Back grey_arrow_rt.gif
 
 
Low Perceived Risk and High HIV Prevalence Among a Predominantly African American Population Participating in Philadelphia's Rapid HIV Testing Program
 
 
  "Our findings support the CDC's guidelines for routine opt-out testing for all Americans aged 13-64,27 and underscore the need for routine, opt-out testing in public clinics, which can effectively diagnose individuals who may underestimate their HIV risks.....greater efforts are needed to encourage more individuals to undergo HIV testing in Philadelphia.....Moreover, because stigma can inhibit persons from seeking HIV testing, routine testing can also help destigmatize HIV testing..... routine HIV testing can be scaled and added as a complement to existing services in public clinics that serve urban populations.....Expanding rapid HIV testing in urban areas should be an important public policy priority"
 
"city-wide HIV incidence in Philadelphia is 114 per 100,000, the sixth highest of any metropolitan area in the nation.......this large sample represents the entire population of individuals undergoing rapid HIV testing during Philadelphia's public HIV testing program over a more than 2-year period.....our findings suggest that self-perceived HIV risk is an insensitive criterion for HIV screening in this urban, predominately African American population. ......Individuals testing HIV positive dramatically underestimated their own HIV risk; two thirds of individuals who tested HIV positive believed they were at zero or low risk for contracting HIV.......It is noteworthy that 90% (56/62) of all HIV-positive individuals reported either never or only sometimes using condoms......It is noteworthy that cocaine use and having a same sex partner were the strongest predictors of testing positive for HIV in this urban population
.....The rapid HIV testing program in Philadelphia identified a population exhibiting multiple high-risk behaviors, including low rates of condom use, substance use, and exchanging sex for money or drugs. However, individuals engaging in high- risk behaviors typically did not perceive themselves at risk for contracting HIV. In our sample, many individuals never used condoms and a large proportion of the others only used condoms infrequently; only 11% of individuals reported always using condoms. Many also reported more than five sexual partners, and individuals reporting more than five sexual partners were more likely to perceive their HIV risk as zero or low and less likely to use condoms"
 
AIDS PATIENT CARE and STDs Volume 25, Number 4, 2011
 
Amy Nunn, Sc.D.,1 Nickolas Zaller, Ph.D.,1 Alexandra Cornwall, B.A.,1 Kenneth H. Mayer, M.D.,1 Elya Moore, Ph.D.,2 Samuel Dickman, A.B.,1 Curt Beckwith, M.D.,1 and Helena Kwakwa, M.D., M.P.H.3 1Alpert Medical School of Brown University and The Miriam Hospital, Division of Infectious Diseases, Providence, Rhode Island. 2Murdoch Children's Research Institute, Melbourne, Australia. 3Philadelphia Department of Public Health, Philadelphia, Pennsylvania.
 
Abstract
 
African Americans are disproportionately infected with HIV/AIDS. Despite Centers for Disease Control and Prevention (CDC) guidelines recommending routine opt-out testing for HIV, most HIV screening is based on self- perceived HIV risks. Philadelphia launched a rapid HIV testing program in seven public health clinics in 2007. The program provides free rapid oral HIV tests to all patients presenting for health services who provide informed consent. We analyzed demographic, risk behavior, and HIV serostatus data collected during the program between September 2007 and January 2009. We used multivariable logistic regression to estimate the association between behavioral and demographic factors and newly diagnosed HIV infection.
 
Of the 5871 individuals testing for HIV, 47% were male, 88% were African American, and the mean age was 34.7 years. Overall HIV prevalence was 1.1%. All positive tests represented new HIV diagnoses, and 72% of individuals reported testing previously. Approximately 90% of HIV-positive individuals and 92% of individuals with more than five recent sex partners never, or only sometimes, used condoms. Two thirds of individuals testing positive and 87% of individuals testing negative assessed their own HIV risk as zero or low. Individuals reporting cocaine use and ever having a same sex partner both had 2.6 times greater odds of testing positive. Condom use in this population was low, even among high-risk individuals. Philadelphia's program successfully provided HIV testing to many underserved African Americans who underestimate their HIV risk. Our results nevertheless suggest greater efforts are needed to encourage more individuals to undergo HIV testing in Philadelphia, particularly those who have never tested.
 
Over 1 million Americans live with HIV/AIDS, more than 20% of whom do not know they are infected. Al- though African Americans represent 13% of the U.S. popu- lation, they account for approximately 45% of new HIV infections, and the rate of HIV infection among African Americans is seven times that of white Americans.1 In- dividuals who do not know their HIV status may unknow- ingly transmit the virus2-6; more than half of new infections in the United States are spread by HIV-positive individuals who do not know they are infected.7,8 Several studies have found that knowing one's HIV status can lead to a reduction in high- risk sexual behavior.9-12 Furthermore, individuals who have been diagnosed with HIV are more likely to initiate highly active antiretroviral therapy (HAART), which lowers viral load and decreases the risk of HIV transmission.13
 
Although African Americans test for HIV at higher rates than individuals of other races, 14-18 they are significantly more likely to present for HIV testing late in their course of infection,7 and the rate of AIDS diagnoses is 10 times higher among African Americans than among whites.19 African Americans with HIV are also less likely to receive anti- retroviral therapy than people of other races.20 In 2006, AIDS- related mortality was almost 2 times higher for African Americans than for whites.21
 
To address nationwide underdiagnosis of HIV and these alarming health disparities, the Centers for Disease Control and Prevention (CDC) adopted routine opt-out HIV screening guidelines for all patients ages 13-64 in medical settings, irrespective of HIV risk. In an effort to reduce barriers to testing, these new recommendations eliminate previous requirements to accompany each HIV test with pretest counseling and separate written informed consent.22
 
Philadelphia's HIV incidence rate is 114 per 100,000 people, approximately five times the national average and the sixth highest of any metropolitan region nationwide.23 Although 43% of the population in Philadelphia is African American,24 of 1123 new HIV cases diagnosed in Philadelphia in 2006, 780 (69%) were among African Americans.23 In 2007, to address the city's high rates of HIV incidence and to comply with the new CDC guidelines, Philadelphia introduced a city-wide rapid testing program in public health clinics. The program has focused on uninsured and underinsured individuals with limited or no access to health services, many of whom are African American. Philadelphia's program has been im- plemented with financial support from the state of Pennsyl- vania and a CDC grant for metropolitan areas most affected by HIV/AIDS. We assessed actual and perceived risks in in- dividuals undergoing rapid HIV testing in Philadelphia's program, analyzed behavioral and demographic factors as- sociated with newly diagnosed cases of HIV, and explain the important role of Philadelphia's rapid HIV testing program in diagnosing African Americans.
 
Results
 
Between September 2007 and February 2009, 5871 unique individuals underwent rapid-testing for HIV in Philadel- phia's public testing program. The vast majority (88%) of in- dividuals reported their race as African American, and most (72%) reported having been previously tested for HIV. The mean age of the population was 34.7 years and the proportion of men and women were similar (47% versus 53%, respec- tively). Sixty-two individuals tested positive for HIV, yielding an overall HIV prevalence of 1.1% (Table 1). All 62 individuals testing positive represented new HIV diagnoses. While we do not have data on the number of individuals entering HIV care for this particular subpopulation of testers, 90% of all indi- viduals who tested positive in this program since 2007 at- tended their first clinical visit.
 
Risk behaviors
 
Table 1 contains HIV-related risk factors reported in the questionnaires administered at the time of testing, stratified by gender. More than a third of women and 27% of men reported never using a condom during sex, and 55% of the population (61% of men and 50% of women) reported using condoms sometimes during sex ( p < 0.001). Nearly four times as many men as women had more than five sex partners in the last 12 months (16% versus 4%, p < 0.001), and 5% and 11% of men and women, respectively, reported exchanging sex for money or drugs (p<0.001). Among those reporting more than five sex partners in the past 12 months, 92% reported never or only sometimes using condoms (Table 2). The vast majority of both men and women self-identified as heterosexual; 7% and 8% reported same sex partners, respectively. Differences in drug and alcohol use were reported between men and women; the largest observed difference was for marijuana use (54% for men and 39% for women, p<0.001). Approximately 4% of men and 2% of women reported heroin use.
 
Perceived HIV risk

 
Approximately 4% of men and 3% of women self-reported their own HIV risk as high, while medical assistants perceived that 12% of men and 6% of women were at high risk for contracting HIV (Table 1). Approximately 85% of men and 88% of women believed they had zero or low risk of HIV infection, compared to medical staff's perception that 60% of men and 72% of women were at zero or low risk of infection.
 
Among individuals reporting never using condoms during sex, 90% perceived themselves to be at zero or low risk of HIV infection (Table 2). To examine whether individuals engaging in high-risk behaviors perceived themselves at risk for HIV infection, we examined potential associations between reported risk factors and perceived risk for HIV infection. Variables included in the final models for self-perceived HIV risk are shown in Table 3. A history of STI was associated with a reduced odds of having a moderate or high perception of HIV risk (OR 1/4 0.55; 95% CI 0.45-0.68). The odds of perceiving one's own HIV risk as high was 2.8 greater if more than five sex partners were reported, and 1.9 times greater for HIV positive participants (95% CI: 1.56-5.10, 1.04-3.37, respectively). There was wide discrepancy between an individual's perceived risk and the medical assistant's risk assessment; individuals who perceived their HIV risks as zero or low were frequently at high risk for contracting HIV. There was an 8.1 times greater odds of the individual believing (s)he was at moderate or high risk (95% CI: 6.71-9.84) if the medical assistant assessed his or her HIV risk as moderate or high.
 
Predictors of HIV Infection
 
We also examined specific predictors for incident HIV in- fection through multivariable analyses (Table 3). HIV testers' assessment of HIV risk was the strongest predictor of HIV- infection in the model. In individuals for whom the medical assistant rated at moderate or high risk for HIV infection, there was a 4.2 greater odds of being HIV infected (95% CI: 2.21-7.99). There was a 2.4 times greater odds of testing HIV positive associated with cocaine use, and a 2.6 times greater odds associated with having a same sex partner (95% CI: 1.30- 4.46, 1.33-5.03, respectively).
 
Individuals reporting more than five sexual partners were less likely to be HIV-positive compared to individuals re- porting fewer than five sexual partners (OR 1/4 0.20; 95% CI: 0.05-0.76). Additionally, reporting a previous HIV test was associated with decreased odds of HIV infection (OR 1/4 0.45; 95% CI: 0.25-0.79). Self-perceived moderate or high-risk for becoming infected with HIV was marginally associated with incident infection (OR: 1.85, 95% CI: 1.02-3.36).
 
It is noteworthy that 90% (56/62) of all HIV-positive individuals reported either never or only sometimes using condoms. Although not statistically significant, it is particularly relevant given that 69% of HIV-positive individuals included in our analysis perceived themselves to be at no or low risk for HIV infection compared with HIV 87% of HIV-negative individuals ( p < 0.001, data not shown).
 
Discussion
 
The rapid HIV testing program in Philadelphia identified a population exhibiting multiple high-risk behaviors, including low rates of condom use, substance use, and exchanging sex for money or drugs. However, individuals engaging in high- risk behaviors typically did not perceive themselves at risk for contracting HIV. In our sample, many individuals never used condoms and a large proportion of the others only used condoms infrequently; only 11% of individuals reported always using condoms. Many also reported more than five sexual partners, and individuals reporting more than five sexual partners were more likely to perceive their HIV risk as zero or low and less likely to use condoms. Interestingly, having more than five sexual partners was not associated with increased risk of HIV infection in this population. This may be attributable to the fact that data on the number of sexual partners had limited specificity because of the way number of sexual partners were categorized during data collection (0, less than 5, and more than 5); this limited our ability to ana- lyze the impact of incremental increases in sexual partners on newly diagnosed HIV infection.
 
It is noteworthy that cocaine use and having a same sex partner were the strongest predictors of testing positive for HIV in this urban population. These findings suggest that sexual networks among drug users and MSM may be equally or more important than numbers of sexual partners in predicting HIV infection in this population. This finding is similar to those in other recent studies among African American populations with high infection rates; for example, an analysis in nearby Washington, D.C. found that nontraditional sexual risk factors such as sexual networks and disclosure of sexual preference among MSM may contribute to elevated HIV rates among African American men.26 These trends suggest that in addition to the provision of routine testing, primary prevention programs in Philadelphia should focus on sexual networks, particularly those of MSM and illicit drug users.
 
Individuals testing HIV positive dramatically underestimated their own HIV risk; two thirds of individuals who tested HIV positive believed they were at zero or low risk for contracting HIV. Additionally, the overwhelming majority of individuals who tested HIV positive did not report consistent condom use. In spite of these reported high-risk behaviors, many individuals undergoing rapid testing for HIV in Philadelphia believed they were at low risk for contracting HIV. Wide discrepancies existed between individuals' self-perception of HIV risk and medical assistants' perceptions of HIV risks; individuals who perceived their HIV risks as zero or low were frequently at high risk for contracting HIV. The exception to this finding was the small group of individuals who perceived themselves at moderate or high risk; these individuals were also perceived as high risk by medical assistants. Individuals reporting a previous STI diagnosis were more likely to perceive their HIV-risk to be low, suggesting that individuals in this population frequently engage in high- risk sexual behaviors but do not fully appreciate the connection between their behaviors and risks of acquiring HIV in- fection. Low perceived HIV risks in this population may explain low rates of condom use; in turn, low rates of condom use may potentiate the spread of HIV infection in this population.
 
Taken together, our findings suggest that self-perceived HIV risk is an insensitive criterion for HIV screening in this urban, predominately African American population. This is supported by other research that finds low perceived risks among African Americans undergoing testing in high prevalence settings.27 Our findings support the CDC's guidelines for routine opt-out testing for all Americans aged 13-64,27 and underscore the need for routine, opt-out testing in public clinics, which can effectively diagnose individuals who may underestimate their HIV risks. Moreover, because stigma can inhibit persons from seeking HIV testing, routine testing can also help destigmatize HIV testing.
 
Most individuals (72%) testing in Philadelphia's program reported having previously tested for HIV. This is unsur- prising given that the population was overwhelmingly Afri- can American; other research demonstrates that African Americans typically test for HIV more frequently than individuals of other races.14-18 Although the survey did not ask where or when individuals previously tested for HIV, high prevalence of repeat testing and repeat testers' reduced odds of testing HIV-positive suggests that populations at high risk for contracting HIV are taking advantage of public HIV testing programs in Philadelphia.
 
Additionally, although rates of new HIV diagnosis in this population are higher than national averages, city-wide HIV incidence in Philadelphia is 114 per 100,000, the sixth highest of any metropolitan area in the nation.23 While the rapid testing program has reached many high-risk individuals, high city-wide incidence rates suggest that even greater efforts are needed to encourage more individuals to undergo HIV testing in Philadelphia. A successful door-to-door rapid HIV testing program among Latinos in North Carolina may offer important lessons for Philadelphia.28
 
Our findings are subject to several limitations. Because our results are based on a brief questionnaire developed for pro- grammatic rather than research purposes, questions related to several HIV risk behaviors were somewhat limited. For ex- ample, data on route of administration of drug use were not collected, and HIV risk behavior questions inquired about ''ever'' use of drugs rather than questions about recent use. Similarly, because questions about condom use were defined as ''never, sometimes, and always'' and the number of sexual partners in the last year was defined as ''zero, less than 5, or more than 5,'' we cannot estimate independent risks associ- ated with marginal changes in condom use or number of sexual partners. Additionally, because the testing program did not utilize a standardized risk assessment instrument, both participant and medical assistants' risk assessments may be subject to some misclassification of risk category. Because this was a testing program evaluation rather than a research study, we also did not measure changes in risk behaviors, risk perception, and testing frequency longitudinally; it is there- fore difficult to assess whether the testing program increased citywide HIV testing or reduced risk-taking behaviors.
 
Finally, because we do not know who has not presented for testing, it is difficult to know whether these results are generalizable to the broader Philadelphia city population. However, this large sample represents the entire population of individuals undergoing rapid HIV testing during Philadelphia's public HIV testing program over a more than 2-year period.
 
Despite CDC guidelines, most HIV screening in the United States is still based on self-reported HIV risk. Nationwide budget constraints continue to present challenges to implementing routine HIV testing.
To date, CDC guidelines have not been accompanied with expansion of the Ryan White Program, a federal program that finances HIV-related health services in cities, states, and local community-based organizations. Until recently, neither has the US Department of Health and Human Services (HRSA) nor the Centers for Medicare and Medicaid (CMS) financed opt-out testing programs recommended by the CDC. Rather, small programs have been sporadically financed by the CDC or state and city governments, including a new CDC program to expand rapid HIV testing that prioritizes expanding testing for African Americans.29 However, in September 2010, the CDC and Health and Human Services (HHS) announced that the 2010 Affordable Care Act's Prevention and Public Health Fund will allocate $30 million to expand HIV prevention programs outlined in President Obama's National AIDS Strategy, in- cluding $4.4 million for expanding HIV testing.30
 
The Philadelphia rapid HIV testing program experience suggests that routine HIV testing can be scaled and added as a complement to existing services in public clinics that serve urban populations. The Philadelphia experience suggests that routine, publicly financed HIV testing can play an important role in diagnosing HIV among high-risk individuals who do not otherwise have access to health services, particularly those who do not believe they are at risk for becoming infected. Expanding rapid HIV testing in urban areas should be an important public policy priority for achieving President Obama's goal of reducing racial disparities in HIV infection.31
 
 
 
 
  icon paper stack View Older Articles   Back to Top   www.natap.org