Tropical Medicine and International Health

volume 5 no 9 pp 657–665 september 2000

Maternity care in rural Nepal: a health service analysis

Albrecht Jahn1, Maureen Dar Iang2, Usha Shah3 and H. J. Diesfeld1

1 Department of Tropical Hygiene and Public Health, Heidelberg University, Heidelberg, Germany

2 Okhaldunga Health Project, Nepal

3 Bheri Zonal Hospital, Nepalganj, Nepal

Summary This study assesses the performance of maternity care and its specific service components (preventive

interventions in antenatal care, antenatal screening, referral, obstetric care) in Banke District, Nepal, using a

set of structure, process, and output/outcome indicators. Data sources included health service documents in

14 first level health units and two hospitals, covering 1378 pregnancies and 1323 deliveries, structured

observations, antenatal exit interviews (n 5 136) and interviews with maternity users (n 5 146). Coverage of

antenatal care (28%) and skilled delivery care (16%) was low. In antenatal care, preventive interventions were

only partially implemented (effective iron supplementation in 17% of users). On average one minute was

spent on individual counselling per consultation. 41% of pregnancies were identified as high risk and 15%

received referral advice, which was followed in only 32%. Hospital deliveries accounted for 9.8% of all

deliveries. Hospital-based maternal mortality was 6.8/1000 births and the stillbirth rate 70/1000. High rates

of stillbirth were observed in breech delivery (258/1000 births), caesarean section (143/1000) and twin

delivery (133/1000). The risk of stillbirth was higher for rural women (RR 2.3; 95% CI 1.51–3.50) and

appeared to be related to low socio-economic status. Emergency admissions were rare and accounted for

3.4% of hospital deliveries or only 0.4% of all expected deliveries. There was hardly any accumulation of

high-risk pregnancies at hospital. The population-based rate of caesarean section was 1.1% (urban 2.3%,

rural 0.2%). The estimated unmet obstetric need was high (82 cases or 61% of expected live-threatening

maternal conditions did not receive appropriate intervention). The limited effectiveness of maternity care is

the result of deficiencies of all service components. We propose a two-pronged approach by starting quality

improvement of maternity care from both ends of maternity services: preventive interventions for all women

and hospital-based obstetric care. Antenatal screening needs to be rationalized by reducing inflated risk

catalogues that result in stereotypical and often rejected referral advice.

keywords safe motherhood, health systems research, maternity care, quality of care, Nepal

correspondence Dr Albrecht Jahn, Department of Tropical Hygiene and Public Health, Heidelberg

University, Im Neuenheimer Feld 324, 69120 Heidelberg, Germany. Fax: 149 6221 565 037, E-mail:

Albrecht.Jahn@urz.uni-heidelberg.de

Introduction

Ten years of Safe Motherhood interventions have failed to

deliver a substantial reduction in the high maternal and

perinatal mortality in many developing countries. This casts

doubts on their health systems’ ability to provide effective

maternity care. Therefore we set out to identify possible causes

for this system failure by following the pathway in maternity

care from preventive interventions in antenatal care (ANC), to

screening and identification of high-risk pregnancies, referral

and hospital-based obstetric care (Figure 1). This approach is

based on the experience that substantial and sustainable

improvements in maternal health cannot be achieved with a

unifocal intervention, but depend on the co-ordinated function

of all service components (Fonn et al. 1998). Hence

identification of high-risk pregnancies and deliveries is only

meaningful in the presence of means for referral and

subsequent obstetric management according to quality

standards as outlined in WHO’s essential elements of referral

level care (WHO 1991a).

This study was conducted in Banke, a rural district in the

mid-west of Nepal. According to WHO, Nepal’s maternal

mortality ratio is among the highest in Asia with 1500/100 000

live births (WHO & UNICEF 1996); national data are lower

(539/100 000) but still high (Pradhan et al. 1996). Perinatal

mortality is also elevated with 75/1000 live births (WHO

TMIH611

© 2000 Blackwell Science Ltd 657

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

1996a). Maternal health is a national health priority and

several interventions have been implemented in recent years

under the national safe motherhood programme (MoH 1996).

These include community information and mobilization,

expanding services through the introduction of a new cadre

(mother-child health workers) and strengthening hospitals’

capacity to provide essential obstetric care. However, use of

maternity services on national level is still low (ANC coverage

29% and institutional delivery care coverage 10%) (Pradhan

et al. 1996).

According to the national guidelines (UNICEF & MoH

1996), maternity services aim to help families take appropriate

decisions through health information and counselling, to

provide basic antenatal and delivery services to all pregnant

women and to ensure referral and adequate obstetric care to

high-risk mothers and obstetric emergencies. The objective of

this study was to assess maternity services along the main

respective service components (Figure 1), using the framework

of structure, process and output/outcome quality (Donabedian

1988). Specifically, we investigated the performance of main

service components, appropriateness of referral practices and

use of referral level obstetric care in relation to risk status and

obstetric complications, and the overall system performance in

terms of coverage of obstetric need.

Materials and methods

Setting

The fieldwork was done in Banke District, Mid-Western

Development Region, Nepal from June to September 1997.

The district covers a plain area with a population of 338 000

(projection from 1991 census) and 13 520 expected deliveries

per year (crude birth rate 41.5/1000 pop.). Most people live in

rural areas (83.4% rural, 16.6% urban). The public maternity

facilities include the Bheri zonal hospital in the district capital

Nepalganj at referral level, and at first service level three

primary health care centres (PHC), 10 health posts (HP) and

30 sub-health posts (SHP). All but two of the first-line units are

located in rural areas. There are also private clinics and one

private hospital in the district capital providing delivery care.

Bheri zonal hospital is the main provider of obstetric care

for the district. Its actual target population in Nepalganj

© 2000 Blackwell Science Ltd658

Figure 1 Maternity service organization.Antenatal Care I:

Preventive activities for all pregnant women:

Prevention of malaria and anaemia

Tetanus vaccination

Health information/education/counselling

Target group:all pregnant women

Antenatal Care II:

Screening: Identification of risk factors andhealth problems

History taking

Physical examination

Bio-medical tests(e.g. sero-test for syphilis)

Antenatal Care III:

Interventions according to risk factors andhealth problems identified:

Management at first service level

Referral of high risk pregnancies anddeliveries

Essential Obstetric Care at Hospital:

Essential elements of referral level care (WHO,1991):

(1) Surgical obstetrics (2) Anaesthesia (3) Med.treatment and essential drugs, (4) Bloodreplacement (5) Monitoring of labour(6) Neonatal special care

Emergencies

Self-referralsEmergencies

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

extends to neighbouring districts, many of which lack

permanent obstetric services. Outflow of patients is likely to be

minimal, as the next reliable referral facilities are very far away

(Kathmandu 14 h drive, India 6 h drive). The zonal hospital

was staffed with doctors (one obstetrician) and six qualified

nurses. It offers comprehensive obstetric care including

emergency services on a 24-h basis and ANC on 5 days per

week.

Maternity services in first-line units (ANC and delivery care)

were provided by staff nurses (training 3 years) and auxiliary

nurse midwives (training 18 months). In primary health centres

(PHC) there were auxiliary nurse midwives, and in health posts

(HP) and subhealth posts (SHP) maternal child health workers

(training 10 weeks).

Study design

As suggested by St Leger et al. (1992), we pursued a pragmatic

study approach: based on our health service model (Figure 1),

we identified indicators along the categories of structure,

process and output/outcome (Donabedian 1988) and selected

the appropriate tools for data collection. These included the

analysis of health service data (Ronsmans et al. 1999),

structured observation with checklists and interviews with

health workers and their clients.

Indicators used for the assessment of ANC were

accessibility; availability of essential drugs and equipment;

coverage of ANC and delivery care; coverage of tetanus

immunization and anaemia prophylaxis; duration and content

of counselling; completeness of implementation and sensitivity

of screening; rate of referral advice in high-risk pregnancies;

referral compliance; and user satisfaction. Indicators used for

the assessment of hospital-based obstetric care were

availability of essential elements of obstetric care according to

WHO (WHO 1991a); rate of hospital deliveries; CS rate; time

lag in emergency management; stillbirth rate; maternal

mortality rate; and rate of obstetric emergency admissions.

Indicators for the overall performance of maternity services

were coverage of obstetric care for specific risk groups (Jahn

et al. 1998) and major obstetric interventions (MOI) for

absolute maternal indications (AMI) according to the unmet

obstetric need approach (De Brouwere et al. 1996; Belghiti

et al. 1998). Many of these indicators were stratified for urban

and rural origin. The national maternity care guidelines of

Nepal (UNICEF & MoH 1996) were used as standard unless

specified otherwise.

Selection of health facilities

A stratified random sample of first-line units was done with

rural PHCs and HPs in one stratum and SHPs in another

stratum. The only urban HP was selected purposely. Five

PHC/HPs and nine SHPs were selected. Their distance from

Nepalganj ranged from 6 to 60 km.

Data sources

The data on structural quality were obtained by checklists

covering 14 tracer items for essential equipment, drugs and

consumables at first-line health facilities and 22 items at the

hospital. These checklists were developed along WHO

standards (WHO 1996b) and national maternity guidelines

(UNICEF & MoH 1996) jointly with health professionals in

Nepal. Antenatal care (ANC) and referral were assessed by

structured observation of first antenatal consultations

(n 5 58), exit interviews with users of ANC (n 5 136) and

analysis of antenatal registers in 14 first-line units, the hospital

and private clinics covering 1378, 772 and 873 first antenatal

consultations, respectively. Coverage of ANC in urban areas

was based on first antenatal visits recorded in all facilities

providing antenatal care there. In rural areas, coverage was

calculated based on first visits in the selected 13 rural health

units and their respective target populations.

Coverage of referral level obstetric care was assessed by

analysing hospital registers (delivery and theatre book,

discharge register) and case notes covering 1323 deliveries from

April 1996 to April 1997 of women residing in Banke District.

These sources were also used to estimate the unmet obstetric

need indicator (De Brouwere et al. 1996; Belghiti et al. 1998).

We interviewed the users of maternity services (n 5 146)

during a 3-month period (mid-June to mid-September 1997)

and reviewed their case notes to determine the cause for

admission, referral status, risk status, and cost incurred. These

interviews were also used to check the validity of register data

and revealed consistent information on place of residence, risk

factors and indications for caesarean section. The inpatient

interviews covered 90% of hospital deliveries in the study

period, as some women could not be interviewed because of

early discharge.

The data on the spatial population distribution was

obtained from the District Development Committee Office,

Nepalganj and from Village Development Committees.

Reference values for risk factor prevalence were derived from

the 1996 Nepal Family Health Survey (Pradhan et al. 1996) for

age, parity, previous caesarean section and the national

maternity care guidelines (UNICEF & MoH 1996) for breech

and twin pregnancies.

Results

Overall health infrastructure and accessibility of maternity

services

Forty-two first-line health units serve 338 000 population.

90% live within 5 km from a health facility. However, only

© 2000 Blackwell Science Ltd 659

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

67% of first-line units offered ANC services at facility and/or

outreach points (usually 2–4 days per week) and only 38%

provided skilled assistance in home deliveries. There is no

facility-based delivery care in first-line units. Essential

obstetric care is provided at two hospitals in the district

capital. Almost half (45%) of the district population resides

within 10 km of the district capital, 43% within 11–30 km

and 12% beyond 30 km. But accessibility to referral level care

is hampered by irregular transport and seasonal flooding.

Most antenatal services are free of charge except fees for

medicine such as iron tablets and laboratory investigations

offered at hospital outpatient departments. Hospital care is

rather expensive with 300–3000 rupees (mean 600 5 11 US$)

for vaginal delivery and 1000–13000 rupees (mean 5360 5 96

US$) for a caesarean section (CS) in comparison to the 90

rupees (5 1.6 US$) minimum daily wage for an unskilled

labourer.

Quality of antenatal and delivery care in first line health

facilities

Overall, 69% of items from a tracer list for essential

equipment and drugs were available in the 14 selected

facilities. Availability was good for thermometers (100%),

sphygmomanometers (92%), foetoscopes (92%), and poor for

technical guidelines (7%), urinary catheter (7%), tape

measure (14%), sterile gloves (21%), ergometrine injection

(29%), and consultation room with privacy (43%).

Coverage of ANC was low with 28% (urban 50%; rural

24%). Only 37.5% of first consultations took place before

20 weeks of gestation. Coverage of ANC in rural areas,

stratified for distance from health facilities, was 32%

(# 5 km) and 10% (. 5 km). Health services were involved in

15.8% of deliveries: 9.8% being hospital deliveries and 6%

being home deliveries conducted by first-level health workers.

Most women delivered with a traditional birth attendant

(trained 18%, untrained 66.2% of total deliveries).

Direct observation of 58 antenatal consultations revealed

an average duration of 10 min (5–15). On average only one

minute was spent on health education and counselling . The

main topics were nutrition (85%), personal hygiene (71%)

and a general reassurance to women that ‘everything is all

right’ (86%). Other essential topics received little attention:

place of delivery and individual delivery plan (17%), danger

signs (12%) and family planning (7%). Communication was

mainly unidirectional and paid little attention to

interpersonal aspects. 65% of ANC users in exit interviews

complained that they were not given the time and

opportunity to ask questions. Although technical quality of

antenatal examinations was rated as satisfactory by most

clients (77%), only 43% were satisfied with counselling and

reassurance.

The observed low coverage of iron supplementation is the

result of a series of deficiencies (Figure 2): to start with, only

28% of pregnant women attended ANC. Only 32% of those

who attended for ANC (44 of 136), or 9% of all pregnant

women, got a prescription for iron tablets. Of those with a

prescription, only 65% received or bought iron tablets.

Lapses in compliance resulted in a further drop of the

coverage of effective anaemia prophylaxis, defined as

continuous supplementation for at least 30 days, to 4.8% of

expected pregnancies or 17% of ANC attendees. Coverage of

tetanus vaccination (at least two doses) was 51.4%.

Physical examinations and blood pressure measurements

were performed in 85% to 95% of the consultations, but no

biochemical tests (urine analysis, sero-testing for syphilis)

were performed. Taking the women’s history focused on the

present pregnancy; complications in previous pregnancies

were never asked after. 15% of antenatal attendees received

referral advice in the course of ANC, while 41.5% of

attendees were documented to be high risk according to the

national risk catalogue. Most referral advice (52%) was based

on the demographic risk factors young age (, 19 years),

nulliparity and multiparity. However, referral practices were

inconsistent and even in these risk groups, the majority (65%

to 75%) did not receive a referral advice.

© 2000 Blackwell Science Ltd660

Figure 2 Stepwise loss of coverage of

anaemia prophylaxis.

0Received

effective dosage

30 28.0

% o

f exp

ecte

d de

liver

ies

ANC coverage

25

20

15

10

5

Received irontablets

Received ironprescription

9.0

5.4 4.8

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

Referral advice because of risk factors and symptoms of

higher clinical relevance, such as previous stillbirth, breech

presentation, twin pregnancy, clinical pallor, weakness, and

high blood pressure were rare and covered less than 10% of the

respective risk groups. No referral was documented for risk

factors such as bleeding in pregnancy, history of previous CS

and previous post-partum haemorrhage (PPH). A cross-check

with hospital data revealed that ANC attendees’ compliance

with referral advice was only 32%. The cumulative effect of

low coverage of ANC, poor screening, inconsistent referral

advice and low user compliance (Figure 3) dramatically

reduced the proportion of identified and properly referred

high-risk pregnancies to 3.1% of all high-risk pregnancies

(11.2% of high-risk pregnancies attending ANC; 1.3% of all

pregnancies).

Quality of referral level obstetric care

Essential services (surgery, anaesthesia, medical treatment,

blood replacement, monitoring of labour, management of

high-risk pregnancies, neonatal care) were available.

However, there was no routine use of the partogramme.

Availability of essential equipment and drugs in the zonal

hospital was 75% according to a reference list of 22 items.

From April 1996–97 the two referral hospitals in the

district capital provided obstetric care at 1323 deliveries for

residents of Banke District, representing 9.8% of the 13520

expected deliveries. There was a wide urban–rural disparity

(urban 36% vs. rural 4.5%). The public hospital bore the

bulk of the obstetric workload (89% of deliveries and 97% of

CS). According to interviews with 146 maternity users at the

zonal hospital and review of their antenatal cards, 86% were

self-referred while 14% were referred through health services.

Most of those self-referred gave general considerations of

safety as sole reason for hospital use, and only 17%

mentioned specific health problems such as general weakness,

pain, bleeding and lack of foetal movement (in order of

priority).

The average time lapse between decision for and actual

performance of CS in 20 CS patients investigated was 4.5 h

(range 40 min – 11 h). As women were often discharged

shortly after delivery, the only reliably documented foetal

outcome parameter was stillbirth. The stillbirth rate among

the 1186 babies born in the zonal hospital was 70/1000. High

rates were observed in breech delivery (258/1000), CS

(143/1000), twin delivery (133/1000) and forceps extraction

(83/1000) (Table 1). The risk of stillbirth was much higher for

rural women (RR 2.3; 95% CI 1.51–3.50), possibly due to a

© 2000 Blackwell Science Ltd 661

Figure 3 Effectiveness of screening for high-

risk pregnancy in Banke district.

0Delivered at

hospital

30.0 28.0

% o

f exp

ecte

d de

liver

ies

ANC coverage

25.0

20.0

15.0

10.0

5.0

Advised hospitaldelivery

High riskpregnancy

attending ANC

11.6

4.2

1.3

Place of residence No. of No. of Stillbirth rate

& mode of delivery total births stillbirths (per 1000 total births)

Urban & vaginal 633 25 39

Rural & vaginal 406 37 91

Urban & CS 96 9 94

Rural & CS 51 12 235

Total 1186 83 70

Table 1 Stillbirth (SB) rate by mode of

delivery and place of residence at Bheri zonal

hospital (n 5 1186)

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

higher proportion of emergencies among rural women.

However, emergency admissions were rare and accounted for

only 3.4% of hospital deliveries or 0.4% of all deliveries in

the district. Four of the five emergency cases observed in the

study period were of urban residence. Risk profiles in urban

and rural women were similar except for a higher rate of

multiparity among rural women (13% vs. 20%). Eight

mothers died in 1171 deliveries (hospital-based MMR

683/100).

Overall maternity service performance in terms of care for

high-risk pregnancies and obstetric complications

The extent to which women in need are covered by adequate

obstetric care is influenced by all service components of

maternity care. Coverage of obstetric care is therefore a good

indicator of the overall performance of maternity services.

We measured this coverage using two approaches: (1) the

coverage of obstetric care for predefined risk groups

according to official risk factors and (2) the coverage of

major obstetric interventions (MOI) for absolute maternal

indications (AMI).

Comparison of risk profiles of users of hospital maternity

care and pregnant women in general showed that there is no

marked risk selection towards referral level care; the vast

majority of high-risk pregnancies were not covered by

obstetric care (Table 2). Nulliparity is the only factor leading

to a considerable accumulation at referral level. The

prevalence of other conditions is similar for both groups or

even lower among hospital users as in the case of multiparity.

This indicates an avoidance of obstetric care and thus rather

the opposite of the desired behaviour. As the prevalence of

pre-eclampsia varies widely even within countries, it was

omitted from Table 2. However, the observed hospital

prevalence of 1.2% is at the low end of values observed in

population-based studies (1.2–8.2%) (WHO 1991b). Severe

anaemia was never mentioned as an indication for

institutional referral or as a reason for admission.

CS was performed in 155 deliveries, resulting in a hospital-

based CS rate of 11.7%. The population-based CS rate is

1.1% of all deliveries in the district (urban 2.3% vs. rural

0.2%). Thirty-five percent of these CS were performed for

absolute maternal indications (obstructed labour 22%,

antepartum haemorrhage 9%, transverse lie 2%, uterine

rupture 2%). According to the unmet obstetric need

approach, this yields a rate of obstetric interventions for

absolute maternal indications of 3.9/1000 expected deliveries

against an estimated minimum requirement of 10/1000

(Belghiti et al. 1998). This is equal to 82 women (61%) per

year with life-threatening complications left untreated in

Banke District.

Discussion

According to WHO, the objectives of pregnancy-related

health services are (1) to prevent complications or to reduce

their adverse effects by early detection and treatment (2) to

ensure access to clean and safe delivery care and (3) to ensure

that essential obstetric care is made available to all women in

need (WHO 1996b). Health services in Banke District are still

far from meeting these objectives:

• At present, ANC confers little benefit to its users.

Efficacious preventive interventions such as anaemia

prophylaxis are poorly implemented. Antenatal

screening has a low sensitivity and fails to trigger an

adequate use of referral level obstetric care.

• Pregnancy outcomes in hospital are unsatisfactory and

show a wide urban–rural disparity.

• Neither institutional referral nor self-referral lead to a

considerable accumulation of high-risk pregnancies and

obstetric emergencies in the hospital maternity. Thus

© 2000 Blackwell Science Ltd662

Frequency of RF Frequency of RF Obstetric coverage

in population in hospital deliveries of risk group

Risk factor % % %

Nulliparity 23.0* 44.1 18.8

APH 3.0† 4.3 14.1

young age (# 19) 18.7* 21.0 11.0

multiple pregnancy 1.4† 1.5 10.5

breech presentation 3.0† 2.0 6.5

previous CS 0.9* 0.5 5.4

multiparity (. 4) 38.6* 13.7 3.5

* 5 Pradhan et al. (1996)

† 5 Unicef & MoH (1996)

Table 2 Prevalence of risk factors (RF)

among pregnant women in the population

and hospital delivery cases, Bheri zonal

hospital (n 5 1171)

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

most women in need of obstetric care either because of

their risk status or because of delivery complications

give birth at home without skilled assistance.

• Coverage of modern maternity services is low, even in

comparison with other countries in the region.

Quality of antenatal care

ANC is an essential element of maternity care. A core set of

efficacious diagnostic and preventive interventions, including

the tracer interventions used in this study, are advocated by

WHO and others (Enkin et al. 1995; WHO 1996b, c; Villar &

Bergsjo 1997). However, turning their beneficial potential into

real health gains depends on the technical quality of

implementation, user compliance and the functioning of

corresponding parts of the health system such as the referral

system. This study reveals shortcomings in all of these areas:

The implementation of preventive interventions is rather

incomplete; only a minority of users receives iron

supplementation (Figure 2) and individual counselling is

almost absent. Only tetanus vaccination reaches a reasonable

level of implementation. While most ANC users had their

blood pressure checked and were physically examined, target

conditions of these procedures such as pre-eclampsia and

multiple pregnancy were hardly ever detected. Low sensitivity

of antenatal screening is a common problem in developing

and developed countries. It has contributed to a rethinking of

the risk approach and related screening procedures (Jahn &

Kowalewski 1998) and is discussed below in the context of

obstetric coverage of high-risk pregnancies.

Quality of obstetric referral level care

The stillbirth rate at the zonal hospital in Banke (7%) is

higher than hospital or population-based stillbirth rates

reported from Nepal previously (1.5–3.4%) (WHO 1996a).

However, hospital data are difficult to compare because of

differences in user profiles (Mancey-Jones & Brugha 1997),

e.g. hospitals with a good reputation may attract more

emergencies. Given the low proportion of emergencies and

the modest risk selection in our study, the observed high rates

of stillbirths and maternal mortality in hospital appear to

reflect inappropriate management at the hospital maternity.

This is confirmed by the observation that every seventh CS

(14.3%) and every 4th breech delivery (23.4%) ended in a

stillbirth, which is hardly acceptable in a referral level facility.

This is partly explained by the considerable time lag between

indication and obstetric intervention. The patient’s social

background may also contribute: rural women had a 2.3

times higher risk of stillbirth than urban women. Late arrival

at hospital and a higher proportion of emergencies would

offer a plausible explanation. However, risk profiles and CS

rates were rather similar in both groups. Obstetric emergency

admissions were rare and mainly from the urban population.

Therefore, eventual differences in biomedical risk factors do

not provide a satisfactory explanation for the grossly

divergent stillbirth rates. Thus we hypothesize that quality of

obstetric care is also related to the client’s socio-economic

status, putting low-status rural women at a disadvantage.

Referral and obstetric coverage of high-risk pregnancies and

obstetric emergencies

There are two approaches to assess coverage of obstetric care:

firstly, coverage can be assessed according to specific

predefined risk groups. Like most ANC programmes, the

Nepalese maternity care guidelines stipulate referral for high-

risk pregnancies in anticipation of complications (UNICEF

& MoH 1996). Only 3.5–18.8% of the respective risk groups

were covered by obstetric care (Table 2). It is particularly

worrying that the majority of women with a previous CS

deliver at home, instead of having a trial of labour in a

hospital setting. In other developing countries, this risk group

was found to have a significant preference for hospital

delivery, resulting in a coverage of obstetric care well above

50% (Voorhoeve et al. 1984; Jahn et al. 1998).

Low effectiveness of antenatal risk assessment has also

been observed in other countries and many risk factors have

shown to be poor predictors of adverse outcomes (Maine

1991; McDonagh 1996; Geefhuysen et al. 1998; Jahn et al.

1998). The Safe Motherhood Initiative has thus

recommended abandoning the risk approach in ANC (Inter-

Agency Group for Safe Motherhood 1997). However,

national programmes are still built on it, and there is

reasonable evidence to justify antenatal identification and

referral for specific risk groups such as women with a history

of CS, breech presentation, multiple pregnancy and

hypertension (WHO 1996d; Villar & Bergsjo 1997). The

WHO trial, evaluating a new ANC programme restricted to

interventions with proven efficacy, includes screening

procedures in the intervention arm (Villar et al. 1998). Our

findings provide no definite answer to the current controversy

on risk assessment in ANC. However, we can conclude that

risk catalogues and related referral guidelines are often not

operational because they classify an unacceptably large

proportion of pregnancies (in our study above 40%) as high

risk and mask the wide differences in predictive properties

and clinical importance across risk factors.

The unmet obstetric need approach compares post hoc the

number of actually performed obstetric interventions (MOI

for AMI) with the estimated need for these life-saving

interventions (Belghiti et al. 1998). As delivery complications

are often unforeseeable, the observed difference is mainly an

indicator for emergency obstetric care. Our results show that

© 2000 Blackwell Science Ltd 663

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

only 39% of the estimated minimal need for these life-saving

interventions is covered. This corresponds to the low rate of

intrapartum referrals (0.37% of home deliveries) against an

estimated prevalence of 5% of severe delivery complications

even in low-risk pregnancies (Rooks et al. 1989; WHO

1996b). Thus, antenatal as well as emergency obstetric care

miss the majority of their respective target groups.

Acceptability of maternity services

The majority of Nepalese women have no pregnancy-related

contact with modern health services. Coverage of ANC

(national 29%, Banke District 28%) and skilled delivery

assistance (national 10%, Banke District 16%) are at the

bottom end of global statistics and compare unfavourably

with the south-central Asia region (ANC 52%; skilled

assistance 34%) and with the developing world in general

(ANC 65%; skilled assistance 53%) (WHO 1997). Though

far from being satisfactory, geographical accessibility in

Banke with one static maternity care service outlet per 12 000

population plus about 100 outreach service points compares

well with neighbouring countries. Even within a 5-km range

from health facilities, coverage of ANC is only 32%. This

indicates that there are strong reasons for non-use of

maternity services beyond geographical accessibility. These

may include cultural barriers, perceived low quality of care,

perceived discrimination of rural people and lack of

perceived health gain (Annis 1981; Thaddeus & Maine 1994;

MacRorie 1998; Kowalewski et al. 2000). Further research is

needed to clarify this issue.

Implications for health services

The overall low effectiveness of maternity services is the

cumulative result of deficiencies throughout the pathway

from preventive interventions to hospital-based obstetric care.

As it is almost impossible to address all of these

simultaneously, we propose a two-pronged and stepwise

approach by starting from both ends of maternity care

(preventive interventions in pregnancy and hospital-based

obstetric care). Interventions in these two areas produce an

immediate health benefit and there is a wide scope for

improvement even within limited resources (Mbaruku &

Bergström 1995). Provision of obstetric care, offering a

substantially better survival chance for high-risk pregnancies

than home management, is a prerequisite for meaningful

antenatal screening interventions and referral. Preventive

interventions to be intensified would include tetanus

vaccination and anaemia prophylaxis. Obstetric care has

been shown to improve considerably after the introduction of

systematic participatory evaluation procedures, which may

include perinatal and maternal audit, team-based analysis of

routine data and a streamlining of organizational procedures

(Mbaruku & Bergström 1995; Mancey-Jones & Brugha

1997).

The focus of ANC needs to be shifted from schematic and

ineffective allocation of risk labels to competent and

individualized counselling in order to empower mothers and

their families to take informed decisions. Lack of counselling

is also a major complaint of ANC users in this study.

Screening still has a place in this process as women’s key

question ‘Is my pregnancy all right?’ requires a technically

competent answer. The proposed re-orientation of ANC

towards the individual client implies abandoning long-

standing ritualistic practices and adopting new attitudes and

skills. This includes a modification of policies and maternity

guidelines, which still operate on a risk concept that classifies

a large proportion of pregnant women as high risk in a self-

defeating manner and compromises the credibility of

compliant health workers. Risk catalogues and referral

guidelines should be restricted to a core set of locally defined

and accepted risk factors and symptoms.

Acknowledgements

We gratefully acknowledge the co-operation and invaluable

help of Dr M. Kidwai, District Medical Officer and all staff

at the Zonal Hospital Nepalganj and the first-line health

units in the district. Ms. A. de Graaf was very helpful in

organizing the field visits. Mr G. Falkenhorst participated in

the development of indicators and Ms. M. Kowalewski

reviewed earlier drafts. The study was supported by the

Federal Ministry of Research and Technology, Germany

(grant 01 KA 9301/3) and the German Academic Exchange

Service (DAAD).

References

Annis S (1981) Physical access and utilization of health services in

rural Guatemala. Social Science and Medicine 15, 515–523.

Belghiti A, De Brouwere V, Kegels G & van Lerberghe W (1998)

Monitoring unmet obstetric need at district level in Marocco.

Tropical Medicine and International Health 3, 584–591.

De Brouwere V, Laabid A & van Lerberghe W (1996) Estimation des

besoins en interventions obstétricales au Maroc. Une approche

fondée sur l’analyse spatiale des déficits. Revue d’Epidemiologie et

de Sante Publique 44, 111–124.

Donabedian A (1988) The quality of care – How can it be assessed?

Journal of the American Medical Association 260, 1743–1748.

Enkin M, Keirse MJNC, Renfrew M & Neilson JP (1995) A Guide to

Effective Care in Pregnancy & Childbirth, 2 edn. Oxford University

Press, Oxford.

Fonn S, Xaba M, Tint K, Conco D & Varkey S (1998) Maternal

health services in South Africa During the 10th anniversary of the

WHO ‘Safe Motherhood’ Initiative. South African Medical Journal

88, 697–702.

© 2000 Blackwell Science Ltd664

Tropical Medicine and International Health volume 5 no 9 pp 657–665 september 2000

A. Jahn et al. Maternity care in rural Nepal

Geefhuysen CJ, Isa AR, Hashim M & Barnes A (1998) Malaysian

antenatal risk coding and the outcome of pregnancy. Journal of

Obstetrics and Gynaecology Research 24, 13–20.

Inter-Agency Group for Safe Motherhood (1997) The safe

motherhood action agenda: Priorities for the next decade.

Colombo.

Jahn A & Kowalewski M (1998) The risk approach in antenatal care:

Pitfalls of a global strategy. Curare 15, 195–209.

Jahn A, Kowalewski M & Kimatta SS (1998) Obstetric care in

Southern Tanzania: Does it reach those in need? Tropical Medicine

and International Health 3, 926–932.

Kowalewski M, Jahn A & Kimatta SS (2000) Why do at-risk mothers

fail to reach referral level care? Barriers beyond distance and cost.

African Journal of Reproductive Health 4, 100–109.

MacRorie RA (1998) Births, deaths and medical emergencies in the

district: a rapid participatory appraisal in Nepal. Tropical Doctor

28, 162–165.

Maine D (1991) Safe motherhood programs: Options and Issues,

Columbia University, New York.

Mancey-Jones M & Brugha RF (1997) Using perinatal audit to

promote change: a review. Health Policy and Planning 12, 183–192.

Mbaruku G & Bergström S (1995) Reducing maternal mortality in

Kigoma, Tanzania. Health Policy and Planning 10, 71–78.

McDonagh M (1996) Is antenatal care effective in reducing maternal

morbidity and mortality? Health Policy and Planning 11, 1–15.

MoH (1996) Safe Motherhood Policy of His Majesty’s Government of

Nepal. Journal of the Nepal Medical Association 34, 162–164.

Pradhan A, Aryal RH, Regmi G, Ban B & Govindasamy P (1996)

Nepal Family Health Survey. FHD/MoH Nepal, Kathmandu.

Ronsmans C, Achadi E, Sutratikto G, Zazri A & McDermott J (1999)

Use of hospital data for Safe Motherhood programmes in South

Kalimantan, Indonesia. Tropical Medicine and International

Health 4, 514–521.

Rooks JP, Weatherby NL, Ernst EKM et al. (1989) Outcomes of care

in birth centers: The national birth center study. New England

Journal of Medicine 321, 1804–1811.

St Leger AS, Schnieden H & Walsworth-Bell JP (1992) Evaluating

Health Services’ Effectiveness. Open University Press, Milton Keynes.

Thaddeus SA & Maine D (1994) Too far to walk: Maternal Mortality

in Context. Social Science and Medicine 38, 1091–1110.

UNICEF & MoH (1996) National Maternity Care Guidelines, Nepal.

FHD/MoH Nepal & UNICEF, Kathmandu.

Villar J, Bakketeig LS, Donner A et al. (1998) The WHO Antenatal

Care Randomised Controlled Trial: rationale and study design.

Paediatric and Perinatal Epidemiology 12, 27–58.

Villar J & Bergsjo P (1997) Scientific basis for the content of routine

antenatal care. Acta Obstetricia et Gynecologica Scandinavica 76,

1–14.

Voorhoeve AM, Kars C & van Ginneken JK (1984) Modern and

traditional antenatal and delivery care. In: Maternal and Child

Health in Rural Kenya – an Epidemiological Study, 2nd edn (eds

JK van Ginneken & AS Muller) Croom-Helm and African Medical

Research Foundation, London and Nairobi, pp. 309–322.

WHO (1991a) Essential Elements of Obstetric Care at First Referral

Level. Macmillan Press Ltd, London.

WHO (1991b) Hypertensive Disorders of Pregnancy: Report of a

WHO/MCH interregional collaborative study. WHO, Geneva.

WHO (1996a) Perinatal Mortality. WHO, Geneva.

WHO (1996b) Mother-Baby-Package: Implementing Safe

Motherhood in Countries. Maternal Health and Safe Motherhood

Programme, WHO, Geneva.

WHO (1996c) Care in Normal Birth: a Practical Guide. Maternal and

Newborn Health / Safe Motherhood Unit, WHO, Geneva.

WHO (1996d) Report of the Technical Working Group on Antenatal

Care. WHO, Geneva.

WHO (1997) Coverage of Maternity Care: a Listing of Available

Information, 4th edn. WHO, Geneva.

WHO & UNICEF (1996) Revised 1990 Estimates of Maternal

Mortality: a New Approach by WHO and UNICEF. WHO,

Geneva.

© 2000 Blackwell Science Ltd 665