Pi

RMa

b

Sc

d

e

a

ARRAA

KIIRP

1

srStr

f

RCT

0d

Drug and Alcohol Dependence 100 (2009) 9–16

Contents lists available at ScienceDirect

Drug and Alcohol Dependence

journa l homepage: www.e lsev ier .com/ locate /drugalcdep

revalences and correlates of non-viral injecting-related injuries and diseasesn a convenience sample of Australian injecting drug users�

. Dwyera, L. Toppb, L. Maherb, R. Powerc, M. Hellarda, N. Walshd,. Jaunceyb, A. Conroye, J. Lewisa, C. Aitkena,∗

Centre for Epidemiology and Population Health Research, Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne 3001, AustraliaViral Hepatitis Epidemiology and Prevention Program, National Centre in HIV Epidemiology and Clinical Research,t. Vincent’s Medical Centre, Darlinghurst, NSW 2010, AustraliaCentre for Harm Reduction, Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne 3001, AustraliaDepartment of Epidemiology and Preventive Medicine, Monash University, Clayton, Victoria 3800, AustraliaQueensland Needle and Syringe Program, Queensland Health, Brisbane, Queensland 4000, Australia

r t i c l e i n f o

rticle history:eceived 17 March 2008eceived in revised form 25 August 2008ccepted 25 August 2008vailable online 14 November 2008

eywords:njecting drug usenjecting-related injuries and diseasesisk factorsrotective factors

a b s t r a c t

Background: The prevalences and correlates of non-viral injecting-related injuries and diseases (IRIDs) inAustralian injecting drug users (IDUs) remain unknown.Methods: A cross-sectional survey of IDUs was conducted in six sites across Australia’s eastern states toinvestigate IRID experience among Australian IDU. Correlates of IRIDs were explored using logistic andnegative binomial regression analyses.Results: 393 IDUs were recruited. Lifetime experience of non-serious IRIDs was common (e.g., ‘dirty hit’68%); potentially serious and serious IRIDs were less commonly experienced (e.g., abscess 16%; gangrene<1%). Factors independently associated with potentially serious or serious IRIDs in the previous 12 monthswere: injecting in sites other than arms (Adjusted Odds Ratio 3.0, 95% confidence interval 1.7–5.4), inject-ing non-powder drug forms (5.0, 2.2–11.2), unstable accommodation (2.0, 1.1–3.5), being aged 25 years

or older (4.3, 1.7–10.6) and not always washing hands before injection (9.3, 2.1–41.8). Factors indepen-dently associated with multiple IRIDs in the preceding 12 months were using three or more injecting sites(Adjusted Incidence Rate Ratio 1.5, 95% CI 1.1–2.0), injecting in sites other than arms (1.7, 1.3–2.2), usingnon-powder drug forms (1.9, 1.4–2.5), injecting daily or more often (1.7, 1.3–2.2), current pharmacotherapyexperience (1.5, 1.1–1.9), and not always washing hands before injecting (1.9, 1.2–2.9).Discussion: Some IRIDs are widespread among Australian IDUs. Observed associations, particularly the

ashin

aihw

protective effect of handw

. Introduction

The process of injecting can give rise to both vascular andoft-tissue injuries and expose injecting drug users (IDUs) to a

ange of organisms that can cause localised or systemic infections.ome injecting-related injuries and diseases (IRIDs), such as infec-ive endocarditis and deep vein thrombosis (DVT), are serious andequire urgent medical attention. Others (e.g., abscesses, cellulitis),

� Supplementary information for this article is available. Please see Appendix Aor more information.∗ Corresponding author at: Centre for Epidemiology and Population Healthesearch, Macfarlane Burnet Institute for Medical Research and Public Health, 85ommercial Rd, Melbourne 3004/GPO Box 2284, Melbourne 3001, Australia.el.: +61 3 9282 2111; fax: +61 3 9282 2138.

E-mail address: aitken@burnet.edu.au (C. Aitken).

G1

l2sDKmagca

376-8716/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.drugalcdep.2008.08.016

g, have useful public health implications.© 2008 Elsevier Ireland Ltd. All rights reserved.

lthough requiring less urgent care, are potentially life-threateningf neglected. Others still are minor but common conditions (e.g.,ives), and may contribute significantly to reduced health andellbeing among injecting drug users (Ciccarone et al., 2001; deliudice, 2004; Pieper and Templin, 2001; Woodburn and Murie,996).

The literature on non-viral IRIDs is relatively limited, confinedargely to case-reports (Aboltins and Daffy, 2005; Cassoux et al.,002; Crossley, 2003; Feeney and Fairweather, 2003) and outbreaktudies from hospital presentations (Brett et al., 2005; Centres forisease Control and Prevention Update, 2001; Finn et al., 2003;imura et al., 2004). Prevalence and incidence studies are uncom-

on although notable exceptions exist. In Glasgow, Morrison et

l. (1997) surveyed current IDUs attending Needle Syringe Pro-rams (NSPs), who also underwent medical examination by alinical team. Clinical assessment indicated a prevalence of 21%ctive abscess, 49% thrombosis, 84% prominent scarring/bruising

1 ohol D

ainsmrIaIl1aI1a

i‘gts

sp(Fweh(

actcetisaqiebl(

hIodpi

2

Nltlcaa

twssrmo

awitcpIpeatC

lbucncdipvesdFi

munwd‘tutfbd

2

amiilapsed/blocked’ veins – suggestive of thrombosis); (ii) potentiallyserious, i.e., requiring treatment for resolution and capable of esca-lation (abscess, cellulitis, thrombophlebitis, oedema, puffy hands

0 R. Dwyer et al. / Drug and Alc

nd 76% other injecting-related problems, including oedema, fasci-tis, arterial damage, limited venous access and DVT. There wereo significant differences between clinical assessment and IDUelf-reports of current problems. In San Francisco, among a com-unity sample of 169 IDU, Binswanger et al. (2000) found that 32%

eported abscesses, cellulitis or both. In their cohort study of 1585DU in Vancouver, Lloyd-Smith et al., 2005 found 21% self-reportedn abscess in the preceding 6 months. Prevalence and incidenceRID studies most commonly investigate abscess and cellulitis, fol-owed by endocarditis. A study in Amsterdam between 1986 and994 found an abscess incidence rate of 33/100 person years andn endocarditis incidence rate of 1.3/100 person years among 758DUs (Spijkerman et al., 1996) while Tsao et al. (2002) identified86 cases of endocarditis (2.4% of 7795 hospital visits) among IDUttending a San Francisco hospital between 1994 and 1999.

Associations have been documented between IRIDs and specificnjecting practices such as subcutaneous injection, referred to asskin-popping’ (Binswanger et al., 2000; Passaro et al., 1998) androin injection (Rhodes et al., 2006a,b). Injecting anywhere otherhan arms is considered high risk, increasing the likelihood of IRIDuch as thrombosis and DVT (Derricot et al., 2003).

Associations have also been documented between IRIDs andpecific drug types and forms. In the UK and Australia, for exam-le, the injection of temazepam capsules has been linked with DVTAitken and Higgs, 2002; Woodburn and Murie, 1996). In Australia,rance and India, injecting-related injuries have been associatedith the injection of sublingual buprenorphine tablets (Jenkinson

t al., 2005). Finally, the black-tar heroin available in San Franciscoas been associated with increased incidence of wound botulismPassaro et al., 1998).

In Australia, the Illicit Drug Reporting System (IDRS) collectsnnual self-report information from IDUs on five injecting-relatedomplications (difficulty injecting, scarring/bruising, ‘dirty hits’,hrombosis and infection/abscess) experienced in the month pre-eding the survey (O’Brien et al., 2007). Among 914 IDU, 7% reportedxperiencing infection/abscess in the past month, 6% reportedhrombosis, 45% scarring/bruising and 43% reported difficulty innjecting. In 2006, the annual Australian NSP Survey, which mea-ures blood-borne virus (BBV) prevalence and risk behavioursmong clients of Australian NSPs, included additional self-reportuestions concerning experience of a small series of IRID andnjecting-related problems. Of 1961 IDUs, 43% reported havingxperienced problems finding a vein and 40% reported scarring andruising around their injection sites. Self-reported lifetime preva-

ence of abscess was 27%, thrombosis 12% and endocarditis (4%)Topp et al., 2008).

Although these studies provide some relevant data no compre-ensive assessment of IRIDs has been undertaken among Australian

DU, and the factors that increase the likelihood and severityf IRIDs have not been systematically investigated. This paperescribes findings from an exploratory study investigating therevalence and associations of a range of IRIDs among IDUs located

n metropolitan and regional settings across eastern Australia.

. Methods

A cross-sectional survey of current IDUs was conducted duringovember–December 2006 in three Australian states: Queens-

and, New South Wales and Victoria. Participants were recruited

hrough NSPs and street and social networks. One metropolitanocation (Brisbane) and one regional location (Toowoomba) werehosen in Queensland, one metropolitan location (outer Sydney)nd one regional location (northern rivers—specifically Lismorend Nimbin) were chosen in New South Wales and one metropoli-

s

t

ependence 100 (2009) 9–16

an location (Melbourne) and one regional location (Warrnambool)ere chosen in Victoria. Regional sites were urban centres out-

ide capital cities (defined as metropolitan sites) which wereelected on the basis of availability of NSPs that could facilitateecruitment. To be eligible, participants must have injected at leastonthly in the preceding 6 months, and have been aged 18 years or

lder.Interviewer teams in each state were comprised of a mix of peers

nd experienced research assistants. Interviewers were providedith an overview of IRIDs and their symptoms and received train-

ng in administration of the survey but they were not medicallyrained and were not required to diagnose IRIDs. Interviewers dis-ussed confidentiality and anonymity issues with participants, whorovided written informed consent prior to survey administration.

nterviews took an average of 45 min to complete; and partici-ants were reimbursed AUD$ 30 for their time and out-of-pocketxpenses, in accordance with accepted procedures for researchmong IDU (Fry and Hall, 2003; Fry et al., 2005). Ethics approval forhe study was obtained from six separate Human Research Ethicsommittees across the three states.

The survey instrument was developed from a comprehensiveiterature review on IRIDs, with clinical advice on IRIDs providedy specialist Medical Practitioners with expertise in injecting drugser health who were members of the study team. The domainsovered by the questionnaire were: injection practices and tech-iques; use of sterile injecting equipment; socio-environmentalontexts of injecting; drug use data (e.g., age first injected; allrug types and forms ever injected, injected last 12 months and

njected last 4 weeks); pharmacotherapy (methadone, buprenor-hine, naltrexone) experience; injecting sites on the body; recentenepuncture problems; experiences of IRIDs (detail below); gen-ral health (including the SF-8TM (QualityMetric, Lincoln, RI, USA));elf-reported BBV status; and demographic information. Up to 40etailed questions were asked relating to each separate domain.ull details of the questionnaire are available from the correspond-ng author on request.

The questionnaire assessed experience of 23 specific IRIDs. Theedical terms for these IRIDs (e.g., thrombophlebitis) were not

sed as these were considered unlikely to be accurately recog-ised or remembered by respondents. Instead, the questionnaireas designed such that, for each IRID, interviewers asked respon-ents about their experience of a specific set of symptoms (e.g.,

tender/painful, swollen, red tracks along veins’). Respondents hado have experienced every symptom associated with each partic-lar IRID to be classified by interviewers as having experiencedhat IRID. This allowed for identification of IRIDs as well as dif-erentiation between similar IRIDs.1 As IRIDs were not diagnosedy medical practitioners, the findings are indicative, rather thaniagnostic, of any given injecting-related complication.

.1. Data analysis

IRIDs were categorised as follows: (i) non-serious, i.e., commonnd generally self-resolving, and/or unlikely to require treat-ent (transient redness; transient swelling; hives; ‘dirty hit’ –

ndicative of transient bacteraemia; accidentally hitting, but notnjecting into, an artery; numbness or pins and needles; and ‘col-

yndrome, injecting sinus); and (iii) serious, i.e., potentially or

1 Further detail on IRIDs and associated symptoms can be viewed by accessinghe online version of this paper at http://dx.doi.org.

ohol D

iD

meeofrtIarcoitstcwuadiseiiatdlvtummaeI(npcomHo

tpoasbsfwcsgtIm

w(Itaaisutfc

3

3wumrbT

(tscp

o8ph4ibuiiRfI(

3

oorsUtliFmw

R. Dwyer et al. / Drug and Alc

mminently disabling and/or life-threatening (systemic infections,VT, gangrene, amputation and venous ulcer).

Two binary categorical outcomes were generated; past 12-onth experience of non-serious IRIDs (yes/no) and past 12-month

xperience of potentially serious/serious IRIDs (these two IRID cat-gories collapsed due to low prevalence of serious IRIDs). Theseutcomes were examined separately because of the major dif-erences in the severity (and associated medical interventionequired) between the outcomes. Univariate analysis of poten-ial correlates of non-serious IRIDs and potentially serious/seriousRIDs, was undertaken with associated (p ≤ 0.1) independent vari-bles included in multivariate logistic regression analysis. Variableselating to ‘injecting practices and technique’ in the 4 weeks pre-eding the survey were: number of injecting sites used; typesf injecting sites used; applying pressure to injecting site afternjection; jacking-back (pulling back on the plunger during injec-ion); using a tourniquet; types of needle and syringes used;elf-administering injections; frequency of injecting; reusing fil-ers; saving filters for drug residue; storing pre-prepared syringes;leaning injection site with an alcohol swab prior to injection;ashing hands before mixing and injecting drugs; backloading;sing acidifiers (lemon juice, vinegar) to prepare drug solutionnd using heat to prepare drug solution. Variables relating to theomain ‘use of sterile equipment’ (last 4 weeks) were using a ster-

le needle and syringe for each injection and using ampoules ofterile water (provided by NSPs) to prepare drug solution. Socio-nvironmental context variables were number of people usuallynjected within the last 4 weeks and injecting environments usedn the last 4 weeks (public places – e.g., public toilets, streets andlleyways – versus private places – e.g., private homes). Other fac-ors entered into the models were: pharmacotherapy experience;rug forms (powders versus powders and/or other forms — tablets,

iquids, gel-capsules) injected last 12 months (this is an aggregateariable coded from the detailed drug use data); number of drugypes injected in the last 12 months; duration of injecting drugse; mental health diagnosis ever; dental problems in the last 12onths; imprisonment in the last 12 months; stability of accom-odation in the last 12 months (unstable accommodation defined

s homelessness or residing in four or more different places);ducation; employment; country of birth (Australia or overseas);ndigenous Australian identity; age; gender; and recruitment sitemetropolitan or regional). A stepwise backwards elimination tech-ique was used. Variables that were not significant and were nototentially confounding (i.e., a variable was considered potentiallyonfounding if its elimination substantially altered the coefficientsf any remaining variables) were systematically removed from theodel until a subset of wholly significant correlates remained. Theosmer–Lemeshow test was used to assess overall goodness-of-fitf the final model.

The total number of IRID types (including non-serious, poten-ially serious and serious) experienced by participants in thereceding 12 months were summed to generate a count measuref the range and severity of IRID experience (hereafter referred tos multiple and serious IRIDs). As a simplified way of reflectingeverity in this measure, non-serious IRIDs (n = 8) were weightedy a factor of 1, potentially serious IRIDs (n = 8) by a factor of 5, anderious IRIDs (n = 10) by a factor of 12. These weights were derivedrom the IRID prevalences observed among the sample. This crudeeighting procedure ensured that a single episode of a more clini-

ally significant IRID such as endocarditis would generate a higher

core than experience of, for example, eight separate self-resolving,enerally non-clinically significant IRIDs (e.g., transient redness,ransient swelling, hives). The distribution of multiple and seriousRIDs scores was skewed; of a possible score between 0 and 168, the

ean multiple and serious IRIDs score for the preceding 12 months

3

fw

ependence 100 (2009) 9–16 11

as 3.6 (S.D. = 5.5), the median score was 2 and the range was 0–51IQR 1–4); 75% scored less than 5, indicating experience of either noRID types or only non-serious ones. Due to over-dispersion, nega-ive binomial regression was used to assess correlates of multiplend serious IRIDs experienced in the preceding 12 months, withll correlates significant at p ≤ 0.1 on univariate analysis includedn the multivariate analysis. The incidence rate ratio (IRR) mea-ure of association produced by negative binomial regression wassed to represent the change in the expected count of the mul-iple and serious IRIDs measure when an individual was positiveor a particular correlate compared to being negative on thatorrelate.

. Results

A total of 393 IDUs were interviewed; their average age was3 years (S.D. 8.5, range 18–65 years, median 32 years) and 73%ere male. The majority (74%) were unemployed, and 33% hadnstable accommodation (either homeless or living in four orore different places) in the previous 12 months. Two-third of the

espondents were born in Australia. Of the overseas-born, 58% wereorn in Vietnam, 8% in New Zealand and 6% in the United Kingdom.hirty participants (8%) identified as Indigenous Australians.

Sixty percent of respondents rated their health in positive termseither good, very good or excellent) on the general health item ofhe SF-8TM, whereas 16% rated it as very good or excellent. Forty-ix percent reported a diagnosed mental health problem (mostommonly depression, 22%). Self-reported hepatitis C exposurerevalence was 54%; self-reported HIV prevalence was below 1%.

Nearly 60% of interviewees had been injecting drugs for 10 yearsr more, and 14% for 5 years or less. In the preceding 12 months,0% had injected heroin, 56% amphetamine, 37% morphine sul-hate, and 29% buprenorphine. In the preceding 4 weeks, nearlyalf reported injecting daily or more frequently in the precedingweeks, few (17%) always used sterile water to mix drugs for inject-

ng but most (79%) made some attempt to wash or clean their handsefore preparing and injecting drugs. Seventy percent reportedsing a sterile needle and syringe for every injection in the preced-

ng 4 weeks. Nearly a quarter of interviewees had used a high-risknjecting site (anywhere other than arms) in the previous 4 weeks.espondents reported ever using a median of six (IQR 3–8.5) dif-

erent injecting sites (includes left and right sides) on their body.n the preceding 4 weeks, respondents had used a median of twoIQR 1–3) different injecting sites.

.1. IRID prevalences

Nearly all respondents (95%) had lifetime experience of at leastne non-serious IRID, with 79% of respondents reporting at leastne non-serious IRID in the preceding 12 months. Forty percent ofespondents reported lifetime experience of at least one potentiallyerious IRID, the most common of which was skin abscess (16%).nder one-quarter reported at least one potentially serious IRID in

he previous 12 months. Fourteen percent of respondents reportedifetime experience of at least one serious IRID, with 4% report-ng ever suffering septicaemia (the most common serious IRID).ive percent reported at least one serious IRID in the preceding 12onths. Participants also reported multiple episodes of IRIDs. Thisas particularly true for non-serious IRIDs (Table 1).

.2. Correlates of non-serious IRIDs

Multivariate logistic regression analysis (Table 2) showed thatactors associated with a non-serious IRID in the last 12 monthsere: injecting in three or more injecting sites in the last 4 weeks;

12 R. Dwyer et al. / Drug and Alcohol Dependence 100 (2009) 9–16

Table 1Prevalence of self-reported injecting-related injuries and diseases (lifetime and past 12 months).

Lifetime (n = 393) Last 12 months (n = 393) More than 1 episode last 12 months

n % n % na %

Non-serious IRIDsTransient redness near inject site 166 42 107 28 80/103 78Transient swelling near inject site 176 45 109 28 78/107 73Hives 220 56 151 38 118/149 79Transient bacteraemia—‘Dirty Hit’ 266 68 135 34 53/133 40Hit an artery when injecting 86 22 34 9 10/32 31Numbness/pins and needles 76 19 48 12 24/47 51Collapsed/blocked veins 188 48 102 26 25/94 27Other (burning in veins, skin problem, needle broke off in vein) 5 1 1 <1 0 –

Potentially serious IRIDsSkin abscess 65 16 27 7 9/27 33Internal abscess 12 3 4 1 0 –Cellulitis 56 14 27 7 20/27 74Thrombophlebitis 56 14 25 6 17/24 71Pitting oedema 29 7 17 4 6/16 38Puffy hands syndrome 28 7 15 4 4/14 29Injecting sinus 19 5 11 3 1/11 9Other (allergic reaction) 4 1 3 <1 2/2 100

Serious IRIDsEndocarditis 12 3 5 1 0 –Septicaemia 17 4 5 1 1/4 25Septic arthritis 4 1 1 <1 0 –Osteomyelitis 2 <1 1 <1 0 –Serious infection (unspecified) 9 2 2 <1 0 –Deep vein thrombosis 13 3 5 1 1/5 20Gangrene 3 <1 1 <1 0 –Amputation 3 <1 1 <1 0 –Venous ulcer 6 1 3 <1 0 –

des inl

ptesl4g

3

TI

P

N

N

P

E

S

U

Other (myocarditis, tumour, renal failure) 3

a Not all respondents experiencing an IRID provided data on the number of episoast 12 months.

harmacotherapy experience; injecting six or more different drugypes in the last 12 months and completing secondary or tertiary

ducation. Factors that were significantly protective against non-erious IRIDs were: always using sterile water to mix drugs in theast 4 weeks; and always self-administering injections in the last

weeks. The Hosmer–Lemeshow test indicated the model was aood fit to the data (p = 0.553).

ldim

able 2ndependenta correlates of any experience of non-serious (NS) IRIDs in the past 12 month

redictor No NS IRID last12 months, n (%)

NS IRID last12 months, n (%)

umber of injecting site areas last 4 weeksOne to two 51 (21) 196 (79)Three or more 13 (11) 107 (89)

umber of drug types used last 12 monthsOne to five 59 (22) 212 (78)Six or more 6 (6) 94 (94)

harmacotherapy experienceNever 27 (32) 57 (68)Past 13 (11) 103 (89)Current 24 (14) 147 (86)

ducationNot complete secondary 20 (22) 72 (78)Secondary/tertiary 45 (16) 234 (84.)

elf-administering injections last 4 weeksNot always 7 (11) 57 (89)Always 58 (19) 249 (81)

sed sterile water last 4 weeksNot always 42 (15) 244 (85)Always 16 (28) 42 (72)

a Independent of other variables in the model.

<1 1 <1 0 –

the last 12 months so on may be lower than for those experiencing the IRID in the

.3. Correlates of potentially serious or serious IRIDs

Factors associated with potentially serious or serious IRIDs in theast 12 months, as shown in Table 3, were: injecting non-powderrug forms in the last 12 months; injecting in sites other than arms

n the last 4 weeks; having unstable accommodation in the last 12onths; aged 25 years or older; and not always washing/cleaning

s.

Unadjusted OR(UOR) (95% C.I.)

p Adjusted OR (AOR)(95% C.I.)

p

2.14 (1.11–4.11) 0.022 2.24 (1.02–4.90) 0.043

4.36 (1.82–10.45) 0.001 5.12 (1.72–15.28) 0.003

3.75 (1.80–7.84) <0.001 4.64 (1.98–10.90) <0.0012.90 (1.55–5.44) 0.001 5.61 (2.63–11.97) <0.001

1.44 (0.80–2.60) 0.221 2.25 (1.11–4.53) 0.024

0.053 (0.23–1.21) 0.133 0.30 (0.10–0.84) 0.022

0.45 (0.23–0.88) 0.019 0.46 (0.22–0.97) 0.042

R. Dwyer et al. / Drug and Alcohol Dependence 100 (2009) 9–16 13

Table 3Independent correlates of any experience of potentially serious (PS) or serious (S) IRIDs in the past 12 months.

Predictor No PS or S IRID last12 months, n (%)

Any PS or S IRID last12 months, n (%)

Unadjusted OR(UOR) (95% C.I.)

p Adjusted OR(AOR) (95% C.I.)

p

Injecting site areas last 4 weeksArms only 183 (81) 42 (19)Arms and/or other sites 84 (64) 48 (36) 2.49 (1.53–4.06) <0.001 3.05 (1.7–5.44) <0.001

Drug forms used last 12 monthsPowders only 94 (93) 8 (7)Powders +/or other forms 173 (68) 83 (32) 5.64 (2.62–12.15) <0.001 4.96 (2.19–11.24) <0.001

Wash/clean hands before mixing and injecting drugs last 4 weeksAlways 43 (93) 3 (7)Not always 225 (72) 88 (28) 5.60 (1.69–18.54) 0.008 9.35 (2.09–41.84) 0.003

Accommodation last 12 monthsStable 187 (79) 51 (21)Unstable 81 (67) 40 (33) 1.81 (1.11–2.95) 0.021 1.98 (1.12–3.49) 0.018

Age groupa

18–24 years 49 (84) 9 (16)25 years or more 218 (73) 82 (27) 2.05 (0.96–4.36) 0.058 4.29 (1.74–10.60) 0.002

Used acid to mix drugs last 4 weeksNo 221 (73) 83 (27)Yes 44 (90) 5 (10) 0.30 (0.12–0.79) 0.014 0.25 (0.09–0.71) 0.009

Country of birth

ategor

hppH

3

tmTtsawfatt

4

rbwIap4osns

4

s

bhoefpat2rcctpifpbfsip

4

oifiraf2

Australia 173 (72) 66 (28)Overseas 95 (79) 25 (21)

a Median age in 18–24 category was 22.5 years, median age in 25 years or more c

ands before mixing and injecting in the last 4 weeks. Significantlyrotective factors were using an acid (i.e., lemon juice or vinegar) torepare drug solution in the last 4 weeks and being overseas-born.osmer–Lemshow test was non-significant (p = 0.924).

.4. Correlates of multiple and serious IRIDs

Factors associated with experience of multiple and serious IRIDypes in the preceding 12 months (Table 4) included using three or

ore injecting sites in the last 4 weeks (Adjusted IRR = 1.5, p = 0.003.he interpretation of the IRR is that injecting in ≥3 injecting sites inhe last 4 weeks increased the expected multiple and serious IRIDscore by a factor of 1.5, compared to injecting in <3 injecting sitesnd holding all other factors constant). Other significant correlatesere: injecting in sites other than arms; using non-powder drug

orms in the last 12 months; injecting once a day or more; and notlways washing hands before mixing and injecting drugs. Factorshat were significantly protective were completing secondary orertiary education and metropolitan residence.

. Discussion

Most IRIDs defined in this study as non-serious (e.g., transientedness, hives, ‘collapsed’/blocked veins) had been experiencedy large majorities of IDUs; the most common non-serious IRIDas the ‘dirty hit’ (indicative of transient bacteraemia) at 68%.

n contrast, the most common potentially serious IRID was skinbscess (16%), and less than a quarter of participants reported anyotentially serious IRID in the previous 12 months. Likewise, only% reported ever suffering septicaemia (the most common seri-us IRID), and 1% or fewer reported experiencing each particularerious IRID in the preceding 12 months. Thus, the prevalence ofon-serious IRIDs was higher than the prevalence of potentiallyerious IRIDs which, in turn, was higher than that of serious IRIDs.

.1. Non-serious IRIDs

Concurrent use of multiple injecting sites and polydrug use wereignificantly associated with having a non-serious IRID, as was

upiwb

0.69 (0.41–1.16) 0.209 0.44 (0.23–0.84) 0.012

y was 33 years.

oth current and past pharmacotherapy programme experienceistory. This was the case after adjusting for frequency, and length,f injecting drug use. The association between pharmacotherapyxperience and non-serious IRIDs is most likely due to a higherrequency of methadone and/or buprenorphine injection amongeople enrolled on pharmacotherapy programmes (Jenkinson etl., 2005). The injection of both these substances has been showno increase the risk of injecting-related complications (Darke et al.,002; Jenkinson et al., 2005). The viscosity of methadone increasesisk of venous trauma and thrombosis, while small particles fromrushed buprenorphine tablets may form microemboli which canause venous obstruction or impaired blood supply if transportedhrough capillaries. Additionally, buprenorphine diverted fromharmacotherapy programmes in Australia is usually ‘smuggled’

n the recipient’s mouth, thereby increasing risks of bacterial orungal infection (Cassoux et al., 2002). The use of sterile water torepare drugs, and self-administration of injections, were found toe protective for non-serious IRIDs; both are logical mechanismsor reducing risk of bacterial exposure, and it seems plausible thatelf-injection generally produces less vascular damage than beingnjected by others as people will be more sensitive to their ownain signals.

.2. Potentially serious and serious IRIDs

The injection of non-powder drug forms (e.g., tablets, liquidsr capsules) and injecting in high-risk injecting sites were signif-cantly associated with potentially serious or serious IRIDs. Thesendings are consistent with other research identifying increasedisk of harm from the injection of non-powder drug forms (Aitkennd Higgs, 2002; Feeney and Gibbs, 2002; Robinson et al., 2000) androm injection in high-risk sites (Derricot et al., 2003; Rhodes et al.,006a,b). The observed association between more serious IRIDs and

nstable accommodation may be reflective of a greater likelihood ofublic injecting and the associated difficulties of maintaining ideal

njecting practices (e.g., comfort, safety and hygiene) among thoseith inadequate housing (Rhodes et al., 2006a,b). Older age may

e a risk factor due to the general deterioration of tissue, cell and

14 R. Dwyer et al. / Drug and Alcohol Dependence 100 (2009) 9–16

Table 4Independent correlates of experiencing multiple IRID types in the past 12 months.

Predictor n Mean (S.D.) Median (range) Unadjusted IRR (95% C.I.) p Adjusted IRR (95% C.I.) p

IRID count 288 3.6 (5.5) 2 (0–51)

Number of injecting site areas last 4 weeksOne to two sites 195 2.8 ± 3.7 1 (0–23)Three or more sites 89 5.4 ± 7.8 3 (0–51) 1.92 (1.44–2.57) <0.001 1.51 (1.15–1.98) 0.003

Injecting site areas last 4 weeksArms only 191 2.7 ± 3.3 1 (0–18)Arms and/or other sites 95 5.5 ± 7.9 3 (0–51) 2.05 (1.55–2.72) <0.001 1.66 (1.26–2.18) <0.001

Drug forms used last 12 monthsPowders only 90 2.0 ± 4.3 1 (0–39)Powders +/or others 198 4.3 ± 5.8 2 (0–51) 2.16 (1.60–2.93) <0.001 1.89 (1.41–2.53) <0.001

Wash/clean hands before mixing and injecting drugs last 4 weeksAlways 40 1.8 ± 1.7 1 (0–7)Not always 248 3.9 ± 5.8 2 (0–51) 2.11 (1.38–3.24) 0.001 1.90 (1.24–2.90) <0.001

Injecting frequency last 4 weeksLess than once per day 155 3.0 ± 4.0 1 (0–23)Once per day or more 133 4.4 ± 6.7 2 (0–51) 1.46 (1.11–1.93) 0.008 1.35 (1.04–1.76) 0.024

Pharmacotherapy experienceNever or past 155 2.8 ± 3.3 2 (0–23)Current 133 4.7 ± 7.2 2 (0–51) 1.71 (1.30–2.25) <0.001 1.47 (1.13–1.90) 0.003

EducationNot secondary 71 5.2 ± 8.7 2 (0–51)Secondary/tertiary 216 3.1 ± 3.8 2 (0–23) 0.60 (0.44–0.82) 0.001 0.69 (0.52–0.92) 0.012

S

0.

i

ohopttfitpbIdld(faIt

bh2eitaofioib

nVa

4

iftoaimatp

4

eIstRaeeT

iteRegional 75 4.3 ± 6.8 2 (0–51)Metropolitan 213 3.4 ± 4.9 2 (0–39)

mmune system function associated with ageing (Pawelec, 2006).Several factors were shown to be associated with reduced risk

f serious or potentially serious IRIDs — notably washing/cleaningands before mixing and injecting drugs, a fundamental tenetf hygienic practice. While NSPs and other IDU services alreadyromote handwashing as a way of reducing blood-borne virusransmission, our data provide sound evidence that this is effec-ive in reducing occurrence of non-viral IRIDs. This is an importantnding because – while handwashing is universally promoted inhe literature and in IDU health promotion – we are unaware ofrior research that demonstrated a strong protective associationetween handwashing and occurrence of non-viral IRID among

DUs. The significant protective effect of using an acid to preparerug solution is most likely due to the anti-bacterial properties of

emon-juice and/or vinegar (Brett et al., 2005). Although preparingrugs with lemon juice has been associated with fungal infectionsNewton-John et al., 1984) and acids provide an ideal environmentor spore-forming bacteria, such as the Clostridium species (Brett etl., 2005), Australia is yet to observe a Clostridium outbreak amongDU such as occurred in the UK in early 2000, and no fungal infec-ions were reported in this study.

Previous research has demonstrated that among some overseas-orn Australian communities, injecting-related harm is markedlyigher than among Australian-born populations (Maher et al.,001; Reid et al., 2002). In light of this, the significant protectiveffect for overseas-born observed in the current study is counter-ntuitive. The lack of a significant association may be explained byhe adjustment made for more harmful injecting practices (suchs injection in high-risk injecting sites) which may occur in someverseas-born populations. However, we are unable to account

or the observed protective effect of this factor. Interpretations hampered by the complexity of the measure which includedverseas-born from English-speaking backgrounds (culturally sim-lar to the majority Australian population — i.e., of Anglo-Saxonackgrounds), as well as overseas-born from a diverse range of

rtitA

78 (0.57–1.07) 0.132 0.63 (0.46–0.85) 0.003

on-English speaking and culturally dissimilar backgrounds (e.g.,ietnam, other Asian countries, South America, Southern Europend so on).

.3. Multiple and serious IRIDs

As in previous models, concurrent use of multiple injecting sites,njection in high-risk injecting sites and injecting non-powder drugorms were all significantly associated with the experience of mul-iple and serious IRIDs. Additionally, frequent injectors (once a dayr more) were more likely to experience multiple and serious IRIDs,s were those who did not always practice handwashing prior tonjection. Metropolitan residence was protective, arguably because

etropolitan areas provide superior access to injecting equipmentnd IDU services (Day et al., 2005). Secondary or higher educa-ion, although identified as a risk factor for non-serious IRIDs, wasrotective against experiencing multiple and serious IRIDs.

.4. Strengths and limitations

This unique study produced a comprehensive snapshot of thepidemiology of non-viral IRIDs among Australian IDU. Nearly 400DUs provided data on their experience of IRIDs. The sample wasimilar to other convenience community-recruited samples of Aus-ralian IDU (National Centre in HIV Epidemiology and Clinicalesearch, 2007; O’Brien et al., 2007) with respect to demographicnd drug use characteristics. Nevertheless, as our data were gen-rated from a population subgroup for which no sampling framexists, there are inevitable limitations on their representativeness.his limitation is exacerbated by our unavoidable reliance on self-

eport of IRIDs by participants rather than diagnosis by medicallyrained personnel. Clinical assessment and diagnosis would havemproved confidence in the prevalence estimates generated byhe study. It is also possible that studying IDUs in three easternustralian states does not permit generalisation to IDUs in other

ohol D

jpnsbIielt

iiCncSt2ssdeWom

aaeCiddmotghwp

peicaocar(iip

iaSapwohp

wpsocdd

R

ms

C

dafissaaLtastatcgibctrficibtaadt

C

A

Iir

R. Dwyer et al. / Drug and Alc

urisdictions, due to differences such as drugs of choice, injectingractices and models of NSP. For the same reasons, our results mayot apply to IDUs in the UK, USA or elsewhere. The cross-sectionaltudy design means that causal inferences regarding associationsetween injecting and drug use practices and the occurrence of

RIDs cannot be made. Furthermore, there are inherent difficultiesn linking practices engaged in during the last 4 weeks with IRIDsxperienced anytime in the preceding 12 months. Due to all theseimitations the findings remain indicative rather than definitive ofhe prevalences and associations of IRIDs among IDUs.

Consistent with international literature, our study found signif-cant associations between the occurrence of IRIDs and high-risknjecting practices, as well as injection of non-powder drug forms.omparison of prevalence rates found in the Australian IDRS isot possible due to different time-frames — 4 weeks in the IDRSompared with either lifetime or 12 months in the present study.ome IRID prevalences in the current study were comparable tohose reported in the 2006 Australian NSP survey (Topp et al.,008) (e.g., 3% and 4% respectively for endocarditis), others wereubstantially higher (e.g., 12% and 48% respectively for thrombo-is/blocked veins). These differences may be due to differences inata collection methods; the NSP survey was self-completed andmployed medical terms, augmented by a lay-term for some IRIDs.e consider our self-report findings were strengthened by the use

f symptoms to identify IRIDs, rather than relying on medical ter-inology with which many IDUs would be unfamiliar.Additionally, this study explored a comprehensive range of IRIDs

nd elicited detailed data on injection techniques and practices,llowing greater specificity in identifying associations of IRIDs. Forxample, in contrast to the recent Australian NSP Study (Nationalentre in HIV Epidemiology and Clinical Research, 2007) and other

nvestigations (Darke et al., 2001; Spijkerman et al., 1996) weid not observe a significant association between IRIDs and gen-er, whereas the aforementioned studies found that women wereore likely to experience these complications. It may be that in

ther studies gender is actually a marker for riskier injecting prac-ices (e.g., non self-administration of injections) produced throughendered social roles (MacRae and Aalto, 2000). Our study compre-ensively measured injecting technique and practices, and genderas insignificant once adjustment was made for riskier injectingractices.

The apparently lower prevalences of IRIDs in Australia com-ared with international reports is probably reflective of differingnvironmental and structural conditions — better access to inject-ng equipment and education in Australia, no black-tar heroin, norack-cocaine injecting, low, if not no, prevalence of skin-popping,nd greater access to welfare and medical systems than in somether countries, such as the USA. Nevertheless, patterns of drug usehange rapidly; increased injection of diverted oral buprenorphinend the recent surge in temazepam capsule injection in Australiaesulted in increases in serious IRIDs among some IDU populationsAitken and Higgs, 2002; Jenkinson et al., 2005). Furthermore, IDUsn the current study reported local and systemic infections, indicat-ng that bacterial exposure and significant bacterial outbreaks areossible.

To reduce the occurrence of IRIDs in Australia and elsewhere,t is imperative that IDUs are given the means to inject as safelys possible, in terms of equipment, knowledge and environments.upervised injecting facilities are useful in this regard (Stoltz etl., 2007). The protective effects of handwashing, sterile water and

reparing drugs with an acid, and increased risk of IRIDs associatedith injecting non-powder drug forms, suggest that the occurrence

f IRIDs could be substantially reduced through reinforcing basicarm reduction messages and training around hygienic injectingractices and safer injecting/vein care, provision of portable hand-

A

i

ependence 100 (2009) 9–16 15

ashing materials (e.g., waterless hand cleaner), and increasedromotion of drug filtering. Near-universal experience of non-erious IRIDs, and a lifetime prevalence of potentially serious IRIDsf 40% implies the existence of a considerable public health burden;linicians, and other health workers, need to be alert to these con-itions in their IDU clients to prevent progression to more seriousisease.

ole of funding source

Funding for this study was provided by the Australian Govern-ent, Department of Health and Ageing, Australia. The funding

ource had no further involvement in the study.

ontributors

Robyn Dwyer designed the study, coordinated and oversaw ofata collection, analysed and interpretation of data, drafted therticle, revised it critically for important intellectual content andnal approval of the version to be published. Guarantor of thetudy. Campbell Aitken designed the study, coordinated and over-aw of data collection, interpretation of data, drafted the articlend revised it critically for important intellectual content and finalpproval of the version to be published. Guarantor of the study.ibby Topp designed the study, oversaw of data collection, revisedhe article critically for important intellectual content and finalpproval of the version to be published. Lisa Maher designed thetudy, oversaw of data collection, interpretation of data, revisedhe article critically for important intellectual content and finalpproval of the version to be published. Robert Power designedhe study, revised the article critically for important intellectualontent and final approval of the version to be published. Mar-aret Hellard designed the study, revised the article critically formportant intellectual content and final approval of the version toe published. Nick Walsh designed the study, revised the articleritically for important intellectual content and final approval ofhe version to be published. Marianne Jauncey designed the study,evised the article critically for important intellectual content andnal approval of the version to be published. Andrew Conroy con-epted and designed the study, revised the article critically formportant intellectual content and final approval of the version toe published. Jenny Lewis analysed and interpreted data, revisedhe article critically for important intellectual content and finalpproval of the version to be published. Robert Kemp conceptednd designed the study, oversaw of data collection. Belinda Lloydesigned the study, oversaw of data collection. All authors con-ributed to and have approved the final manuscript.

onflict of interest

None declared.

cknowledgements

The authors are grateful to the people who participated in theDU survey and contributed their experiences of injecting-relatednjuries and diseases. The authors also thank the anonymouseviewers for their thoughtful comments.

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at doi:10.1016/j.drugalcdep.2008.08.016.

1 ohol D

R

A

A

B

B

C

C

C

C

D

D

D

d

D

F

F

F

F

F

J

K

L

M

M

M

N

N

O

P

P

P

R

R

R

R

S

S

T

6 R. Dwyer et al. / Drug and Alc

eferences

boltins, C.A., Daffy, J.R., 2005. Letter to the editor: fungal endophthalmitis in intra-venous drug users injecting buprenorphine contaminated with oral Candidaspecies. Med. J. Aust. 182, 427.

itken, C., Higgs, P., 2002. Letter to the editor: severe vein damage caused bytemazepam injecting. Aust. N. Z. J. Public Health 26, 79.

inswanger, I.A., Kral, A.H., Bluthenthal, R.N., Rybold, D.J., Edlin, B.R., 2000. Highprevalence of abscesses and cellulitis among community-recruited injectiondrug users in San Francisco. Clin. Infect. Dis. 30, 579–581.

rett, M.M., Hood, J., Brazier, J.S., Duerden, B.I., Hahne, S.J.M., 2005. Soft tissue infec-tions caused by spore-forming bacteria in injecting drug users in the UnitedKingdom. Epidemiol. Infect. 133, 575–582.

assoux, N., Bodaghi, B., Lehoang, P., 2002. Presumed ocular candidiasis in drugmisusers after intravenous injection of oral high dose buprenorphine (Subutex).Br. J. Opthamol. 86, 940–941.

entres for Disease Control and Prevention Update, 2001. Clostridium novyi andunexplained illness among injecting drug users—Scotland, Ireland, and England,April–June 2000. Ann. Emerg. Med. 37, 107–108.

iccarone, D.H., Bamberger, J.D., Kral, A.H., Edlin, B.R., Hobart, C.J., Moon, A., Mur-phy, E.L., Bourgois, P., Harris, H.W., Young, D.M., 2001. Soft tissue infectionsamong injection drug users—San Francisco, California, 1996–2000. JAMA 285,2707–2709.

rossley, M., 2003. Temperate pyomyositis in an injecting drug misuser. A difficultdiagnosis in a difficult patient. Emerg. Med. J. 20, 299–300.

arke, S., Ross, J., Kaye, S., 2001. Physical injecting sites among injecting drug usersin Sydney, Australia. Drug Alcohol Depend. 62, 77–82.

arke, S., Topp, L., Ross, J., 2002. The injection of methadone and benzodiazepinesamong Sydney injecting drug users 1996–2000: 5-year monitoring of trendsfrom the Illicit Drug Reporting System. Drug Alcohol Rev. 21, 27–32.

ay, C. Day, C., Conroy, E., Lowe, J., Kimber, J., Page, J., Dolan, K., 2005. Drug useand harms among rural and metropolitan injecting drug users: Findings fromthe Rural Injectors Project. NDARC Technical Report No. 234. National Drugand Alcohol Research Centre, The University of New South Wales, Sydney,NSW.

el Giudice, P., 2004. Cutaneous complications of intravenous drug abuse. Br. J.Dermatol. 150, 1–10.

erricot, J., Preston, A., Hunt, N., 2003. The safer injecting briefing: aneasy to use comprehensive reference guide to promoting safer injecting.http://www.exchangesupplies.org. [Accessed on 4 November 2005].

eeney, G.F.X., Fairweather, P., 2003. Groin tissue necrosis requiring skin graftfollowing parenteral abuse of buprenorphine tablets. Drug Alcohol Rev. 22,359–361.

eeney, G.F.X., Gibbs, H.H., 2002. Digit loss following misuse of temazepam. Med. J.Aust. 176, 380.

inn, S.P., Leen, E., English, L., O’Briain, S., 2003. Autopsy findings in an outbreak ofsevere systemic illness in heroin users following injection site inflammation.Arch. Pathol. Lab. Med. 127, 1465–1470.

ry, C., Hall, W., 2003. Key issues in determining the ethics of research subject pay-ment: the special case of drug abuse epidemiology. Australasian Epidemiologist

10, 41–47.

ry, C.L., Ritter, A., Baldwin, S., Bowen, K., Gardiner, P., Holt, T., Jenkinson, R.A., John-ston, J., 2005. Paying research subjects: a study of current practices in Australia.J. Med. Ethics. 31, 542–547.

enkinson, R., Clark, N., Fry, C., Dobbin, M., 2005. Buprenorphine diversion and injec-tion in Melbourne, Australia: an emerging issue. Addiction 100, 197–205.

T

W

ependence 100 (2009) 9–16

imura, A., Higa, J., Levin, R., Simpson, G., Vargas, Y., Vugia, D., 2004. Outbreak ofnecrotizing fasciitis due to Clostridium sordelli among black-tar heroin users.Clin. Infect. Dis. 38, e87–91.

loyd-Smith, E., Kerr, T., Hogg, R.S., Li, K., Montaner, J.S.G., Wood, E., 2005. Prevalenceand correlates of abscesses among a cohort of injection drug users. Harm Reduc.J. 2, 24.

acRae, R., Aalto, E., 2000. Gendered power dynamics and HIV risk in drug-usingsexual relationships. AIDS Care 12, 505–515.

aher, L., Sargent, P., Higgs, P., Crofts, N., Kelsall, J., 2001. Risk behaviours of youngIndo-Chinese injecting drug users in Sydney and Melbourne. Aust. N. Z. J. PublicHealth 25, 50–54.

orrison, A., Elliot, L., Gruer, L., 1997. Injecting-related harm and treatment-seekingbehaviour among injecting drug users. Addiction 92, 1349–1352.

ational Centre in HIV Epidemiology and Clinical Research, 2007. Australian NSPSurvey National Data Report 2002–2006. National Centre in HIV Epidemiologyand Clinical Research, The University of New South Wales, Sydney,NSW.

ewton-John, H.F., Wise, K., Looke, D.F., 1984. Role of the lemon in disseminatedcandidiasis of heroin abusers. Med. J. Aust. 140, 780–781.

’Brien, S., Black, E., Degenhardt, L., Roxburgh, A., Campbell, G., de Graaff, B., Fether-ston, J., Jenkinson, R., Kinner, S., Moon, C., White, N., 2007. Australian Drug Trends2006: findings from the Illicit Drug Reporting System (IDRS). NDARC Mono-graph No. 60. National Drug and Alcohol Research Centre, The University of NSW,Sydney, NSW.

assaro, D.J., Werner, S.B., McGee, J., MacKenzie, W.R., Vugia, D.J., 1998. Woundbotulism associated with black tar heroin among injecting drug users. JAMA279, 859–863.

awelec, G., 2006. Mini review. Immunity and ageing in man. Exp. Gerontol. 41,1239–1242.

ieper, B., Templin, T., 2001. Chronic venous insufficiency in persons with a historyof injection drug use. Res. Nurs. Health 24, 423–432.

eid, G., Higgs, P., Beyer, L., Crofts, N., 2002. Vulnerability among Vietnamese illicitdrug users in Australia: challenges for change. Int. J. Drug Policy 13, 127–136.

hodes, T., Stoneman, A., Hope, V., Hunt, N., Martin, A., Judd, A., 2006a. Groin inject-ing in the context of crack-cocaine and homelessness: from ‘risk boundary’ to‘acceptable risk’? Int. J. Drug Policy 17, 164–170.

hodes, T., Kimber, J., Small, W., Fitzgerald, J., Kerr, T., Hickman, M., Holloway, G.,2006b. Public injecting and the need for ‘safer environment interventions’ inthe reduction of drug-related harm. Addiction 101, 1384–1393.

obinson, G.M., Kemp, R., Lee, C., Cranston, D., 2000. Patients in methadone mainte-nance treatment who inject methadone syrup: a preliminary study. Drug AlcoholRev. 19, 447–450.

pijkerman, I.J.B., van Ameijden, E.J.C., Mientjes, G.H.C., Coutinho, R.A., van den Hoek,A., 1996. Human immunodeficiency virus infection and other risk factors forskin abscesses and endocarditis among injection drug users. J. Clin. Epidemiol.49, 1149–1154.

toltz, J.-A., Wood, E., Small, W., Li, K., Tyndall, M., Montaner, J., Kerr, T., 2007. Changesin injecting practices associated with the use of a medically supervised saferinjection facility. J. Public Health 29, 35–39.

opp, L., Iversen, J., Conroy, A., Salmon, A.M., Maher, L., 2008. Prevalence and predic-tors of injecting-related injury and disease among clients of Australia’s Needle

and Syringe Programs. Aust. N. Z. J. Public Health 34, 34–37.

sao, J.W., Marder, S.R., Goldstone, J., Bloom, A.I., 2002. Presentation, diagnosis, andmanagement of arterial mycotic pseudoaneurysms in injection drug users. Ann.Vas. Surg. 16, 652–662.

oodburn, K.R., Murie, J.A., 1996. Vascular complications of injecting drug misuse.Br. J. Surg. 83, 1329–1334.