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

Tu.C.2533 -Tu.C.2537

Tuesday July 9, 1996

Selected Results:

Parameter
General coping
Mother warmth
Father warmth
Current family cohesion
significant other emotional supp.
Marital harmony
*p less than.05; "p less than.01

Frequency of Needle Sharing
With Familiar People With St argers
(Pearson correlation)
-0.23* 0!
-0.18* 0.25
-0.03 0. 13
0.07 023"
0.17 -0.25"
-0.18" -0.19"

Conclusion: Overall the findings indicate that risk factors from several areas (personality
relationship with parents, siblings, significant other and friends) increase the probability of
needle-sharing behavior among HIV+ and HIV- female IDUs.
JR. Masci, Elmhurst Hospital 79-01 Broadway Elmhurst, NY USA I 1 373 Telephone: [phone] FAX [phone]
Tu.C.2533
HIV RISK BEHAVIORS AND DEPRESSIVE SYMPTOMATOLOGY AMONG INJECTION
DRUG USERS IN LOS ANGELES COUNTY
Lopez-Gomez I*, Ford W*, Kerndt P *,Weber M, Monterroso E. Los Angeles County
Department of Health Services, HIV Epidemiology Program, Los Angeles, CA USA
Centers for Disease Control and Prevention, Atlanta, GA USA
Objective: To describe the association between depressive symptormatology and HIV risk
behaviors among a cohort of street-recruited injecting drug users.
Methods: Between August, 1994 and January, 1995, four hundred forty four treet-recruited
injection drugs users were tested for HIV and administered a baseline HIV risk behavior
survey Levels of depressive symptoms were determined by the Center for Epidemiologic
Scale -Depression Scale (I 2-item abridged version).
Results: Overall HIV seroprevalence was 1.8%. Eighteen per-cent of the study participants
were identified as having high levels of depressive symptoms. Participants with high levels of
depressive symptoms were more likely to engage in high risk sexual and drug injection
behavior than were those with low levels of depressive symptoms. For example, participants with high levels of depressive symptoms were more likely (p <.05) to report illegal
sources of money sharing contaminated needles, shooting in galleries or public venues, a
greater number of steady and casual sex partners, and exchanging money and/or dru gs for
sex. Levels of depressive syreptomatology were not differentiated by age or race/ethnicity
However, females exhibited a three-fold increased risk for high depressive syrnpton itology
when compared to mrales (OR = 2.92; 95% C.I. 1.73, 4.94).The association between highf
levels of depressive syrmptornatology and high HIV risk behavior remained essentially
unchanged after adjustment for gender.
Conclusions:The association between HIV risk behaviors and high levels of depressive
symptoms among IDUs supports the notion that depression screening should become a
standard modality for drug treatment and rehabilitation in this population. In general, our
research suggests that efforts to prevent the spread of HIV in the IDU population must
consider not only risk behaviors but also psychological aspects that may underlie IDU
behavior:
Javier Lopez-Gomez, 600 S. Commonwealth Ave., Suite 1920 Los Angeles, CA 90005.
Telephone: [phone] Fax: [phone]
Tu.C.2534
HIV RISK BEHAVIORS AND BELIEFS OF HOMELESS AND NON-HOMELESS
OUT-OF-TREATMENT DRUG USERS
Malotte, C. Kevin, Wood MM, Rhodes F. California State University Long Beach, California
Objective: To examine differences in HIV risk-related behaviors and beliefs between homeless and non-homeless out-of-treatment injection drug users (IDU) and crack cocaine users.
Methods: Nine hundred seventy-five (677 men, 298 women) active injection drug and/or
crack cocaine users, recruited through street outreach, completed standardized interviews
regarding use of shared needles, bleaching of needles, condom use with main and casual
partners, and choice of sexual partners. Beliefs about these behaviors, belhavioral inrtentions,
and perceived self-efficacy to engage in risk-reduction activities were also assessed. Stage of
behavior change (a summary measure of behavior and intention) for HIV risk related
behaviors was calculated.
Results: Nearly half (47.6%) of the respondents reported that they considered themselves
to be currently hormeless. Men were more likely to report homrelessness than women
(51.6%9 vs. 38.6%). Injection drug users who did not use crack cocaine were less likely to
report hormelessness (35.5%) than either non-IDU crack users (49.7%) or IDUs whio also
used crack (54. 1%).There were 653 IDUs in the sample; 75% of both homeless and nonhomeless IDUs reported that they used their own, non shared, needle the last time they
injected drugs, and 49% of each group reported use of their own needle every time they
injected. No difference between groups was found in positive or negative beliefs about use
of own needle, but homeless individuals were less confident of their ability to get new needles and put off injection if they did not have their own needle. Stage of change for not
sharing needles did not differ between homeless and non-homeless grou ps. Men stage of
change also did not differ between groups for bleaching shared needles, quittin, drugs, condom use with main partner, and condom use with occasional partners. Homeless individuals
were at a lower stage of change for not having sex with IDU partners (p-.001 ) partners
who had other partners (p.0003), and partners of unknown HIV status (p.0I). Specific
behavioral beliefs and perceived self-efficacy also differed between groups for most behaviors.
Conclusions: Homeless drug users did not engage in significantly greater risk behavior than
other drug users except in choice of sexual partners. Specific belieRs and perceptions of
ability to engage in risk reduction behavior- did differ between groups, however red should
be assessed in risk-reduction interventions.
C. Kevin Malotte, Ctr for Behavioral Res and Servs, 920 Pacific Ave, Long Beach, CA 908 13
Telephone: [phone] Fax: [phone] Email: [email protected]

Tu.C.2535
RISK TAKING BI:HAVIORS AMONG SHORT-TERM INJECTING DRUG USERS
ASSOCIATED WITI IfY'E OF DRUG INJECTED
C rfein Richea, I M(, Brown D, Huang H, Pitanlril SS,Vlahov D. The REACH
Study Johns 1 ' nr,of Hygiene and Public Health, Bailtimore, Maryland. USA
Objectives: Io c--, e: 7r, i-gh risk behaviors among injecting drug users (IDU) who
injected cocaine with i it iut heroin compared with those who only injected heroin.
Methods: Standardized inter,rews were conducted to assess risk of human immurnodeficiency virus (HIV) infecti nrmong 18 to 25 year old IDUs recruited via street outreach in
Baltimore. Sera colleted were tested for antibodies to fIV. Data were arnalyzed using chisqcuare and logistic r egi -sirn methods.
Results: In 218 IDUs. w o.-bserved that cocaine injection was associated with HIV infection
even after controlir for trading sex for money or drugs and sexual preference in a logistic
regression model (OR 7.8, 95% CI 1.8- i.5). Cocaie inlectors (N= 147) were significantly
more likely thanHe oin rjectrs (N- 7f) to be white (22% vs. 3%, p<0.001 ), have been
homeless in the past,ear ( 0% vs. 35%, p0.036), been in drug treatment in the past year
(45% vs. 27%, p0.010), and have fir t injected at age 19 or younger (54% vs 34%
p--0.006).They w e sirilar isth retrd to gender education, source of income, and use of
needle exchange. Significantly more ocaine injectors needed help injecting (47% vs. 31%,
p=0.025), used needles th.t appeared to be dirty (36% vs. 14%, p=0.001 ),"bickloaded
drugs (29% vs. 10%. p=0.00), injected for more than 3 years (48% vs. 23%, p<0.001). and
in ected more tha orice d lyduring the past six months (4 3% vs. 21%, p=0.002).
Differences in seu r i rcfi e incruded hiving h ad an IDU sex partner (63% vs. 39%,
p-0O.00 I), having anal,interi co. ssince it nitiationr (23% vs. 10%, p- 0.026), being a victim of
sexual assault ( 34% vs 20(, p:--0.030), anid trading sex for monrey or drugs since nitiation
(40% vs. 27%, p=0.04). Age at sexual debut and number of ifetime sex partners were not
significantly different.
Conclusions:TIhe assocition between HIV infection and injecting cocainre is complicated by
the observation 'at,care inifecting may be a marker of other high risk behaviors related
tc HIV transmission. Prevention efforts should consider a difference in risk depending upon
drug of choice.
R.S. Garfein, Johns Hopnst ifpH, 624 N. Broadway, Room 897. Baltimore, ME) 21205
elephone:410 S- lit 0 PFfAX:41 05505 385 e51 a-MIr,rft I01v worsder em. rD gov
Tu.C.2536
RISK CHARACTERISTICS OF INJECTION DRUG USERS (IDUS) PARTICIPATING IN
THE VACCINE PREPAREDNESS STUDY (VPS)
Metzger; David S",Vr isdy i ' Marmor M' ( Gross M *University of Pennsylvanra,
" 'New York University,. Abt Associates and the HIV Network for Eficacy Trials
(HIVNEF)
Background: fhe Vaccine Preparedness Study (VPS) enrolled 4884 HIV-seronegative individuals from 8 sites in 7 US les berit e-n March and October of 1995 This study is being
implemented in prepar,tior for efficacy trias of HIV vacncies and other preventive nterventrions.The cohorrt irfclres gay mer, I.r and woresn at eevated risk of HIV infection.
IDUs were recruited ii Philrfdelphi and New York (Cty.
Objective: To exarnre the demnographic and isk caracterist-s of IDUs enrolled in the
VPS.
Methods: A va iety of outre ch strategres were used to r ecr urt HIV seronegative IDUs with
a history of drug rnectioni fsthe rontfrs prior to enrollment.
Results: A total of 1046 ID1)s were enrolled. Of these, 340 (30%) were "roliover subjects
frrom previous vaccine preparedness cohorts whtle 706 were rewly recruited. thirty-one
percent were women; 1% African-Amer ican, 19% L atinro, and 37% white: 63% had cornmpleted high sshool nd 8,.(% were unempoyed it the timne of enr-ollnent. Heroin was the
most frequently injeted dru (88%), 39% reported injecting cocaine, and 45% speedball
(heroin and cocairne rixed) Ihirty seven percent reported shairing needles, 22% used
shooting galleries, 32% shre cottons, 28 v shared rinse water and 4 1% reported shanring
cookers. Although needle exc hange progrars were utilized by 6)% of the cohort, subjects
continued to get needle, frtm other sources including: on the street (57~,). diabetics
( i%), and drug sellers 29% /).f sthe 921 subjects (83%) who reported being sexually
active, 76% report u irsn ridor, less th.n half of the time. When compared to the
rolloever srbjiti is f nrolewere s cn ricrntly more likey to report
drug related r isk be,havi, s Lootir Lp wi h others (58 I. 17,), needle sharing (43% vs.
22%), and usi rg a ne ie after other wi out cl(nin- (25 vs. 12o). Newly recruited subjects were also more lily to report unpritected sexual activsty. Gender and geography
were less saient fr explai risk behaviors.
Conclusion: HIV kisk behav ors are prevlent amorng drug users enrolled in the VPS, particularly those who hiad not participated in prior vaccine preparedrness research.
David Metzger Ph.D. Un iversity of Pennsylvania. 3900 Chestnut Street, Philadelphia, PA USA
191041el: 215 823 6098 Fa: 2I5 82 42418 email: metzoeratresearch.trc.upen.edu
Tu.C.2537
RACIAL/ETHNIC DIFFERENCES IN HIV- I SEROPREVALENCE AMONG INJECTING
DRUG USERS (IDUS): DO SOCIAL NETWORKS MATTER*
Jose. Ben<y Friednain ER'. 5sfeagus A, Curtis 18. lGoldsten HE.' ffdefonso (.0, Des Jail art
DC I. a Natisisal [)evelopnrernt & Reserch Instiutes, New York, NY, USA, ( Bethi Ittraol
lsedical Center NewYork, SY USA
Objective: Steudres in f [thUitled Stots obten report sigher HIV seroprevaece amonig
black and Punto ivRi itt IDUs Irisanrong whsite lfDfs. L~rffe-ences in rndrvidual-level nexual
and drug risk behv or of ID1s seldom explain the raeal/ethn cd fferences n HIV. This
sludy examines how sonl rind risk retworks of IDUs contfuie to racial/ethnrc vanations
in EHIV son opreval ere.
Methods: 662 IDUs (175 black, 264 Puere to Rican, 223 white) were street-recruited
thrcough ethnographically dirtedf chain referral front Bushawiek, New York City and HIV
tested. Scubjects worn interviewed about shi seuua ifnd dirug rtsk beraviors. tfhe charactersties red ritk behaviors of piionp/ wills s-shorn they used dirugs or hasree in the prioc 30
days, and their risk behsavior s with tftose people. Net-onrk links between study part-erpants
were ascertaned through a uanety of methods Egoimentric network measures were
derived from self reports I inkages amorng study part ipants were analyzed using graph-theoetc algebraic techniques to derive sociometric network cr es.

357