Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Mo.C.462 - Mo.C.465

Monday, July 8, 1996

Results: HIV prevalence was 30.1 % in 1990. it decreased to 17.6% in 1991 and then
remained quite steady during the following four years (15.8% in 1992, 18.0% in 1993. 19.0%
in 1994 and 16.3% in 1995).We observed 32 anti HIV seroconversions during 846 persons/yearn s (p/y) of follow up, with an overal incidence rate of 3.8 per 100 p/y.The inci-r
dence rate was 3.6 during the 1992 93 and 4.0 per 100 p/y during the 1994-95 period.The
anti-HIV incidence was higcher among females than males (5.5 vs 3.3 per 100 p/y) and
among younger IDUs, considenng both age (6.3 among the patients <24 years older vs 3.1
per 100 p/y among those 224 years; RR=2.1, 95% Cl I.0-4.2. p=0.04) and duration of drug
use (8.3 among VDUs with a history of drug use <3 years vs 3.5 per 100 p/y among those
with 3 years, RR=2.3: 95% C I. 51, I p o0.03). Only 130/998 IDUs ( 3.0%) reported
syringe-sharing and this practice was more common among those who seroconverted than
among those who remained negative, but the difference was not statistically significant (5.0
vs 3.5 per 1 00 py). By contrast, IV seroconversion was associated with having an HIV
positive partner ( 2.3 vs 2.6 per!00 p/y; RR=4.8, 95% C12.4-9.7, p<0.001).
Conclusions: Our data show a continuous occurrence of new infections among IDUs in
Rome.The risk of ienfection is higher soon after starting drug injection. Sexual t ransmission
seems to play n a more relevant role in the 1990s while drug injecting rik behavior sneems to
be less common than in the past.The use of d ata obtained through different epidemiological methods allows to better identify current risk factors and target prevention programs.
Mauro Zaccarelli,Via Portuens e 33, 001 49- Roma, ItalyTel: 39-6-55383417;
Fax [phone] emain: spa llianz( )flashnet.it
Mo.C.462
DRUG INJECTORS' SOCIAL NETWORKS CAN BE RISK FACTORS FOR
SYRINGE SHARING
Neaigus, Alan, Fedman SRp, Jose B*, Curtis R*, Goldsten M*, Ildefonso G*, De s Jarlais
DC**. "National Development & Research Institutes, Inc., NewYork, NY, USA; **Beth
Israel Med. Ctrs, New York, NY, USA
Objective:Jo determine whether injecting drug users' (IDUs) social networks help explain
why some IDUs engage in receptive syringe sharing (injecting with a syringe after another
injector used it first).
Methods: Face-to-face structured interviews and HIV antibody testing were conducted with
767 street-recruited IDUs in New York City between 199I and 1993; 40% (272/687) were
HIV infected. Subjects were asked about their: -IV drug and sexual risk behaviors in the 30
days prior to the interview: relationship to, risk behaviors with ad characteristics of, up to
10 social networ k members with whom they injected drugs or had sexual relationships; and
which network emembers injected with ea ch other or shared syringes with each other As
previously publ ished, a comparison of linked pairs of independently-interviewed subjects
indicrated tlat their reports about network members were reliable. Adjusted odds ratios
(AOR) re all significant at p < 0.05.
Results: Receptve syringe sharing (RSS) was engaged in by 35% (267/765) of subjects. Of
the 267 engaging in RSS, 52% shared with a friend, 28% with a spouse or lover, 25% with a
running partner, but only 5 with a stranget. 52% (I 38) of those who shared reported that
at least 2 other members of their soc Ilnetwork shared syringes with each other. In simultaneous multiple Iogistic regression (N=733) (controlling for- the number of IDU network
members, drug use, current treatment s Ae tatus, selling syringes, and demographic variables),
subects were more kely to share syr nges if their network member s shared nyringes with
each other AOR, 6.29); rntwor E embers with whom subjects injected were their
sex partners (ACR 1.66), people with whom they were in daily contact (AOR 2.56),
or people with whom n they had injected for more than I year (AOR 2.60).When HIV
status wa s added to the equa ition (N 658) these network variables remained significant.
Conclusions: More than 15 I yeitars after the onset of the dHIV epidemic in New York City, a
substantial propor tion of IDUs in ethis s tudy reported tha t they engage in receptive syringe
sharing. However, syringe sharing is much more likely to occur within IDU social networks
than across netwo ks with st rangers. Socia influence frome IDU social network s may
encourage syringe sharing. Interventions should develop mechanisms to use such network
influence to reduce ra ither than encourage syringe sha ring.
Alan Neaigus, Nat i onal Development and Researc h Institut es, Inc. I I Beach Street,
NewYork,NY 100o13, USA.Phone [phone] x514 Fax [phone]
email alneadix.netcorm.com
Mo.C.463
EXPLAINING AN OUTBREAK: DETERMINANTS OF HIV SEROCONVERSION
AMONG INJECTION DRUG USERS IN VANCOUVER
Patrick, David M 1,3, Strathdee SA2,3, Ofner M4, Archibald CP4, Craib K(P2, Cornelisse
PGA 2, Schechter MT2. 3, Eades G 2, O'Shaughnessy MV23, Rekart ML1. I BC Centre for
Disease Control, 2BC Centre for Excellence in HIV/AIDS and 3University of BC,Vancouver,
BC, Canada: 4Laboratory Centre for Disease Control, Ottawa, ON, Canada.
Objective: A ma jor increase in new posite HIV tests among injection drug users (IDUs)
prompted a case ontro investgation to establish factor s associated with seroconversion.
Methods: Caonsstore IDDs with a cow posite DIV tent docurerted aster Januatry I,
1994, who had a dorunaene negative tnt within the prior 1 8 morthn. Controns required
two documented nogitree tosts in the note time fiame. Subjects completed a questionearre oe demographic. pnyoho-norial, needle usng ned nexunl fictors focusing oe the rotor:
tent enterval hoetweon the two DIV tents denaod shove).
Results: Eighty errne cases B Imale) ard 192 1I29 masen controlt participated. Canes and
coetirols wetre similan with r epect to age (p 0.562), gender (p-0.820), othnicity (p-0.1I92),
education (p-0 61I6) ned rotor tent interaI (p OS I). Canon were more i kely to report
shatring i gs (OR 2 275. 95% U1:.48,5.13), core fiequtort uno of needle exchange
(p-0o.0008) and eoreo frequent eliectron of cocaine (p 0.0 13) thae did coetrols durreg the
renttt irterval. hemasn rates alto reported r higher frequency of sex trade clreet roetacts than did female contro (p-O.O2 I). leversno asnocratrons were observed for use of
noe-in notior diru, (OR -0.21, 95%(Cl 0.08,0.57). ntablo housing (OR -0.51: 95%(2Cf
0.30,0.87) rid vaginal sex DOR <or iral r0.35, 95%C1: 0.1r9,0.68 <red OR fur hemalsnot-0.18:
35%Cl: 0.0,0.65). Mlult site Iogi tic regrentron analynen for sll nubects demonstrsted
shhsting r gs with a istrnger 3aj OR -2.96, p <0.001), unstable housing (adj. OR -c 2.08: p
-0.028, vaginal tex (asdj. OR -0.44: p:0.004j sod TUC (etariijuana/hashish) use (ad1. OR:0.40: Ia- 0.001 to lao independently asiatted with nororonvornroe.

Conclusions: Needle sharing continues to play i key role ia HIV transmission in Vancouver.
Programs that address the role of housing, spec fic management of cocaine addiction and
the use of non- injection drugs to help IDUs control their condition require evaluation.
Dr DM Patrick, 828 West 10th Avenue,Vanc ouver British C ol umbiaVSZ ILt 8 Canada
Telephone: [phone] Fax: [phone] emnail: dmpatric(bcsc02.gov.bc.ca
Mo.C.464
YOUNGER AGE OF INITIATING INJECTING DRUG USE IS ASSOCIATED WITH
RISKY BEHAVIORS AND HIV INFECTION IN SHORT-TERM INJECTORS
Doherty, Meg Ct, Garfein RS*, Monter roso E', Brow n Dfr, HuangH', Park M e,VlahovD.
*The REACH Study Johns Hopkins U., Baltimore, Maryland o Centers for Disease Control,
Atlanta, Georgia.
Objective:To describe the events surrounding initiation of injecting drug use (IDU) and
other risks of HIV infection, by age of inilt ation, in a cohort of short-term in ectors.
Methods: In Baltimore, 18-25 year old IDUs (N==218) recruited ma street outreach, were
interviewed regarding the Rcircumstances of in itiating IDU the characteristics of persons present at initiation, and lifetime and curren t HIV related inectnge and sexual risks. Blood was
assayed for anti bodies to HIV. Data weren analyzed u sineg logistic regression methods.
Results: IDUs who initiated at < 19 years (N a103) compared to those at -9 years
(N-: 115) were more likely to be HIV+ (19% vs. II%, p=.14), white (24~ vs. 8% p<.005)
and have < 12 years of education (69% vs. 53%, p<.05). At rst njection, most were initiated
by the same sex peers, but younger initiates (<19) were more likely to reporti larger net
work (>2 other people) at initiation (OR-2.1, 95% Cl 1.2 -=.9) and risky mnecting Ibehaiors
esuch as: cocaine inecti on (OR2.3, 95% Cl 1.3-4. I ),"shooting gallery use (OR 22., 95 CI
1.2-3.8), "backloading" (OR-2. I, 95% Cl I. I -4.0) and being assisted n ijectin g by a pers on
known to be HIV infected (OR=3.6, 95% CI 1.1 -16.7), Younger ni tiates also reported being
assisted by >2 different people before being able to inject on their own (OR-2.4, 95% CI
1.3-4.6). In terms of HIV-related sexual practices, younger initates more often repomted
being bisexual/gay (OR-6.4, 95%e Cl 2.0-19.8), having anal intercourse (OR-3.7, 95% Cl
1.8-7.7), a history of sexual assault (OR-.8, 9500 Cl 1.0-3. '-), and trading sex for money or
drugs after starting to inject (OR-1.6, 95% Cl 0.9-2.9). Ian n ultvariate model, tradong sex
for money or drugs was associated with HIlV prevalence among younger initiates (OR= 3.1,
95% Cl 1.1-9.3) when adjusting for cocaine injection and being bisexual/gay In a mode for
the entire cohort, regardless of age of initiation, trading sex for money or drugs (OR= 2.5,
95%o CI 1.1 5.6), cocaine injection (OR=7.8, 95% CI 1.8-34.5), and being bisexual/gay
(OR=3.2 95% CI, I. I-9.4) were associated with HIV
Conclusions: Among short-term injectors, younger age of initiation was associated w ith
more firequently reported risky drug use and se practices, as well as, higher- rates of HIV
infection. Even though youngerinitiates practice risky injecting behaviors, risky sexual practices are more important for HIV infection in younger initiates; interventions should target
these behaviors.
M.C Doherty Johns Hopkins SHPH 624 N. Broadway. Rmi. 397, Baltimore, MD 2 205
Telephone: [phone] Fax: [phone] e-mail: mdohcrty(%phnet.sph.lhu.edu
Mo.C.465
THE RELATIVE IMPORTANCE OF YOUNG AND RECENT-ONSET INJECTION DRUG
USERS IN THE AMSTERDAM HIV EPIDEMIC, 1986-1995
Fennema, Johan S.A., vin Ameijden, E.J.C., van den Hoek,(].AR., Cout nho, R.A. Murcpal
Health Service, department of Public Health and Enwronment, Amsterdam, the Netherlands
Objective: To study whether young and/or recent onset tnecting drug users (IDU) ire at
increa sed risk for HIV infectione, andr to ass s tre ds from e86 through 1995 n. isk behR
iors and HIV incidence among subgroups of IDU,
Methods: Associations of age (<26, 26-30, >30 year d tiene snce o nset of injection (<4,
4-6, >6 years) with HIV prevalence and borrowing used needles (BOR) are nas esed
among current IDU entering the Amsterdam cohor t between 1989- 95, and ompared with n
IDU recruited in streetsurveys 1990 and 1993. HIV incidence (INC) is computed using only
repeat visits. Logistic and Poisson reg ression s used for trend analyoses 1986-1995 in BOR
and HIV- INC.
Results: Sizable portions of IDU were y oung (< 26 years: 4-30%) orn recent -onset IDU
(within last 3 years; 17-21 %). Between 41% and 63% t of young IDU was also recent-onset
IDU. HIV prevalence was 12- 24 among young IDU and 5-12% amon g recent-onset o IDU.
Youneg IDU more often (p<0.05o) reported BOR than older IDU. Recent onset IDU were
lenss often aware of their HIV serostatus than moren remoten a e-onset IDU. Between 1986 and
1995, BOR declined significahntly ein all nubgroups, except among young IDUTrends in BOR
per subgroupns differed significantly for age, but enot for onset of nject ion. H IV INC decned
significantly from 18/I 00 person-years (PY) in 1986 to 5/100 PY i99rDmA in gher HIV -INC
was associated with younger age (RR.31 per category; p0.02) and recent-onscet of IDU
(RR: I.41 per category: p-0. 005). After adjustment for mutt collineity of age, time nice
onset of IDU and calendar time, a higher HIV-INC was associated with recent-onset of IDU
(p C.04) but no longer with age (p-0.67).Trends in the HIINC per subgroups of age and
time since oneet of IDU did not differ nign rhocneil>.
Conclusions: Despite the DIV epidoeaic, important pmroporn nof diu agsem- tntt nect
meg. hIV prevalence in alt-eddy moderate no high among yoecg and meceont-nmet IDU. DIV
incidnce wan, and still ns highost among moent-ontet IDU. Hove, rites of brmm-r-og son
highest among IDU aged under 26, and have not demrented rn the at 10 yers
Nonrandom patterns 01 borrowing (lie ronalsc among IDU na e xplacin ns cict
meg observations. IDO who have necently star ted injocting Ia eld ho approhednote
actvely for DIVteng and counslling.

U
L
O
C
0
z
JSA FenneNa, MD; Municipal Health Service, Nieuwe Achtern gracht
Amsterdam, Netherlands.Telephone: [phone] Fax: +31-2C
Email: [email protected]

VWT

41