation, whereas most of thepopulation of PNG live in rural areas.

• There is an inequity of services acrossthe country.

• Services seem to focus more onphysical disability than other types ofdisability.

226


• Transport is a major problem both inaccessing clients and clientsaccessing services.

• It appears that awareness ofrehabilitation services is limited.

• Resources for most of the smallerCBR projects are a problem.Questionnaires were often returnedwith requests for financial assistanceor equipment.

• Almost all CBR is being carried out byNGOs. Among these, Callan Servicesfor Persons with a Disability has thebiggest share.

• Government-run rehabilitation projectsare based mainly in acute hospitalswhich lack the resources required tocarry out an effective outreachprogram.

• There is a lack of funding and suitablyqualified personnel to fulfill therehabilitation needs of PNG.

Recommendations

The organizations involved in rehabilitationneed to work harder to inform otherprofessionals as well as the community atlarge about what they do, both in terms ofrehabilitation and disability prevention. Thiscould be facilitated through communityawareness work and community screeningto identify disabled persons and wherepossible promote health and preventdisability.

Alternative sources of funding are requiredto ensure that rehabilitation reaches the ruralpopulation, and funding needs to be backedup by appropriate personnel and training.This will hopefully be partially addressed bythe opening of the Diploma in Physiotherapycourse at Divine Word University in Madang.

There needs to be a proactive movement

of rehabilitation into the community to meetthe demands and needs of the community.CBR is ideally suited to this situation and ispivotal to responding to the needs of disabledPapua New Guineans in a culturallyappropriate, sustainable and empoweringway (12).

Greater collaboration between existingrehabilitation services (both NGO andgovernmental) and increased resources arerequired. There is also a need for greatercommunication between the SpecialEducation Resource Centres, teachertraining colleges and rehabilitation providersaround PNG to support the integration ofdisabled children into mainstream schoolwherever possible. This is being addressedby the Department of Education, as theSpecial Education Resource Centres havebeen actively involved in conductinginservice training for primary and elementaryteachers (14). Plans to expand this trainingto secondary school teachers also exist.

If CBR is to continue to grow in the presenteconomic climate the reliance on NGOs issensible, as they may be able to access morefunding than the government is able to spendon rehabilitation. Their work needs to besupported and monitored by the governmentto ensure that it is appropriate andsustainable and that it links into othergovernment structures where appropriate, forinstance education and vocational training.

Areas for further research

Further research needs to be conductedto establish the quality and capacity of therehabilitation services that exist at present,to address transport needs of the disabledand rehabilitation workers, and to investigatethe role of schools in identifying disabledchildren.

REFERENCES

1 Papua New Guinea Department of Health. PapuaNew Guinea National Health Plan 2001-2010.Volume 2. Port Moresby: Department of Health,2000:84.

2 World Health Organization. International

227


Classification of Functioning, Disability and Health(ICF). Geneva: World Health Organization, 2001:3-4.

3 International Labour Organization, UnitedNations Educational, Scientific and CulturalOrganization, World Health Organization.Community-Based Rehabilitation for and with Peoplewith Disabilities. Joint Position Paper. Geneva:World Health Organization, 1994.

4 Mitchell R. The research base of community-basedrehabilitation. Disabil Rehabil 1999;21:459-468.

5 United Nations. The Standard Rules on theEqualization of Opportunities for Persons with aDisability. Adopted by the United Nations GeneralAssembly at the Forty-eighth Session, 20 Dec 1993(resolution 48/96). New York: United NationsDepartment of Public Information, 1994.

6 Papua New Guinea Department of Education.Papua New Guinea National Education Plan 1995-2004. Volume B. Port Moresby: Department ofEducation, 1996:76.

7 Powell N. Physiotherapy in Mount Hagen GeneralHospital: an audit of activity over a six-month period.PNG Med J 2001;44:24-35.

8 van Amstel H, Dyke T, Crocker J. Would peoplewith a disability in the highlands benefit from acommunity-based rehabilitation program? PNG Med

J 1993;36:316-319.9 Hamilton J, van Zwanenberg TD. Three years’

experience of the East New Britain project for thedisabled. PNG Med J 1988; 31:163-167.

10 Barbotte E, Guillemin F, Chau N, LorhandicapGroup. Prevalence of impairments, disabilities,handicaps and quality of life in the generalpopulation: a review of recent literature. Bull WorldHealth Organ 2001;79:1047-1055.

11 Melanesian Institute. Attitudes and practicestowards people with a disability. A survey in theWosera area of the East Sepik Province, Papua NewGuinea. Unpublished report. Melanesian Institute,Goroka, 2003.

12 Byford J, Veenstra N, Gi S. Towards a method forinforming the planning of community-basedrehabilitation in Papua New Guinea. PNG Med J2003;46:63-80.

13 Papua New Guinea Department of Education.The state of education in Papua New Guinea, March2002. In: Education Reform Facilitating andMonitoring. Port Moresby: Papua New GuineaDepartment of Education, 2002:105-106.

14 Rutten G. Study guide – associate course in physicalrehabilitation, Part 4 of 4. Unpublished. FatherAntonine Services, Aitape, 2002:139.

228


List of Medical Research Projects in Papua New Guinea

Approved or Noted

By the Medical Research Advisory Committee in 2003

Sanguma (Witchcraft) – Its medical, socio-legal implications

Dr Thomas Kubu and Dr Brother Andrewssf (Accidents and Emergencies Unit,Goroka Base Hospital, PO Box 392, Goroka,Eastern Highlands Province 441, Papua NewGuinea)

Merging bacterial diseases and lymphaticfilariasis mediated immunosuppression inPapua New Guinea

Prof. John Reeder, Prof. David Durrheim,Dr N Ketheesan, Dr Richard Speare and MrJeffrey Warner (Papua New Guinea Instituteof Medical Research, PO Box 60, Goroka,Eastern Highlands Province 441, Papua NewGuinea)

Genotyping Trichomonas vaginalisinfections and their association with HIV inPapua New Guinea

Dr Jacqueline Upcroft, Dr Peter Upcroft,Prof. John Reeder and Dr Peter Siba(Queensland Institute of Medical Research,300 Herston Rd, Herston, Queensland 4006,Australia)

Duffy antigen expression in SEAovalocytosis

Dr Alfred Cortés, Dr Stephen Opat, DrErica Wood and Dr Rosemary Sparrow(Papua New Guinea Institute of MedicalResearch Madang, PO Box 378, Madang,Madang Province 511, Papua New Guinea)

Determination of receptors forPlasmodium falciparum rosetting in PapuaNew Guinea

Dr Alexander Rowe, Mr Ian Cockburn, DrAlfred Cortés and Prof. John Reeder(Institute of Cell, Animal and PopulationBiology, University of Edinburgh, West MainsRd, Edinburgh EH9 3JT, United Kingdom)

Factors influencing attendance atantenatal clinics and treatment-seekingbehaviour of pregnant women for malariaand anemia – does non-attendance impacton low infant birthweight?

Ms Rachael Hinton, Prof. John Reeder andMs Olive Oa (Papua New Guinea Instituteof Medical Research Maprik, PO Box 400,Wewak, East Sepik Province 531, PapuaNew Guinea)

Phytochemical diversity and humanhealth: dietary change and type 2 diabetesmellitus in transitional communities of PapuaNew Guinea

Mr Patrick Owen, Dr Lohi Matainaho andDr Timothy Johns (School of Dietetics andHuman Nutrition, Macdonald Campus ofMcGill University, 21,111 Lakeshore Road,Saint-Anne-de-Bellevue, Montreal, QuebecH9X 3V9, Canada)

Sexual and reproductive agency in theTrobriand Islands: mediating HIV/AIDSmessages

Ms Katherine Lepani (Gender RelationsCentre, Research School of Pacific and AsianStudies, The Australian National University,Canberra, ACT 0200, Australia)

CD8-T-cell memory to liver stagePlasmodium falciparum in humans

Dr James Kazura, Prof. John Reeder andDr Moses Bockarie (Division of GeographicMedicine and International Health, CaseWestern Reserve University School ofMedicine, 2109 Adelbert Road, Cleveland,Ohio 44106, USA)

Love, marriage and HIV: a multi-site studyof gender and HIV risk

Dr Holly Wardlow (University of Toronto,Anthropology Department, 100 St George

229


Street, Toronto, Ontario, Canada)

Arbovirus surveillance in clinicalpresentations of meningoencephalitis inMadang

Dr Harin Karunajeewa, Dr Peter Siba, DrCecilia Pakula, Dr Cheryl Johansen, Prof.Tim Davis, Dr Moses Bockarie, Dr DavidSmith, Dr Annette Broom and Dr JosephAmban (University Department of Medicine,Fremantle Hospital, PO Box 480, Fremantle,WA 6160, Australia)

Project to investigate the contribution ofparvovirus B19 to episodes of severeanaemia among children 6 months to 5 yearswith malaria in Papua New Guinea

Prof. Yvonne Cossart, Dr James Wildig,Dr Peter Siba and Dr Ivo Mueller (Room 604,Blackburn Building D06, Department ofInfectious Diseases, University of Sydney,NSW 2006, Australia)

Case-control study on severe malarialanaemia in the Wosera, East Sepik Province

Dr Ivo Mueller, Prof. John Reeder, DrSimon Foote and Dr Louis Schofield (PapuaNew Guinea Institute of Medical Research,PO Box 60, Goroka, EHP 441, Papua NewGuinea)

Social effects of new medical technologiesin Papua New Guinea

Alice Street (Department of SocialAnthropology, Free School Lane, CambridgeCB2 3RF, United Kingdom)

What are the causes of unexplained feverin children presenting to the children’soutpatient department at PMGH?

Mr Clement Morris (School of Medicineand Health Sciences, University of PapuaNew Guinea, PO Box 5623, Boroko, NCD111, Papua New Guinea)

Assessment of HIV/AIDS knowledge,attitudes and behaviour of high schoolstudents in Papua New Guinea

Dr M Kiromat, Dr N Pomat, Dr G Oswyn,Dr M Tovilu, Dr I Hwaihwanje, Dr C Pakule,Dr G Kiage, Dr J Amini, Dr L Bussim, Dr N

Tefuarani, Dr D Mokela, Dr M Baki, Dr VVetuna, Dr P Ripa and Prof. John Vince(Division of Paediatrics, School of Medicineand Health Sciences, University of PapuaNew Guinea, PO Box 5623, Boroko, NCD111, Papua New Guinea)

Investigation of alleged disease outbreakin the Fane and Mekeo areas of CentralProvince, Papua New Guinea

Dr Tukutau Taufa (School of Public Healthand Tropical Medicine, Anton Breinl Centre,James Cook University, Townsville,Queensland 4811, Australia)

A comparative investigation of traditionalhealing practices on indigenous men’s health

Mr Adrian Miller (School of Public Healthand Tropical Medicine, Anton Breinl Centre,James Cook University, Townsville,Queensland 4811, Australia)

Definition of the human response toPlasmodium falciparum HGXPRT

Prof. Michael Good, Prof. John Reederand Dr Danielle Stanisic (QueenslandInstitute of Medical Research, The BancroftCentre, 300 Herston Road, Herston,Queensland 4006, Australia)

Neonatal immunization withpneumococcal conjugate vaccine in PapuaNew Guinea

Prof. John Reeder, Assoc. Prof. DeborahLehmann, Prof. Patrick Holt, Dr PeterRichmond and Mr William Pomat (PapuaNew Guinea Institute of Medical Research,PO Box 60, Goroka, Eastern HighlandsProvince 441, Papua New Guinea)

Baseline study on knowledge, attitude andpractice of children on tobacco use

Dr Timothy Pyakalyia (Department ofHealth, Division of Technical Health Services,Disease Control Branch, PO Box 807,Waigani, NCD 131, Papua New Guinea)

Understanding the pathophysiology andimmunology of acute lung injury in malaria

Prof. Isi Kevau, Dr Jerry Minei, Assoc. Prof.Nick Anstey, Prof. Peter Sly, Prof. Michael

230


Pain and Dr Graeme Maguire (School ofMedicine and Health Sciences, University ofPapua New Guinea, PO Box 5623, Boroko,NCD 111, Papua New Guinea)

A comparative analysis: the cause/s of lowlevel of supervised delivery rates in Nuku andTelefomin Districts, Sandaun Province

Mr Desak Drorit and Mr Dawa Wera(Coordinator – Sandaun ProvincialAdministration, Technical Health Services,PO Box 84, Vanimo, Sandaun Province 551,Papua New Guinea)

Nursing education as capacity building inPNG

Ms Nayer Kaviani (PO Box 865, Waigani,NCD 131, Papua New Guinea)

Qualitative study to evaluate the HealthPromotion Strategy of HPB, focusing onmalaria, TB and EPI

Dr Alison Heywood and Mr Lindsay Piliwas(Formulation Research/Evaluation Advisor,HSSP, Department of Health, PO Box 807,Waigani, NCD 131, Papua New Guinea)

Molecular epidemiology of hepatitis B virusinfection of the South Pacific

Dr Andrew Masta, Dr Oji Malani, Prof.David Penny, Prof. Mathias Sapuri, Dr PaulMark, Dr Jan Pryor and Ms Abby Harrison(School of Medicine and Health Sciences,University of Papua New Guinea, PO Box5623, Boroko, NCD 111, Papua New Guinea)

Seroepidemiology of dengue in PNGDr John Irima and Dr Samir Dutta

(Microbiology Unit, Port Moresby GeneralHospital, Free Mail Bag, Boroko, NCD 111,Papua New Guinea)

Note:

These projects have been examined andcleared by the MRAC but they have not allstarted, nor is there any guarantee that theyall will, since in many cases this still dependson funding. It should be noted that the projectfunds for the MRAC were deleted from theHealth Budget from 1997 to 2003.

Information about these projects may beobtained from the investigators or from theChairperson of the Medical ResearchAdvisory Committee (Director of Researchand Monitoring, Department of Health, POBox 807, Waigani, NCD 131)

231



PUBLICATIONS OF RELEVANCE TO PAPUA NEW GUINEA AND MELANESIA

Bibliographic Citation List generated from MEDLARS

1 Barnes JL, Warner J, Melrose W, Durrheim D,Speare R, Reeder JC, Ketheesan N.Adaptive immunity in melioidosis: a possible role forT cells in determining outcome of infection withBurkholderia pseudomallei.Clin Immunol 2004 Oct;113(1):22-28.

Melioidosis is a potentially fatal disease causedby the bacterium Burkholderia pseudomallei.Individuals with subclinical melioidosis have noapparent clinical signs or symptoms, and areidentified only by positive serology. The present studyis the first to investigate cell-mediated immune (CMI)responses following in vitro stimulation with B.pseudomallei antigens in peripheral bloodmononuclear cells (PBMC), collected under fieldconditions in Papua New Guinea (PNG) fromindividuals with exposure to B. pseudomallei (n =13). While five had a clinical history of melioidosis(C(+)), the remaining individuals (n = 8) wereseropositive, yet healthy with no clinical history ofmelioidosis (S(+)/C(-)). Proliferation and IFN-gammaproduction were significantly greater in lymphocytecultures from S(+)/C(-) individuals compared to C(+)individuals (p <0.001 and p <0.05, respectively).These findings demonstrate that compared to C(+)patients, individuals with subclinical melioidosis havea stronger CMI response to B. pseudomallei antigensin vitro. Such a response may be essential forprotection against disease progression.

2 Basuni AA, Butterworth L, Cooksley G, LocarniniS, Carman WF.Prevalence of HBsAg mutants and impact ofhepatitis B infant immunisation in four Pacific Islandcountries.Vaccine 2004 Jul 29;22(21-22):2791-2799.

The prevalence rate of hepatitis B virus (HBV)infection in Pacific Island countries is amongst thehighest in the world. Hepatitis B immunisation hasbeen incorporated into national programmes atvarious times, often with erratic supply and coverage,until a regionally co-ordinated programme, whichcommenced in 1995, ensured adequate supply. Theeffectiveness of these programmes was recentlyevaluated in four countries, Vanuatu and Fiji inMelanesia, Tonga in Polynesia and Kiribati inMicronesia. That evaluation established that theprogrammes had a substantial beneficial impact inpreventing chronic hepatitis B infection [Vaccine 18(2000) 3059]. Several studies of hepatitis Bvaccination programmes in endemic countries haveidentified the potential significance of surface genemutants as a cause for failure of immunisation. Inthe study outlined in this paper, we screened infected

children and their mothers for the emergence andprevalence of these variants in specimens collectedfrom the four-country evaluation. Although theopportunity for the emergence of HBV vaccineescape mutants in these populations was high dueto the presence of a considerable amount of the virusin the population and the selection pressure fromvaccine use, there were no “a” determinant vaccineescape mutants found. This suggests that vaccineescape variants are not an important cause for failureto prevent HBV transmission in this setting. OtherHBsAg variants were detected, but their functionalsignificance remains to be determined. The failureto provide satisfactory protection during suchimmunisation programmes reflects the need forachieving and sustaining high vaccine coverage,improving the timeliness of doses as well asimproving ‘cold-chain’ support, rather than theselection of vaccine-escape mutants of HBV.

3 Benet A, Mai A, Bockarie F, Lagog M, ZimmermanP, Alpers MP, Reeder JC, Bockarie MJ.Polymerase chain reaction diagnosis and thechanging pattern of vector ecology and malariatransmission dynamics in Papua New Guinea.Am J Trop Med Hyg 2004 Sep;71(3):277-284.

The ecology and behavior of most of the 11known members of the Anopheles punctulatus groupremain unresolved and only the morphologic speciesAn. farauti, An. koliensis, and An. punctulatus areknown as vectors of malaria in Papua New Guinea.Of 1,582 mosquitoes examined morphologically, 737were identified as An. farauti s.l., 719 as An. koliensis,and 126 as An. punctulatus. All specimens identifiedmorphologically as An. punctulatus were shown tobe An. punctulatus by polymerase chain reaction-restriction fragment length polymorphism analysis,but the An. farauti and An. koliensis morphotypesconsisted of three or more species including An.farauti s.s., An. farauti No. 2, and An. farauti No. 4.The biting cycles and role in malaria transmissionof some of these species are described here for thefirst time. We also show evidence that An. koliensiscould be a sub-complex of two or more species. Theepidemiologic implications of our findings arediscussed.

4 Bettiol L, Griffin E, Hogan C, Heard S.Village birth attendants in Papua New Guinea.Aust Fam Physician 2004 Sep;33(9):764-765.

AIM: To describe the care delivered by villagebirth attendants in a remote area of Papua NewGuinea. METHOD: A qualitative study usingsemistructured interviews with 56 voluntary village

232


birth attendants. RESULTS: Village birth attendantssupervise and assist women with labour and delivery.Many have additional roles in their community. Mostare motivated by a desire to help the women in theircommunity. Common difficulties encountered includeobstetric and logistical problems. The most commonreason for ceasing work is family pressure or lack ofsupport. DISCUSSION: Village birth attendants maycontribute toward an improvement in maternalmorbidity and mortality in remote areas of PapuaNew Guinea.

5 Breman JG, Alilio MS, Mills A.Conquering the intolerable burden of malaria: what’snew, what’s needed: a summary.Am J Trop Med Hyg 2004 Aug;71(2 Suppl):1-15.

Each year, up to three million deaths due tomalaria and close to five billion episodes of clinicalillness possibly meriting antimalarial therapy occurthroughout the world, with Africa having more than90% of this burden. Almost 3% of disability adjustedlife years are due to malaria mortality globally, 10%in Africa. New information is presented in thissupplement on malaria-related perinatal mortality,occurrence of human immunodeficiency virus inpregnancy, undernutrition, and neurologic, cognitive,and developmental sequelae. The entomologicdeterminants of transmission and uses of modelingfor program planning and disease prediction andprevention are discussed. New data are presentedfrom the Democratic Republic of the Congo,Tanzania, Ethiopia, and Zimbabwe on the increasingurban malaria problem and on epidemic malaria.Between 6% and 28% of the malaria burden mayoccur in cities, which comprise less than 2% of theAfrican surface. Macroeconomic projections showthat the costs are far greater than the costs ofindividual cases, with a substantial deleteriousimpact of malaria on schooling of patients, externalinvestments into endemic countries, and tourism.Poor populations are at greatest risk; 58% of thecases occur in the poorest 20% of the world’spopulation and these patients receive the worst careand have catastrophic economic consequences fromtheir illness. This social vulnerability requires betterunderstanding for improving deployment, access,quality, and use of effective interventions. Studiesfrom Ghana and elsewhere indicate that for everypatient with febrile illness assumed to be malariaseen in health facilities, 4-5 episodes occur in thecommunity. Effective actions for malaria controlmandate rational public policies; market forces,which often drive sales and use of drugs and otherinterventions, are unlikely to guarantee their use.Artemisinin-based combination therapy (ACT) formalaria is rapidly gaining acceptance as an effectiveapproach for countering the spread and intensity ofPlasmodium falciparum resistance to chloroquine,sulfadoxine/pyrimethamine, and other antimalarialdrugs. Although costly, ACT ($1.20-2.50 per adulttreatment) becomes more cost-effective asresistance to alternative drugs increases; early useof ACT may delay development of resistance to thesedrugs and prevent the medical toll associated with

use of ineffective drugs. The burden of malaria inone district in Tanzania has not decreased since theprimary health care approach replaced the verticalmalaria control efforts of the 1960s. Despitedecentralization, this situation resulted, in part, fromweak district management capacity, poorcoordination, inadequate monitoring, and lack oftraining of key staff. Experience in the SolomonIslands showed that spraying with DDT, use ofinsecticide-treated bed nets (ITNs), and healtheducation were all associated with disease reduction.The use of nets permitted a reduction in DDTspraying, but could not replace it without anincreased malaria incidence. Baseline data andreliable monitoring of key outcome indicators areneeded to measure whether the ambitious goals forthe control of malaria and other diseases hasoccurred. Such systems are being used forevidence-based decision making in Tanzania andseveral other countries. Baseline cluster samplingsurveys in several countries across Africa indicatethat only 53% of the children with febrile illness inmalarious areas are being treated; chloroquine (CQ)is used 84% of the time, even where the drug maybe ineffective. Insecticide-treated bed nets were usedonly 2% of the time by children less than five yearsof age. Progress in malaria vaccine research hasbeen substantial over the past five years; 35candidate malaria vaccines are in development,many of which are in clinical trials. Development ofnew vaccines and drugs has been the result ofincreased investments and formation of public-private partnerships. Before malaria vaccinebecomes deployed, consideration must be given todisease burden, cost-effectiveness, financing,delivery systems, and approval by regulatoryagencies. Key to evaluation of vaccine effectivenesswill be collection and prompt analysis ofepidemiologic information. Training of persons inevery aspect of malaria research and control isessential for programs to succeed. The MultilateralInitiative on Malaria (MIM) is actively promotingresearch capacity strengthening and has establishednetworks of institutions and scientists throughout theAfrican continent, most of whom are now linked bymodern information-sharing networks. Evidence overthe past century is that successful control malariaprograms have been linked to strong researchactivities. To ensure effective coordination andcooperation between the growing number ofresearch and control coalitions forming in supportof malaria activities, an umbrella group is needed.With continued support for scientists and controlworkers globally, particularly in low-income malariouscountries, the long-deferred dream of malariaelimination can become a reality.

6 Burrows JM, Bromham L, Woolfit M, PiganeauG, Tellam J, Connolly G, Webb N, Poulsen L,Cooper L, Burrows SR, Moss DJ, Haryana SM,Ng M, Nicholls JM, Khanna R.Selection pressure-driven evolution of the Epstein-Barr virus-encoded oncogene LMP1 in virus isolatesfrom Southeast Asia.

233


J Virol 2004 Jul;78(13):7131-7137.The geographically constrained distribution of

Epstein-Barr virus (EBV)-associatednasopharyngeal carcinoma (NPC) in SoutheastAsian populations suggests that both viral and hostgenetics may influence disease risk. Althoughsusceptibility loci have been mapped within thehuman genome, the role of viral genetics in the focaldistribution of NPC remains an enigma. Here wereport a molecular phylogenetic analysis of an NPC-associated viral oncogene, LMP1, in a large panelof EBV isolates from southeast Asia and from PapuaNew Guinea, Africa, and Australia, regions of theworld where NPC is and is not endemic, respectively.This analysis revealed that LMP1 sequences showa distinct geographic structure, indicating that theSoutheast Asian isolates have evolved as a lineagedistinct from those of Papua New Guinea, African,and Australian isolates. Furthermore, a likelihoodratio test revealed that the C termini of the LMP1sequences of the Southeast Asian lineage are undersignificant positive selection pressure, particularlyat some sites within the C-terminal activator regions.We also present evidence that although the Nterminus and transmembrane region of LMP1 haveundergone recombination, the C-terminal region ofthe gene has evolved without any history ofrecombination. Based on these observations, wespeculate that selection pressure may be driving theLMP1 sequences in virus isolates from southeastAsia towards a more malignant phenotype, therebyinfluencing the endemic distribution of NPC in thisregion.

7 Curry C, Annerud C, Jensen S, Symmons D, LeeM, Sapuri M.The first year of a formal emergency medicinetraining programme in Papua New Guinea.Emerg Med Australas 2004 Aug;16(4):343-347.

OBJECTIVE: To describe a programmecatalyzing the development of emergency medicinein Papua New Guinea (PNG). METHODS: Fiveemergency physicians rotated through a new positionof Senior Lecturer in Emergency Medicine in theUniversity of PNG during 2003. The position wasestablished as a consequence of emergencyphysician input supported by AusAID in 2002.RESULTS: Fifth (final)-year medical students andmedical officers in the Emergency Department atPort Moresby General Hospital undertook formal andbedside problem based learning. The first traineesfor a Master of Medicine in Emergency Medicineprogramme were inducted and supported.Emergency Department management was providedwith specialist input. Research projects wereinitiated, dealing with snakebite, chloroquine toxicityand HIV/AIDS. The first year of an emergencynursing curriculum was supported. CONCLUSIONS:There is now considerable enthusiasm for thedevelopment of emergency medicine as the hospitalgeneralists’ specialty. Emergency nursing traininghas also made a start. Limitations on resources willrequire flexibility to sustain the project. Furthersupport by emergency physicians will be needed.

8 de Boer MA, Peters LA, Aziz MF, Siregar B,Cornain S, Vrede MA, Jordanova ES, Kolkman-Uljee S, Fleuren GJ.Human papillomavirus type 16 E6, E7, and L1variants in cervical cancer in Indonesia, Suriname,and The Netherlands.Gynecol Oncol 2004 Aug;94(2):488-494.

OBJECTIVE: Human papillomavirus type 16(HPV 16) has several intratypic variants, and someare associated with enhanced oncogenic potential.For risk determination as well as for future vaccinedevelopment, knowledge about variants is important.Regarding the geographical distribution of HPVvariants and the lack of data from Indonesia andSuriname, we studied the prevalence of HPV 16variants in cervical cancer in these high incidencecountries. Data were compared with TheNetherlands, a low-risk country. METHODS: DNAsamples from 74 formalin-fixed paraffin-embeddedHPV 16-positive cervical carcinomas from Indonesia(Java, N = 22), Suriname (N = 25), and TheNetherlands (N = 27) were amplified using primersspecific for the E6, E7, and part of the L1 regions.Products were sequenced and analyzed. RESULTS:A specific Javanese variant, with mutations 666A inE7 and 6826T in L1, was found in 73% of theIndonesian samples, 56% having an additionalmutation in the E6 open reading frame (ORF; 276G),giving the predicted amino acid change N58S. ThisJavanese variant was also found in threeSurinamese samples, which reflects what could beexpected from migration of Javanese people toSurinam. Other non-European variants wereidentified in Indonesian, Surinamese, and Dutchsamples in 14%, 28%, and 19%, respectively.CONCLUSION: The majority of the HPV 16-positivecervical cancers in Indonesia are caused by aspecific intratypic variant that was rarely found beforein other countries.

9 Donikian JC, Miralles C, Le Pommelet C.[Twenty-two cases of hepatic abscesses in childrenin New-Caledonia.] [Fr]Arch Pediatr 2004 Jul;11(7):867.

10 Fachiroh J, Schouten T, Hariwiyanto B, ParamitaDK, Harijadi A, Haryana SM, Ng MH, MiddeldorpJM.Molecular diversity of Epstein-Barr virus IgG and IgAantibody responses in nasopharyngeal carcinoma:a comparison of Indonesian, Chinese, and Europeansubjects.J Infect Dis 2004 Jul 1;190(1):53-62.

Epstein-Barr virus (EBV)-specific immunoblotanalysis was used to reveal the molecular diversityof immunoglobulin (Ig) G and IgA antibody responsesagainst Epstein-Barr nuclear antigen (EBNA), earlyantigen (EA), and viral capsid antigen (VCA) inserum samples from patients with nasopharyngealcarcinoma (NPC) and control subjects, by use ofimmunofluorescence assay (IFA). Control donors(n=150) showed IgG responses to few EBV proteins– VCA-p18, VCA-p40, EBNA1, and Zebra – andsporadically weak IgA reactivity to EBNA1 and VCA-

234


p18. Patients with NPC stage 1 (n=6) had similarresponse patterns. Patients with NPC stage 2-4(n=132) showed significantly more diverse IgG andIgA responses to EA and VCA proteins – VCA-p18/-p40, EBNA1, Z-encoded broadly reactive activator,and EAd-p47/54, -DNAse, -thymidine kinase, and -p138. No correlation was found between IFA titersand the number of EBV proteins recognized by IgGor IgA. Our results reveal dissimilarity between EBVpolypeptides recognized by IgG and IgA antibodies,which suggests independent B cell triggering events.

11 Hanson JP, Taylor CT, Richards AR, Smith IL,Boutlis CS.Japanese encephalitis acquired near Port Moresby:implications for residents and travellers to PapuaNew Guinea.Med J Aust 2004 Sep 6;181(5):282-283.

12 Huaman MC, Yoshinaga K, Suryanatha A,Suarsana N, Kanbara H.Short report: Polymorphisms in the chloroquineresistance transporter gene in Plasmodiumfalciparum isolates from Lombok, Indonesia.Am J Trop Med Hyg 2004 Jul;71(1):40-42.

The polymorphisms in the Plasmodiumfalciparum multidrug resistance 1 (pfmdr1) and P.falciparum chloroquine resistance transporter (pfcrt)genes, which are associated with chloroquineresistance, were examined in 48 P. falciparumisolates from uncomplicated malaria patients fromthe West Lombok District in Indonesia. The pointmutation N86Y in pfmdr1 was present in 35.4% ofthe isolates and mutation K76T in pfcrt was found inall but one of the samples studied. Identified pfcrthaplotypes were mainly identical to the Papua NewGuinea type S(agt)VMNT (42 of 48, 87.5%), and afew isolates had the Southeast Asia type CVIET (5of 48, 10.4%). Moreover, one P. falciparum isolateharbored the K76N mutation, giving rise to thehaplotype CVMNN, which was not previouslyreported in field isolates. Our findings suggest thatchloroquine resistance in this area might have thesame origin as in Papua New Guinea.

13 Ito A, Wandra T, Yamasaki H, Nakao M, Sako Y,Nakaya K, Margono SS, Suroso T, Gauci C,Lightowlers MW.Cysticercosis/taeniasis in Asia and the Pacific.Vector Borne Zoonotic Dis 2004 Summer;4(2):95-107.

Three taeniid tapeworms infect humans in Asiaand the Pacific: Taenia solium, Taenia saginata, andTaenia asiatica. Although there is continuing debateabout the definition of a new species, phylogeneticanalyses of these parasites have provided multiplelines of evidence that T. asiatica is an independentspecies and the sister species of T. saginata. Herewe review briefly the morphology, pathology,molecular biology, distribution and control optionsof taeniasis/cysticercosis in Asia and the Pacific andcomment on the potential role which dogs may playin the transmission of T. solium. Special attention isfocused on Indonesia: taeniasis caused by T. asiatica

in North Sumatra, taeniasis/cysticercosis of T. soliumand taeniasis of T. saginata in Bali, and taeniasis/cysticercosis of T. solium in Papua (formerly IrianJaya). Issues relating to the spread of taeniasis/cysticercosis caused by T. solium in Papua NewGuinea are highlighted, since serological evidencesuggests that cysticercosis occurs among the localresidents. The use of modern techniques fordetection of taeniasis in humans and cysticercosisin humans, pigs and dogs, with the possible adoptionof new control measures will provide a betterunderstanding of the epidemiology of taeniasis/cysticercosis in Asia and the Pacific and lead toimproved control of zoonotic and simultaneouslymeat-borne disease transmission.

14 Johansen CA, Nisbet DJ, Foley PN, Van Den HurkAF, Hall RA, Mackenzie JS, Ritchie SA.Flavivirus isolations from mosquitoes collected fromSaibai Island in the Torres Strait, Australia, duringan incursion of Japanese encephalitis virus.Med Vet Entomol 2004 Sep;18(3):281-287.

Adult mosquitoes (Diptera: Culicidae) werecollected in January and February 2000 from SaibaiIsland in the Torres Strait of northern Australia, andprocessed for arbovirus isolation during a period ofJapanese encephalitis (JE) virus activity on nearbyBadu Island. A total of 84 210 mosquitoes wereprocessed for virus isolation, yielding six flavivirusisolates. Viruses obtained were single isolates of JEand Kokobera (KOK) and four of Kunjin (KUN). Allvirus isolates were from members of the Culexsitiens Weidemann subgroup, which comprised53.1% of mosquitoes processed. Nucleotidesequencing and phylogenetic analysis of the pre-membrane region of the genome of JE isolateTS5313 indicated that it was closely related to otherisolates from a sentinel pig and a pool of Cx. gelidusTheobald from Badu Island during the same period.Also molecular analyses of part of the envelope geneof KUN virus isolates showed that they were closelyrelated to other KUN virus strains from Cape YorkPeninsula. The results indicate that flaviviruses aredynamic in the area, and suggest patterns ofmovement south from New Guinea and north fromthe Australian mainland.

15 Karunajeewa HA, Ilett KF, Dufall K, Kemiki A,Bockarie M, Alpers MP, Barrett PH, Vicini P, DavisTM.Disposition of artesunate and dihydroartemisininafter administration of artesunate suppositories inchildren from Papua New Guinea with uncomplicatedmalaria.Antimicrob Agents Chemother 2004 Aug;48(8):2966-2972.

A detailed pharmacokinetic analysis wasperformed with 47 children from Papua New Guineawith uncomplicated falciparum or vivax malariatreated with artesunate (ARTS) suppositories(Rectocaps) given in two doses of approximately 13mg/kg of body weight 12 hours apart. Following anintensive sampling protocol, samples were assayedfor ARTS and its primary active metabolite,

235


dihydroartemisinin (DHA), by liquid chromatography-mass spectrometry. A population pharmacokineticmodel was developed to describe the data. Followingadministration of the first dose, the mean maximalconcentrations of ARTS and DHA were 1,085 nmol/liter at 0.9 hours and 2,525 nmol/liter at 2.3 hours,respectively. The absorption half-life for ARTS was2.3 hours, and the conversion half-life (ARTS toDHA) was 0.27 hours, while the elimination half-lifeof DHA was 0.71 hours. The mean common volumesof distribution for ARTS and DHA relative tobioavailability were 42.8 and 2.04 liters/kg,respectively, and the mean clearance values relativeto bioavailability were 6 and 2.2 liters/h/kg for ARTSand DHA, respectively. Substantial interpatientvariability was observed, and the bioavailability ofthe second dose relative to that of the first wasestimated to be 0.72. The covariates age, sex, andalpha-thalassemia genotype were not influential inthe pharmacokinetic model development; but theinclusion of weight as a covariate significantlyimproved the performance of the model. An ARTSsuppositories dose of 10 to 20 mg/kg is appropriatefor use in children with uncomplicated malaria.

16 Leclerc MC, Durand P, Gauthier C, Patot S,Billotte N, Menegon M, Severini C, Ayala FJ,Renaud F.Meager genetic variability of the human malariaagent Plasmodium vivax.Proc Natl Acad Sci USA 2004 Oct 5;101(40):14455-14460.

Malaria is a major human parasitic diseasecaused by four species of Plasmodium protozoa.Plasmodium vivax, the most widespread, affectsmillions of people across Africa, Asia, the MiddleEast, and Central and South America. We havestudied the genetic variability of 13 microsatellite lociin 108 samples from 8 localities in Asia, Africa, SouthAmerica, and New Guinea. Only one locus ispolymorphic; nine are completely monomorphic, andthe remaining three are monomorphic in all but oneor two populations, which have a rare second allele.In contrast, Plasmodium falciparum displaysextensive microsatellite polymorphism within andamong populations. We further have analyzed, in96 samples from the same 8 localities, 8 tandemrepeats (TRs) located on a 100-kb contiguouschromosome segment described as highlypolymorphic. Each locus exhibits 2-10 alleles in thewhole sample but little intrapopulation polymorphism(1-5 alleles with a prevailing allele in most cases).Eight microsatellite loci monomorphic in P. vivax arepolymorphic in three of five Plasmodium speciesrelated to P. vivax (two to seven individuals sampled).Plasmodium simium, a parasite of New Worldmonkeys, is genetically indistinguishable from P.vivax. At 13 microsatellite loci and at 7 of the 8 TRs,both species share the same (or most common)allele. Scarce microsatellite polymorphism mayreflect selective sweeps or population bottlenecksin recent evolutionary history of P. vivax; thedifferential variability of the TRs may reflect selectiveprocesses acting on particular regions of the

genome. We infer that the world expansion of P. vivaxas a human parasite occurred recently, perhaps<10,000 years ago.

17 Nair U, Bartsch H, Nair J.Alert for an epidemic of oral cancer due to use ofthe betel quid substitutes gutkha and pan masala: areview of agents and causative mechanisms.Mutagenesis 2004 Jul;19(4):251-262.

In south-east Asia, Taiwan and Papua NewGuinea, smoking, alcohol consumption and chewingof betel quid with or without tobacco or areca nutwith or without tobacco are the predominant causesof oral cancer. In most areas, betel quid consists ofa mixture of areca nut, slaked lime, catechu andseveral condiments according to taste, wrapped ina betel leaf. Almost all habitual chewers use tobaccowith or without the betel quid. In the last few decades,small, attractive and inexpensive sachets of betelquid substitutes have become widely available.Aggressively advertised and marketed, often claimedto be safer products, they are consumed by the veryyoung and old alike, particularly in India, but alsoamong migrant populations from these areas worldwide. The product is basically a flavoured andsweetened dry mixture of areca nut, catechu andslaked lime with tobacco (gutkha) or without tobacco(pan masala). These products have been stronglyimplicated in the recent increase in the incidence oforal submucous fibrosis, especially in the very young,even after a short period of use. This precancerouslesion, which has a high rate of malignanttransformation, is extremely debilitating and has noknown cure. The use of tobacco with lime, betel quidwith tobacco, betel quid without tobacco and arecanut have been classified as carcinogenic to humans.As gutkha and pan masala are mixtures of severalof these ingredients, their carcinogenic affect canbe surmised. We review evidence that stronglysupports causative mechanisms for genotoxicity andcarcinogenicity of these substitute products.Although some recent curbs have been put on themanufacture and sale of these products, urgentaction is needed to permanently ban gutkha and panmasala, together with the other established oralcancer-causing tobacco products. Further, educationto reduce or eliminate home-made preparationsneeds to be accelerated.

18 Nakano T, Lu L, Liu P, Pybus OG.Viral gene sequences reveal the variable history ofhepatitis C virus infection among countries.J Infect Dis 2004 Sep 15;190(6):1098-1108.

BACKGROUND: The analysis of molecularphylogenies estimated from the gene sequences ofsampled viruses can provide important insights intoepidemiological processes. METHODS: Thedemographic and migration histories of the prevalenthepatitis C virus (HCV) subtypes 1a and 1b wereinferred from viral gene sequences sampled in 5countries. Estimated viral phylogenies were analyzedby use of methods based on parsimony andcoalescent theory. RESULTS: The parsimonymigration analysis suggested that the global subtype

236


1a and 1b epidemics are geographically structured,with asymmetrical movement of HCV strains amongthe sampled countries. The coalescent analysisindicated that subtype 1a infections in the UnitedStates, Brazil, and Indonesia began to increaseexponentially during the 1940s and 1950s, whereasin Vietnam the increase began after the 1970s. Incontrast, subtype 1b infections in these 4 countriesand in Japan began to increase exponentiallybetween 1880 and 1920, with a possible recentdecrease in infection rates in Indonesia and Japan.In the United States, Brazil, and Vietnam, theepidemic growth rates for subtype 1a strains werehigher than those for subtype 1b strains, whereasthe growth rates were similar in Indonesia.CONCLUSIONS: The estimated histories ofmigration and population growth indicated thatpatterns of HCV transmission differ among countriesand viral subtypes.

19 Oku N, Gustafson KR, Cartner LK, Wilson JA,Shigematsu N, Hess S, Pannell LK, Boyd MR,McMahon JB.Neamphamide A, a new HIV-inhibitory depsipeptidefrom the Papua New Guinea marine spongeNeamphius huxleyi.J Nat Prod 2004 Aug;67(8):1407-1411.

A new HIV-inhibitory cyclic depsipeptide,neamphamide A (2), was isolated from a Papua NewGuinea collection of the marine sponge Neamphiushuxleyi. Its structure was established throughinterpretation of spectroscopic data and by acidhydrolysis, derivatization of the free amino acids,and LC-MS analysis of the derivatives.Neamphamide A (2) contains 11 amino acid residuesand an amide-linked 3-hydroxy-2,4,6-trimethylheptanoic acid moiety. The amino acidconstituents were identified as L-Leu, L-NMeGln, D-Arg, D- and L-Asn, two residues of D-allo-Thr, L-homoproline, (3S,4R)-3,4-dimethyl-L-glutamine,beta-methoxytyrosine, and 4-amino-7-guanidino-2,3-dihydroxyheptanoic acid. In a cell-based XTT assay,2 exhibited potent cytoprotective activity against HIV-1 infection with an EC50 of approximately 28 nM.

20 Over M, Bakote’e B, Velayudhan R, Wilikai P,Graves PM.Impregnated nets or DDT residual spraying? Fieldeffectiveness of malaria prevention techniques inSolomon Islands, 1993-1999.Am J Trop Med Hyg 2004 Aug;71(2 Suppl):214-223.

The incidence of malaria in Solomon Islands hasbeen decreasing since 1992. The control programused a combination of methods including DDTresidual house spraying and insecticide-treatedmosquito nets. To determine how much each methodcontributed to malaria control, data were analyzedon monthly incidence and on control activities for 41of 110 malaria zones over the same time period(January 1993 to August 1999). After correction forendogeneity, then spraying, insecticide treatment ofnets, and education about malaria are allindependently associated with reduction in incidentcases of malaria or fever, while larviciding with

temephos is not. The evidence suggests thatalthough impregnated bed nets cannot entirelyreplace DDT spraying without substantial increasein incidence, their use permits reduced DDTspraying. The paper shows that non-experimentaldata can be used to infer causal links inepidemiology, provided that instrumental variablesare available to correct for endogeneity.

21 Pozio E, Owen IL, Marucci G, La Rosa G.Trichinella papuae in saltwater crocodiles(Crocodylus porosus) of Papua New Guinea.Emerg Infect Dis 2004 Aug;10(8):1507-1509.

22 Quinnell RJ, Pritchard DI, Raiko A, Brown AP,Shaw MA.Immune responses in human necatoriasis:association between interleukin-5 responses andresistance to reinfection.J Infect Dis 2004 Aug 1;190(3):430-438. Epub 2004Jun 25.

Cytokine and proliferative responses to Necatoramericanus infection were measured in a treatment-reinfection study of infected subjects from an areaof Papua New Guinea where N. americanus is highlyendemic. Before treatment, most subjects produceddetectable interleukin (IL)-4 (97%), IL-5 (86%), andinterferon (IFN)-gamma (64%) in response to adultN. americanus antigen. Pretreatment IFN-gammaresponses were negatively associated withhookworm burden, decreasing by 18 pg/ml for eachincrease of 1000 eggs/gram (epg) (n=75; p<0.01).Mean IFN-gamma responses increased significantlyafter anthelmintic treatment, from 166 to 322 pg/ml(n=42; p<0.01). The intensity of reinfection wassignificantly negatively correlated with pretreatmentIL-5 responses, decreasing by 551 epg for each 100pg/ml increase in production of IL-5 (n=51; p<0.01).These data indicate that there is a mixed cytokineresponse in necatoriasis, with worm burden-associated suppression of IFN-gamma responsesto adult N. americanus antigen. Resistance toreinfection is associated with the parasite-specificIL-5 response.

23 Rolett B, Diamond J.Environmental predictors of pre-Europeandeforestation on Pacific islands.Nature 2004 Sep 23;431(7007)443-446.

Some Pacific island societies, such as those ofEaster Island and Mangareva, inadvertentlycontributed to their own collapse by causing massivedeforestation. Others retained forest cover andsurvived. How can those fateful differences beexplained? Although the answers undoubtedlyinvolve both different cultural responses of peoplesand different susceptibilities of environments, howcan one determine which environmental factorspredispose towards deforestation and which towardsreplacement of native trees with useful introducedtree species? Here we code European-contactconditions and nine environmental variables for 81sites on 69 Pacific islands from Yap in the west toEaster in the east, and from Hawaii in the north to

237


New Zealand in the south. We thereby detectstatistical decreases in deforestation and/or forestreplacement with island rainfall, elevation, area,volcanic ash fallout, Asian dust transport andmakatea terrain (uplifted reef), and increases withlatitude, age and isolation. Comparative analysesof deforestation therefore lend themselves to muchmore detailed interpretations than previouslypossible. These results might be relevant to similardeforestation-associated collapses (for example,Fertile Crescent, Maya and Anasazi) or the lackthereof (Japan and highland New Guinea) elsewherein the world.

24 Rubin GL, Baird JH.New recommendation on Japanese encephalitisvaccination for travellers to Papua New Guinea.Med J Aust 2004 Sep 6;181(5):283.

25 Sankaranarayanan S, Untoro J, Erhardt J, GrossR, Rosales FJ.Daily iron alone but not in combination withmultimicronutrients increases plasma ferritinconcentrations in Indonesian infants withinflammation.J Nutr 2004 Aug;134(8):1916-1922.

Iron deficiency is a public health problem ininfancy. We assessed the efficacy of ironsupplements in infants with inflammation on ironstatus and subsequent inflammation. This was aprospective, nested, case-control study of 6- to 12-month-old infants participating in the InternationalResearch on Infant Supplementation study,Indonesia. Cases (n = 46) were selected on the basisof their inflammation status at baseline, C-reactiveprotein (>5 mg/l) or alpha-1 acid glycoprotein (>1 g/l); there were 44 controls without inflammation.Infants received 10 mg/day of elemental iron aloneor in combination with multimicronutrients, orplacebo. Blood samples were collected at baselineand at 6 months for determinations of plasma ferritin,zinc, copper, retinol, beta-carotene, alpha-tocopherol, and inflammation status. Data on breast-feeding and acute respiratory infections (ARI) werecollected daily. At baseline, 33% of infants had irondeficiency, and those with inflammation had lowerretinol, beta-carotene, higher concentrations ofcopper and higher rates of ARI compared withcontrols. After 6 months, compared with infants givenplacebo, ferritin concentration increased significantlyin infants administered iron alone independently ofinflammation status at baseline or at the end of thestudy. In those given multimicronutrients with iron,ferritin increased significantly in infants who did nothave inflammation at baseline or at the end of thestudy compared with those given placebo.Consequently, iron alone resolved iron deficiency,whereas multimicronutrients reduced the

deterioration of iron stores compared with placebo(÷2, p <0.05), without enhancing inflammation. Ironalone is recommended in populations in which irondeficiency is a public health problem despite thepresence of inflammation in infants who are stillbreast-feeding.

26 Sweeney AW, Blackburn CRB, Rieckmann KH.Short report: the activity of pamaquine, an 8-aminoquinoline drug, against sporozoite-inducedinfections of Plasmodium vivax (New Guineastrains).Am J Trop Med Hyg 2004 Aug;71(2):187-189.

It was reported in 1946 that the administrationof pamaquine during the incubation period delayedbut did not prevent primary attacks of a New Guineastrain of Plasmodium vivax malaria. The observationthat none of the four test subjects in this study hadrelapses has not previously been published and mayhave important implications for the evaluation ofother 8-aminoquinoline compounds againstrelapsing vivax malaria.

27 Vollaard AM, Ali S, van Asten HA, Ismid IS,Widjaja S, Visser LG, Surjadi Ch, van Dissel JT.Risk factors for transmission of foodborne illness inrestaurants and street vendors in Jakarta, Indonesia.Epidemiol Infect 2004 Oct;132(5):863-872.

In a previous risk factor study in Jakarta weidentified purchasing street food as an independentrisk factor for paratyphoid. Eating from restaurants,however, was not associated with disease. To explainthese findings we compared 128 street food-vendorswith 74 food handlers from restaurants in a cross-sectional study in the same study area. Poor hand-washing hygiene and direct hand contact with foods,male sex and low educational level were independentcharacteristics of street vendors in a logisticregression analysis. Faecal contamination of drinkingwater (in 65 % of samples), dishwater (in 91 %) andice cubes (in 100 %) was frequent. Directlytransmittable pathogens including S. typhi (n = 1)and non-typhoidal Salmonella spp. (n = 6) wereisolated in faecal samples in 13 (7 %) vendors; thegroups did not differ, however, in contamination ratesof drinking water and Salmonella isolation rates instools. Poor hygiene of street vendors compared torestaurant vendors, in combination with faecalcarriage of enteric pathogens including S. typhi, mayhelp explain the association found betweenpurchasing street food and foodborne illness, inparticular Salmonella infections. Public healthinterventions to reduce transmission of foodborneillness should focus on general hygienic measuresin street food trade, i.e. hand washing with soap,adequate food-handling hygiene, and frequentrenewal of dishwater.

Papua New Guinea Institute of Medical Research Monograph Series

ISSN 0256 2901

1. Growth and Development in New Guinea.A Study of the Bundi People of theMadang District.L.A. Malcolm. ISBN 9980 71 000 4, 1970,105p.

2. Endemic Cretinism.B.S. Hetzel and P.O.D. Pharoah, Editors.ISBN 9980 71 001 2, 1971, 133p.

3. Essays on Kuru.R.W. Hornabrook, Editor. ISBN 9980 71002 0 (also 0 900848 95 2), 1976, 150p.

4. The People of Murapin.P.F. Sinnett. ISBN 9980 71 003 9 (also0 900848 87 1), 1977, 208p.

5. A Bibliography of Medicine and HumanBiology of Papua New Guinea.

R.W. Hornabrook and G.H.F. Skeldon,Editors. ISBN 9980 71 004 7, 1977,335p. (with 1976 Supplement, 36p.)

6. Pigbel. Necrotising Enteritis in PapuaNew Guinea.

M.W. Davis, Editor. ISBN 9980 71 0055, 1984, 118p.

7. Cigarette Smoking in Papua NewGuinea.D.E. Smith and M.P. Alpers, Editors.ISBN 9980 71 006 3, 1984, 83p.

8. Village Water Supplies in Papua NewGuinea.

D.E. Smith and M.P. Alpers, Editors,ISBN 9980 71 007 1, 1985, 94p.

9. The Health of Women in Papua NewGuinea.Joy E. Gillett. ISBN 9980 71 008 X,1990, 180p.

10. National Study of Sexual andReproductive Knowledge and Behaviourin Papua New Guinea.

The National Sex and ReproductionResearch Team and Carol Jenkins.ISBN 9980 71 009 8, 1994, 147p.

Monographs 1-5 are case-bound, 6-10 arepaperbacks.Monographs may be obtained from

The Librarian,Papua New Guinea Institute of

Medical ResearchPO Box 60, Goroka, EHP 441,Papua New Guinea

Cost of each Monograph (see below forPostage and Handling):

1,2……………………………………K 5.003,4……………………………………K 8.005……………………………………...K 12.006,7,8,9……………………………….K 6.0010…………………………………….K 12.00

Applications for free copies of anymonograph should be sent to the Director atthe above address.

)aniKGNP(gnildnaHdnaegatsoP

LIAMECAFRUS LIAMRIA

nihtiWGNP GNPnihtiW 1enoZ 4/3enoZ 6enoZ

01,2,1 00.7 00.01 00.02 00.06 00.57

5,4,3 00.41 00.02 00.04 00.09 00.501

9,8,7,6 05.3 00.5 05.01 05.71 05.71

tneiciffusddaesaelp,ycnerrucrehtoynaniedamsitnemyapfI.aniKnitnemyapekamesaelP.aniK=KGP=K.segrahcknabllarevocotsdnuf

THE MEDICAL SOCIETY OF PAPUA NEW GUINEA

Society Membership and Journal Subscription

Membership of the Medical Society of Papua New Guinea is open to all health workerswhether resident in Papua New Guinea or overseas. Members of the Society receive fourissues of the Papua New Guinea Medical Journal each year. The Society organizes anannual symposium and other activities.

Membership dues are:-

Papua New Guinea residents:Members – K150Associate (Student) Members – K20Overseas residents: K200; AU$120; US$80

I wish to join the Medical Society of PapuaNew Guinea as aFull Member

Please indicate your category

Medical Officer [ ]Scientific Officer [ ]Pharmacist [ ]Health Extension Officer [ ]Nursing Officer [ ]Laboratory Technologist [ ]Radiographer [ ]Social Health Worker [ ]Other (Please specify) [ ]

OR a Student Member(for full-time students)

Medical Student [ ]Other Student (Please specify) [ ]

I enclose my membership fee of

K…………..for the year(s)……………

Name: …………………………………

Title: …………………………………..

Address: ……………………………...

………………………………………….

………………………………………….

Telephone: …………………………..

Fax: …………………………………..

Email: …………………………………

(Forward to the Membership Secretary,Medical Society of Papua New Guinea, POBox 60, Goroka, EHP 441, Papua NewGuinea)

INFORMATION FOR AUTHORS

The Papua New Guinea Medical Journal invitessubmission of original papers and reviews on allaspects of medicine. Priority will be given toarticles and subjects relevant to the practice ofmedicine in Papua New Guinea and othercountries in the South Pacific.

Manuscripts are accepted for publication onlywith the understanding that they have not beenpublished nor submitted for publicationelsewhere. All manuscripts will be sent out forreferees’ comments as part of the peer reviewprocess.

Original Articles: Reports of original and newinvestigations or contributions.

Brief Communications and Case Reports:Contents similar to that of original articles buttext should be no more than a total of 4 Journalpages including all figures and tables.

Reviews: Critical analysis of previouslycollected and published information.

Letters: Short reports of clinical experience ortopics of interest. Text should not exceed 2 pagesof the Journal.

Other types of manuscript may also beaccepted for publication at the Editor’s discretion.

Submitted manuscripts should conform to theinstructions set out below. Manuscripts notconforming to these instructions will be returned.

MANUSCRIPTS

Submit the original with a virus-free electroniccopy on disk, and specify the software used. Allsections of the manuscript, including text,references, tables and legends, should be indouble spacing. Manuscripts should not be rightjustif ied. Each paper should include aninformative Summary, Introduction, Patients/Materials and Methods, Results, Discussion andReferences. The title page should include thetitle, full names of all authors, names andaddresses of institutions where the work hasbeen done and full present address of the first orcorresponding author.

References should be kept to a minimum, andmust be in the Vancouver style. Authors shouldcheck all references against the original source.Sample references are shown below.

3 Garner PA, Hill G. Brainwashing intuberculosis management. PNG Med J1985;28:291-293.

4 Cochrane RG. A critical appraisal of thepresent position of leprosy. In: LincicomeDP, ed. International Review of TropicalMedicine. New York: Academic Press,1961:1-42.

ILLUSTRATIONS

Tables and figures should be prepared onseparate pages. Photographs should be glossyprints, either 7 cm or 14.5 cm in width. Graphsand charts should preferably be in Microsoft Wordor Excel. Photomicrographs should have internalscale markers. Each table should have a headingand footnotes which make it understandablewithout reference to the text. Each figure shouldhave a legend; figure legends should be typedtogether on a separate sheet. Indicate the top offigures lightly in pencil on the back.

Abbreviations: Standard abbreviations andunits should be used.

Drug Names: Generic names of drugs shouldbe used.

Orthography: The Shorter Oxford EnglishDictionary is followed.

EDITORIAL MAIL

Manuscripts and other editorialcommunications should be forwarded to:

The Editor, Papua New Guinea Medical Journal, PO Box 60, Goroka, EHP 441, Papua New Guinea Email: [email protected]

SUBSCRIPTIONS AND ADVERTISEMENTS

Communications relating to advertisements orsubscriptions should be addressed to the Journalas above. Matters related to the Society shouldbe addressed to the Medical Society of PapuaNew Guinea, PO Box 6665, Boroko, NCD 111,Papua New Guinea.

Subscriptions: Members of the Medical Societyof Papua New Guinea receive the Journal as partof their annual subscription. Others maysubscribe and should contact the subscriptionsecretary for a price.

Papua New Guinea Medical Journal Volume 47, Number 3-4, September-December 2004

EDITORIAL Malaria epidemics in the highlands of Papua New Guinea L. Dapeng

ORIGINAL ARTICLESThe management of children with cancer in Papua New Guinea: a review of

children with cancer at Port Moresby General Hospital M. Kiromat, J.D. Vince, G. Oswyn and N. Tefuarani

Zinc in human health V.J. Temple and A. Masta

The epidemiology of malaria in the Papua New Guinea highlands: 3. SimbuProvince I. Mueller, J. Kundi, S. Bjorge, P. Namuigi, G. Saleu, I.D. Riley and J.C.Reeder

A comparison of booked and unbooked mothers delivering at the Port MoresbyGeneral hospital: a case-control study F. Failing, P. Ripa, N. Tefuarani and J.Vince

Strongyloides fuelleborni kellyi and other intestinal helminths in children fromPapua New Guinea: associations with nutritional status and socioeconomicfactors S.E. King and C.G.N. Mascie-Taylor

How the poor die in the settlements of Port Moresby, 2003-2004 P. Sims

Antenatal care in Goroka: issues and perceptions G.L. Larsen, S. Lupiwa, H.Paito Kave, S. Gillieatt and M.P. Alpers

Rehabilitation services in Papua New Guinea A. Shaw

MEDICAL RESEARCH PROJECTS IN PAPUA NEW GUINEA


CONTENTS

135

138

146

159

174

181

192

202

215

228

231

VOL. 47, NO 3-4, SEPTEMBER-DECEMBER 2004

Documents

Transcript of VOL. 47, NO 3-4, SEPTEMBER-DECEMBER 2004