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

Mo.C.1414 - Mo.C.1418

Monday, July 8, 1996

Results: Cumulative seroprevalence (P) among injecting drug users was 31.2% (95% CI.
29.0, 33.4) (N = 1,770). P among non-injecting drug users was significantly lo 5er it 7 I%
(95% Cl: 6.9, 7.3) (N = 73,140).Of documented users. 68% were male, 96% were non
white, 80% were older than 29 years of age, 26% had a history of crack cocaine use, add
S2% had a history of exchanging money or drugs for sex.Twenty percent o.sers wer
diagnosed with a genital ulcer (syphilis, herpes or chancroid)- significantly morie than
nonusers (I 0%). Cumulative P was 32% among male users (382/ 1,1 94). P was higher
among users greater than 30 years old (34%, 482/1412), and users diagnosed with a genital
ulcer (39%, 140/355). P was lower among whites (20%, 3 1/1 59) than nonwhites.
Seroprevalence among users ranged from 34% (96/279) in 1990 to 29% (95/326) in 1995
with a peak in 1992 at 37% (125/334). P among female users surpassed that of males in
1993 and 1994.
Conclusions: Seroprevalence among injecting drug users remains relatively high compared
to that in nonusers attending ST)D clinics.These data support the continued need for the
promotion of HIV counseling and testing and the development of prevention programs targeting such patients.
Christopher S. Murrill PhD, MPH, NYCDOH, 346 Broadway Rm. 701, NY NY 10013,
Tel: [phone], Fax: 2 I 2 442 3482, email: murri [email protected]
Mo.C.1414
HIV SEROPREVALNCE TRENDS IN PATIENTS PRESENTING TO A PUBLIC HOSPITAL
IN NYC WITH ADMISSIONS UNRELATED TO AIDS, HIV, OR OF INFECTIOUS
ORIGIN, 1991-1994.
Vavagiakis, PTorian LV, Makki HA, Brennessel D,Wersfuse lB. NewYork City
Department of Health.
Objective: To measure HIV I seroprevalence (P) in a sample of hospital patient,, with an
admitting diagnosis not related to AIDS, HIV, or of an infectious origin at a pubhlic hospital in
New York City,
Method: An unlinked HIV-I serosurvey using remnant serum originally obtained for diagnostic purposes. Demographic, clinical and laboratory results were obtained via medical
chart review. No other sources of data were used. Samples were equally drawn firom male
and female patients and followed an age distribution that is meant to be representative of
the U.S. population.
Results: Between 199 1 and 1994, 13,806 patients were sampled that had admissions unre
lated to AIDS, HIV, or of an infectious origin. Overall P was 2.9% (N + /N 406/13,806).
Males were observed to have a higher P than females (4.3% [247/5737] v. 2.0%
[I 59/8069]). While a decrease of 23% in P was observed for males (4.7% [67/ I 422] in
1991 to 3.7% [50/1368] in 1994) P in females remained stable (2.1% [44/2127] in 1991 to
2.2% [39/1770] in 1994]. P decreased in males Hispanic (4.3% to 1.4%) and white (4.6% to
2.9%) patients but remained stable for male black patients (5.9% to 5.6%). Similar trends
were observed for female patients. Among male patients, P remained stable for all age
groups except for patients aged 35-44 where P decreased 39% (I14. 1% iin 199I to 8,6% in
1994). Among female patients highest cumulative P (4. 1%, 60/1462) was found in females
aged 35 44 years and moderate increases in P among all age groups except for patients age
15-24 where a 40% decrease in P was observed (1.6% in 1991 to 1.0% in 1994).
Conclusions: P remained stable in females who are admitted to this hospital with conditions that do not suggest a possible HIV infection. Although universal HIV C&T is not a reasonable goal, further studies of the profile of patients with unsuspected HIV infections in
general hospital populations should provide direction for targeted counseling programs.
PeterVavagiakis, NYCDOH, 346 Broadway Rm. 701, NewYork, NY 10013
Telephone 2 i 2-442-3480 Fax [phone]
Mo.C.1415
COUNTY, CITY,AND STATE-LEVEL CORRELATES OF U.S.AIDS PREVALENCE AND
MORTALITY: IMPLICATIONS FOR FUTURE HEALTH CARE NEEDS
Zimmerman, Rick S., StrattonTerry 1 University of Kentucky Lexington, KY
Objective: To consider nd explain the extent and ways in which patterns of AIDS incidence and mnortality in the U.S. are relsated to sociodemographic and geographic variables
and are lchangng at the city, county, and state levels.
Methods: AIDS data were taken from the CDC's AIDS surveillance reports for t1984- 1992.
State denmographic and social characteristic data were from the 1990 United States Census
and pop ulation estimates, 1984 -1992. State-level variables from the I1 990 Census included
populaion density, racial/ethnic distribution, rural/ urban, percent lving below the poverty
level, percent of individuals without health ins urance, proportion of never married adult
males, coastal/interior location, and whether one of the top 10 airports in the U.S. was
located in the state Similar Census based data were used for county and city level analyses,
with additional daita about medical personnel and racilities accessed in the Area Resource
File. Multple regression analyses were used.
Results: At the state level, population density and percentage of residents living in rural
areas wereb oth consistently related to AIDS incidence over time.The greater the populaConu densit of and the more urban a state, the greater the AIDS incidence rate. Greater
than 80% of the variance it AIDS incidence was explained by sociodemographic and geographic rirabes at the state level. Coastal states and those with high proportions of never
marred nrales had higher AIDS incidence rates early on in the epidernic, but rot more
rerentle.The only relationship of change in AIDS incidence at the state level was a positive
one with per-cent Black. At the county level, AIDS mortality rates were positively i-elated to
populaion density, negatvely related to percent of married families, positively related no
income, and positv ely related to per-cent of genersl shortterm hospitals with a dedicated
AIDS or ARC unit. About 40% of varriation in AIDS mortality rates were explarraed by these
variables at the county level. At the city level, trends in both incidence and mortality were
increasingly torward medium-sized rities and away from coastal cities.
Conclusions: Differens correelastes are impoe-tant for mortality arid iracidence, with geographin rables more significant for changes in incidence. Different socoo-geographic processes
are importrant across the 3 levels of analysis, with health care variables more important at
the county level Imphicatrons for AIDS health rare needs are discussed.
Kick S. Zimnaer-man, Department of Behavioral Science, Dniversity of Kentucky, 127
CC3MDB, Fexinyton, KY 40536-00386 Tel: [phone]. Fax' 606-323-.5350

Mo.C.1416
SECOND WAVE OF HIV INFECTION WASHES ASHORE: SOUTH BEACH HEALTH
SURVEY OF YOUNG GAY MEN
Webster, Robert D C[arrr' 'W Stempel RR,Taylor R, Bennett B. Dollinger RF. Florida
International Univers ity, INort Miarmi, Florida, USA
Objectives: To estimate the prevalence of HIV I infection among young men who have sex
with men (MSM) and to compare prevalence and risk factors with those reported for
young MSM in San Francsco.
Methods: A three stage probability samplhng scheme was used to identify residential
addresses and elig ble respondents in a four contiguous census tract area of Miami Beach
Florida, known as "South Beach". Unmarried men 18-29 years old residing at a randomly
selected address for at least 30 days were eligible if they reported ever having sex with a
man. After anonymity was assured and informed consent obtained, participants provided a
saliva specimen for HIV I antibody testing by modified ELISA; then were interviewed and
asked to complete a self-admninistered questionnaire. Procedures were similar to those
reported by Osmond DH, et al. Am J Public Health 1994: 84:1933- 1937
Results: Preliminary results for the first 30 MSM enrolled in South Beach (16.7% HIV- ianti
body positive) were remarkably similar to those found in San Francisco (17.6% posit ve). I
both studies. HIV infection was highly associated with receptive anal ntercourse (RAI) with
10 or more partners rin the past 12 months:

Variable:
Age: 18-23 years old
24- 26 years old
27-29 years old
Ethnicity: Hispanic
RAl with 10 partners, past year

%HIV I Antibody Positive
San Francisco (n=420) South Beach (n
4.8 7
10.7 20.0
28.9 176

25.0

n'.s

V
a)
H,
0
0.
C
0

Conclusions: In spte of different methods of specimen collection
only studies with representative samples of young MSM iwng in hi
States showed nearly identical results.
RD Webster FIU, 3000 NE 145 Street, ACI-394F, North Miami, FL)
305 919 5877 Fax: [phone] email:darrowwCtservax.fiu.edu

salva vs. finger prick), the
useholds ri the United
3181, USA Telephone:

Mo.C.1417
HTLV-I/II SEROPREVALENCE AMONG A COHORT OF HIV-INFECTED WOMEN
Telzak, Edward El, Hershow R2, Kalish L3, Hardy W1, Zuckerman E, Levine A4. Delapenha
R, DeHovitz 6, Greenblatt R7, Anastos K8. IBronx-Lebanon Hospital. 2U. of Ihno s, 3New
England Research Institute, 4U. of Southern California, 5Howard University, 6SUNY at
Brooklyn, 7UC San Francisco, 8Montefiore Medical Center, USA.
Objective:-To determine the seroprevalence of and risk factors for HTLV-I/II infection among
HIV-infected women enrolled in the prospective Women's Interagency HIV Study (WIHS)
Methods: From October 1, 1994 to May I, 1995, 1761 women were enrolled at six sites I r
the United States.This analysis is limited to the 533 HIV infected women at four of these
sites for whom complete serologic data are available. An interview-based survey was utilized to collect data on demographics, behavior and medical history HTLV I/I determinations were done using IFA and confirmed and typed by Western blot. Fishers Exact tests
and logistic regression were used to determine bivariate and mulivariate independent predictors for HTLV-I/II infection.
Results: Fifty two (10 %) of the 533 HIV infected women were coinfected with HTLV-1/II. Of
these, 43 (83%) were infected with HTLV II; it was not possible to distinguish HTLV I from
HTLV II for the rermaining 9 women.The race/ethnicity distribution was 54% black 24% white,
and 17% Hispanic; 301 women (56%) were 35 years or older Of the 533 women, 5%
(6/13 1) of participants from New York City 7% (7/96) from the District of Columb a, 10%
(16/165) from Chicago and 16% ( 23/14 ) from San Francisco were confected. H TLV-I/II serprevalence ranged fiom 2 among Hispanics to 15% among blacks Prevalence of coinfecton
for women with _500.200-499 and <200 CD4 cells was 10%( I 1/112), 10% (20/205) and
14% (17/122), respectively. Independent predictons for HTLV-I/I were: intravenous drug use
(OR=I).9, p<0.00 ), age> 35 years (OR=2.4, p=0.05), geographic site (p-0.02) and race/ethnicty (p=0.00 I). CD4 counts, foreign birth, socioeconomic status, number of sexual partners
and a history of sexually transmitted diseases were not associated with seronos try t>
Conclusions: Among HIV-infected women, intravenous drug use is the strongest predictor
of HTLV-I/Il coinfection, as has been previously reported among men. In addition, coinfectuon rates were significantly higher among blacks, among the San Francisco cohort and
among older women. On preliminary analyses, there is no evidence that sexual transmrssion
contributes to the high F-TIVI/Il seroprevalence or that HTLV-I/Il enhances progression of
HIV infection.
Edward E.Telzak, MD Bronx-Lebanon Hospital Center, 8th Flu: Dept. of Med. 1650 Grand
Concourse, Bronx, N.Y 10457
Mo.C.1418
DEMOGRAPHICS & NEEDLE-SHARING OF INJECTING DRUG USERS IN ARIZONA
Boyd, Denise K, Brown, C. Arizona Dept. of Health Services. Office HIV/STD Services Phx.
AZ USA
Objective: To identify the demographic and behavioral profile of injecting drug users in
Arizona.
Methods: Since 1991, all HIV-infected Arizona residents aged at least 18 years reported to
the Department of Health Services (ADHS) have been asked to part iapate in a standard
ized interview, the Supplermental HIV/AIDS Surveillance Project (SHAS) developed in collaboration with the Centers for Disease Control and Prevention and eleven health depart
ments.-The ADHS has developed procedures to ensure patient confidentiality and this pro
ject was reviewed and approved by the departmental human sublect review board. Data
from the HIV/AIDS cases report forms and the questionnaires administered from mid 199 I
through 1995 were analyzed.
Results: Of the 2240 individuals eligible, 49% agreed to be interviewed. Seventy five percent of those completing the interview indicated that they had ever used mind altenrng
drugs. Of these, 37% have injected non-prescription drugs The most common drug of

127