Prevalence estimates regularly used to direct HIV funding are based on reported positive test results, reflecting the inherent shortcoming of reliance on surveillance registries, which only contain reported results of those who decide to be tested . Yet, inadequate risk assessment by providers, and patients themselves, primarily limits testing coverage  to groups at high risk for infection [e.g., men who have sex with men (MSM) and injecting drug users (IDU)]. National data suggest that HIV infection is shifting towards groups not routinely targeted for testing , which might cause surveillance data to underestimate the burden of HIV and also could partially explain why approximately 25% of HIV-infected adults in the United States are presumed to be unaware of their infection .
By including people not previously tested, population-based HIV serosurveys potentially provide more reliable HIV estimates. A national anonymous serosurvey, conducted as part of the 1999–2002 National Health and Nutrition Examination Survey (NHANES), revealed that approximately 0.37% of the household-based US population aged 18–49 years was infected with HIV . Comparisons with estimates based on voluntary HIV testing were not possible because few jurisdictions have mandated HIV reporting or accompanying surveillance registries. Similarly, inadequate sample size did not enable the calculation of local HIV prevalence estimates.
In New York City (NYC), the prevalence of reported HIV/AIDS infections based on voluntary testing through 2004 was 1.6% among adults aged 18–49 years and 1.2% among the total population , the highest in the nation . Attempts to account statistically for undiagnosed or unreported HIV cases in NYC resulted in an overall HIV seroprevalence estimate in the range 1.3–1.6% based on a locally derived estimated range of 12–29% undiagnosed patients . To assess the accuracy of this HIV prevalence estimation, the NYC Department of Health and Mental Hygiene (DOHMH) examined 2004 data from the NYC Health and Nutrition Examination Survey (HANES), the first-ever community-based examination survey in the United States, and tested stored sera for HIV. A match of survey participants to the NYC DOHMH surveillance registry for all reported HIV/AIDS cases, the HIV/AIDS Registry System (HARS), allowed for estimation of the prevalence of reported and unreported (possibly undiagnosed) HIV infections. This information might identify gaps in our understanding of HIV and provide information for HIV prevention and testing programs.
New York City Health and Nutrition Examination Survey
NYC HANES is a population-based, cross-sectional examination survey based on NHANES protocols and conducted by the NYC DOHMH . A three-stage cluster, household-based sampling design was used to select a representative sample of NYC adults aged ≥ 20 years who were neither homeless nor institutionalized. NYC DOHMH Institutional Review Board approval was granted for all study procedures. Details have been described previously .
Briefly, during June–December 2004, participants from randomly selected households were given a full-length interview, including an audio computer-assisted self-interview for sensitive questions, and underwent a brief examination at one of four clinics. Serum was collected during the examination for specified laboratory testing. After being informed that they would not receive results of any future blinded testing, participants were asked to provide additional serum that was archived. The study population of interest included NYC HANES participants who consented and provided blood for the repository.
Selected demographic variables included sex, age, race/ethnicity, education, marital status, origin of birth, and sexual orientation. Self-perceived risk of a current sexually transmitted disease (STD) or HIV infection was grouped as high/medium or low/none. Herpes simplex virus-2 (HSV-2) serostatus was identified by testing for HSV-2 antibody.
Risk behaviors of interest included a history of vaginal, anal, or oral sex; lifetime number of sex partners; and drug-use history. Behaviors that can increase HIV risk were defined as a history of needle use for drugs (asked of those who reported previous drug use) and having multiple sex partners during the year preceding the interview. Participants with multiple sex partners during the previous year were asked if they most recently engaged in sexual activity with someone other than a main partner. Finally, excluding women who never had a male partner, anyone who reported multiple partners in the past year and any sexual activity during the past month was asked how many times they had sex without using a condom during that month.
One category of transmission risk was ascribed for each NYC HANES participant using the existing HARS surveillance algorithm . Limited to the details collected by NYC HANES, transmission risk was categorized as MSM, IDU, or neither. Men who reported both needle use and sex with other men were categorized as IDU .
Identifying previously reported HIV using Health and Nutrition Examination Survey data and surveillance data
To identify if NYC HANES participants had been reported with HIV before their survey participation, they were matched with data within HARS. Complete matches were categorized on the basis of last name; first name; and month, day, and year of birth. Independent reviewers examined incomplete matches using information from both NYC HANES and HARS (i.e., sex, race, and social security number). Only complete matches or a match agreement by two reviewers were characterized as having a previous HIV/AIDS diagnosis.
Blinded HIV serosurvey for the Health and Nutrition Examination Survey
New, randomly assigned identification numbers (ID2) were generated for each original ID number to delink data from all identifying information. The NYC DOHMH Public Health Laboratory tested the anonymized specimens using a commercially available HIV-1/HIV-2 enzyme immunoassay with confirmatory Western blot analysis. Results were categorized as positive, negative, or inconclusive. Remnant sera from inconclusive specimens were tested for HIV-1 RNA by PCR with a detection limit of 400 copies/ml. Using ID2 numbers, the test results were merged with the dataset of selected NYC HANES variables and HARS match status.
Observations were weighted to represent the total household NYC adult population and to account for survey nonresponse. A propensity-score adjustment  to the original weights corrected for component nonresponse among the HIV-tested sample. Estimates were calculated in Stata version 7.0 (College Station, Texas, USA), allowing for specification of the complex survey sampling to ensure appropriate variance estimates.
The NYC population prevalence of HIV was estimated and prevalence odds ratios (OR) were calculated with 95% confidence intervals (CI) for HIV infection by demographics, health, and behavior. The weighted prevalence of unreported HIV was also estimated as the proportion of HIV-infected persons identified from the NYC HANES serosurvey that did not match a HARS case. Finally, the prevalence of demographic and transmission risk categories of HIV-infected adults identified by NYC HANES was compared with that of adults reported to HARS . Statistical significance for differences in prevalence for bivariable comparisons was determined at α = 0.01 level by using the Pearson χ2 statistic. Estimates with relative standard errors of > 30% were noted as unreliable.
Health and Nutrition Examination Survey response rates
A household-based eligibility screening survey was completed for 3388 of 4026 (84%) randomly selected households. Screening identified 3047 eligible survey participants, of whom 1999 (66%) completed the face-to-face interview and at least one component of the examination, for an overall response rate of 55%. The majority of survey participants [1626 (81%)] consented and provided blood for future testing, resulting in a 45% response rate for the current study.
HIV-related risk behaviors and risk perceptions
Comparable to the 2000 census for NYC (Table 1), the majority of NYC HANES participants were female (53.8%), nonwhite (61.0%), and aged ≥ 40 years (57.3%). Approximately half were born in the United States (52.2%) and married or living with a partner (52.8%).
Most participants (88.7%) self-identified as heterosexual or straight; 9.7% of men reported a history of MSM activity (Table 2). The vast majority (94.2%) reported ever having had sex, but 1 in 10 men (10.8%), compared with 1 in 20 women (4.5%), reported ≥ 50 lifetime partners (P < 0.01). HSV-2 prevalence was twice as high among women (34.8%) than men (18.1%; P < 0.01). Approximately all participants (95.6%) thought they had a low or no risk for having HIV or other STD.
Participation in sexual or needle-use activities that can increase HIV infection risk differed by gender (Table 2). Men (25.1%) and women (15.5%) reported having multiple sex partners during the previous year, of whom approximately twice as many men as women most recently engaged in sex with someone other than a main partner (40.7% versus 23.8%). Men were more likely than women to report a history of needle use (2.5% versus 0.7%; P < 0.01). Overall, 18.1% of this group of NYC adults (321) had engaged in at least one of these risky sexual or needle-use behaviors, of whom 92.2% thought they had a low or no risk of having HIV or an STD.
HIV prevalence and risk perception in New York City
Using the HIV serosurvey results, the overall HIV prevalence among NYC household adults was estimated to be 1.4% (21 individuals; 95% CI, 0.8–2.5); the age-adjusted prevalence was 1.5% (95% CI, 0.8–2.6) (Table 3). Overall, HIV prevalence was significantly higher among persons who perceived their risk for currently having HIV or an STD to be high or medium, compared with those who thought they were at low or no risk (28.6% versus 0.2%; P < 0.01). All of the HIV-infected IDU and MSM reported a high/medium risk of having HIV or an STD, but 26.1% (n = 3) of HIV-infected non-IDU/non-MSM perceived themselves to be at low/no risk for HIV or STD infection.
HIV reporting completeness
Of the 21 adults with HIV infection, one (5.1%) was identified as HIV infected and not reported in HARS (95% CI, 0.7–29.9). Three individuals who matched to a record in HARS repeatedly tested HIV negative. One other HIV-negative person matched to a HARS record dated after participation in NYC HANES. Blood specimens for another three participants, who did not match to any HARS records, resulted in inconclusive HIV status. Viral load testing, possible for only two of the three, did not meet the HIV-1 RNA detection limit.
HIV-infected persons identified by NYC HANES and those reported to NYC HARS through 2004 did not differ significantly in their demographic and transmission risk profile (Fig. 1). A higher proportion of HIV-infected adults in NYC HANES had a transmission risk that was neither MSM nor IDU (57.9%) than did those in HARS (47.4%).
From the first population-based HIV serosurvey in NYC, we estimate that approximately 1.4% of household adults aged ≥ 20 years are infected with HIV, consistent with the 2004 HIV prevalence measure using cases reported to HARS (1.6%). Data also suggest that as few as 5% of HIV infections in the adult household population are undiagnosed. Approximately 18% of NYC adults reported engaging in risky sexual or needle-use activity, but the majority perceived their HIV/STD risk to be minimal. Disparities between risk behaviors and perception indicate the need for a broader approach to HIV education and screening to be more effective in reaching those who are unaware of their risk.
Self-reported risk behavior is often considered unreliable, but participants provided this information using state-of-the-art audio computer-assisted self-interviews, which has been shown to elicit more accurate information about risk behaviors than face-to-face or telephone interviews . The high proportion of participants who reported both engaging in behaviors that can increase their HIV risk and perceiving themselves to be at low risk of infection suggests that adults continue to be complacent about protecting themselves and seeking testing [13,14]. Studies elsewhere report that patients delay testing or, among those who do get tested, often do not return for their test results, resulting in HIV infections that have progressed to AIDS by the time they are diagnosed . Additionally, the hesitancy of physicians to take lengthy sexual histories can affect testing rates . Routine HIV testing is therefore recommended, in combination with improved prevention and education messages [17–19], to reduce the proportion of undiagnosed infections. The use of rapid HIV testing can also ensure timely diagnosis and treatment of an infection at the time of testing  in order to limit further progression and transmission [21,22].
The majority of HIV-infected persons in both NYC HANES and HARS did not have a reported transmission risk of either MSM or IDU. In NYC, heterosexual behavior is now the second highest reported transmission risk among newly diagnosed persons . Nationally, the proportion of new HIV infections from heterosexual contact has increased faster than other modes of exposure . The increasingly important role of heterosexual transmission in the persistence of the pandemic supports recent Centers for Disease Control and Prevention recommendations calling for routine HIV screening, irrespective of traditional risk-group identification [17,24,25].
Differences in demographic and health characteristics between NYC and the United States as a whole help to explain the observed differences in HIV prevalence (1.4% versus 0.37%). In NYC, 13.7% of men aged 18–64 years reported having sex with men during the past year , whereas only 4% of men in the United States reported having sex with another man during adulthood . The higher proportion of ethnic and racial minorities in NYC, who continue to be disproportionately affected by the HIV pandemic , also contributes to higher overall HIV rates. Additionally, HSV-2 infection, which has been shown to double the risk for HIV infection , is more prevalent in NYC than nationally  and is similarly concentrated among black and Hispanic adults. The higher HSV-2 prevalence might reflect more widespread high-risk sexual behavior in NYC or the role of HSV-2 infection in facilitating HIV transmission.
This study had certain limitations. HIV testing and HARS matching were imperfect, suggesting that our estimates might be a conservative representation of the burden of HIV in NYC. Our findings are limited to household adults and not further generalizable to homeless and institutionalized persons (both of whom might have a higher prevalence of diagnosed and undiagnosed infections [30–35]), nor to persons who were unable to participate because of poor health or inability to visit the local health clinic. Consequently, our estimate of 5% unreported HIV, compared with the national estimate of 25%, might be explained in part by the lack of participation or eligibility of groups who were more likely to be infected or less likely to seek testing and medical care. Nonetheless, undiagnosed HIV among NYC's household adult population might be low, owing in part to aggressive, local efforts to increase testing.
The majority of associations were in the expected direction , but sample sizes were too small to test for significant risk factors or differences between subgroups. Low response rates to NYC HANES might introduce bias, although weighting adjusted for nonresponse. Also, HARS is limited by New York State HIV/AIDS reporting guidelines, which did not require named HIV reporting until 2000  and all CD4 cell count and viral load results until June 2005. Therefore, our estimate of unreported HIV might include persons who know of their infection but were never reported to HARS or did not seek HIV care.
Our findings provide evidence that local surveillance-based prevalence measures are relatively reliable estimates among NYC household adults. This conclusion might not be valid for other populations that engage less in voluntary testing. Additional community-level HIV prevalence studies are needed to address local and subgroup differences in testing and prevention needs. The NYC HANES serosurvey which, to our knowledge represents the first community-wide serosurvey in the United States suggests that measurable discordance between risk behaviors and perceptions of minimal infection risk exist, demonstrating gaps in traditional risk-based testing approaches.
We gratefully acknowledge the NYC HANES participants and staff who made this study possible. We also appreciate the support of the NYC DOHMH Public Health Laboratory, especially Lisa Forgione and Drs Amado Punsalong and William Oleszko.
Note: The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
1. Brookmeyer R, Yasui Y. Statistical analysis of passive surveillance disease registry data. Biometrics 1995; 51:831–842.
2. Beckwith CG, Flanigan TP, del Rio C, Simmons E, Wing EJ, Carpenter CCJ, et al
. It is time to implement routine, not risk-based, HIV testing. Clin Infect Dis 2005; 40:1037–1040.
3. Kellerman S, Begley E, Boyett B, Clark H, Schulden J. Changes in HIV and AIDS in the United States: entering the third decade. Curr HIV/AIDS Rep 2004; 1:153–158.
4. Glynn M, Rhodes P. Estimated HIV prevalence in the United States at the end of 2003. National HIV Prevention Conference
. Atlanta, June 2005 [abstract T1-B1101].
5. McQuillan GM, Kruszon-Moran D, Kottiri BJ, Kamimoto LA, Lam L, Cowart MF, et al
. Prevalence of HIV in the US household population: the National Health and Nutrition Examination Surveys, 1988 to 2002. J Acquir Immune Defic Syndr 2006; 41:651–656.
6. New York City Department of Health and Mental Hygiene. New York City HIV/AIDS Annual Surveillance Statistics 2004.
New York City: New York City Department of Health and Mental Hygiene; 2005.
7. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report, 2004.
Atlanta, GA: Centers for Disease Control and Prevention; 2005.
8. New York City Department of Health and Mental Hygiene. HIV Prevalence in New York City, 2002: Estimating the Total Number of People Living with HIV and AIDS in NYC.
New York City: New York City Department of Health and Mental Hygiene; 2004.
9. Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey Questionnaire (or Examination Protocol, or Laboratory Protocol)
. Atlanta, GA: Centers for Disease Control and Prevention for the US Department of Health and Human Services; 1999–2000.
10. Thorpe LE, Gwynn RC, Mandel-Ricci J, Roberts S, Tsoi B, Berman L, et al
. Study design and participation rates of the New York City Health and Nutrition Examination Survey, 2004. Prev Chronic Dis 2006; 3:A94.
11. Brookhart MA, Schneeweiss S, Rothman KJ, Glynn RJ, Avorn J, Sturmer T. Variable selection for propensity score models. Am J Epidemiol 2006; 163:1149–1156.
12. Ghanem KG, Hutton HE, Zenilman JM, Zimba R, Erbelding EJ. Audio computer assisted self interview and face to face interview modes in assessing response bias among STD clinic patients. Sex Transm Infect 2005; 81:421–425.
13. Boily MC, Bastos FI, Desai K, Masse B. Changes in the transmission dynamics of the HIV epidemic after the wide-scale use of antiretroviral therapy could explain increases in sexually transmitted infections: results from mathematical models. Sex Transm Dis 2004; 31:100–113.
14. Boily MC, Godin G, Hogben M, Sherr L, Bastos FI. The impact of the transmission dynamics of the HIV/AIDS epidemic on sexual behaviour: a new hypothesis to explain recent increases in risk taking-behaviour among men who have sex with men. Med Hypoth 2005; 65:215–226.
15. Centers for Disease Control and Prevention. Advancing HIV prevention: new strategies for a changing epidemic: United States, 2003. MMWR Morb Mortal Wkly Rep
16. Centers for Disease Control and Prevention. Recommendations for incorporating human immunodeficiency virus (HIV) prevention into the medical care of persons living with HIV. Clin Infect Dis
17. Branson BM, Handsfield HH, Lampe MA, Janssen RS, Taylor AW, Lyss SB, et al
. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep 2006; 55:1–17, quiz CE11-14.
18. Bayer R, Fairchild AL. Changing the paradigm for HIV testing: the end of exceptionalism. N Engl J Med 2006; 355:647–649.
19. Frieden TR, Das-Douglas M, Kellerman SE, Henning KJ. Applying public health principles to the HIV epidemic. N Engl J Med 2005; 353:2397–2402.
20. Franco-Paredes C, Tellez I, del Rio C. Rapid HIV testing: a review of the literature and implications for the clinician. Curr HIV/AIDS Rep 2006; 3:169–175.
21. Pilcher CD, Eron JJ Jr, Galvin S, Gay C, Cohen MS. Acute HIV revisited: new opportunities for treatment and prevention. J Clin Invest 2004; 113:937–945.
22. Pilcher CD, Tien HC, Eron JJ Jr, Vernazza PL, Leu SY, Stewart PW, et al
. Brief but efficient: acute HIV infection and the sexual transmission of HIV. J Infect Dis 2004; 189:1785–1792.
23. Neal JJ, Fleming PL, Green TA, Ward JW. Trends in heterosexually acquired AIDS in the United States, 1988 through 1995. J Acquir Immune Defic Syndr Hum Retrovirol 1997; 14:465–474.
24. Jenkins TC, Gardner EM, Thrun MW, Cohn DL, Burman WJ. Risk-based human immunodeficiency virus (HIV) testing fails to detect the majority of HIV-infected persons in medical care settings. Sex Transm Dis 2006; 33:329–333.
25. Chen Z, Branson B, Ballenger A, Peterman TA. Risk assessment to improve targeting of HIV counseling and testing services for STD clinic patients. Sex Transm Dis 1998; 25:539–543.
26. Manning SE, Thorpe LE, Ramaswamy C, Hajat A, Marx MA, Karpati AM, et al
. Estimation of HIV prevalence, risk factors, and testing frequency among sexually active men who have sex with men, aged 18–64 years: New York City, 2002. J Urban Health 2007; 13:13.
27. Michael RT, Wadsworth J, Feinleib J, Johnson AM, Laumann EO, Wellings K. Private sexual behavior, public opinion, and public health policy related to sexually transmitted diseases: a US–British comparison. Am J Public Health 1998; 88:749–754.
28. Serwadda D, Gray RH, Sewankambo NK, Wabwire-Mangen F, Chen MZ, Quinn TC, et al
. Human immunodeficiency virus acquisition associated with genital ulcer disease and herpes simplex virus type 2 infection: a nested case-control study in Rakai, Uganda. J Infect Dis 2003; 188:1492–1497.
29. Xu F, Sternberg MR, Kottiri BJ, McQuillan GM, Lee FK, Nahmias AJ, et al
. Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States. JAMA 2006; 296:964–973.
30. Hammett TM, Harmon MP, Rhodes W. The burden of infectious disease among inmates of and releasees from US correctional facilities, 1997. Am J Public Health 2002; 92:1789–1794.
31. Kerker B, Bainbridge J, Li W, Kennedy J, Bennani Y, Agerton T, et al. The Health of Homeless Adults in New York City: A Report from the New York City Department of Health and Mental Hygiene and Homeless Services, 2005.
New York: New York City Department of Health and Mental Hygiene; 2005.
32. Bucher JB, Thomas KM, Guzman D, Riley E, Dela Cruz N, Bangsberg DR. Community-based rapid HIV testing in homeless and marginally housed adults in San Francisco. HIV Med 2007; 8:28–31.
33. Herndon B, Asch SM, Kilbourne AM, Wang M, Lee M, Wenzel SL, et al
. Prevalence and predictors of HIV testing among a probability sample of homeless women in Los Angeles County. Public Health Rep 2003; 118:261–269.
34. Altice FL, Marinovich A, Khoshnood K, Blankenship KM, Springer SA, Selwyn PA. Correlates of HIV infection among incarcerated women: implications for improving detection of HIV infection. J Urban Health 2005; 82:312–326.
35. Sabin KM, Frey RL Jr, Horsley R, Greby SM. Characteristics and trends of newly identified HIV infections among incarcerated populations: CDC HIV voluntary counseling, testing, and referral system, 1992–1998. J Urban Health 2001; 78:241–255.
36. Centers for Disease Control and Prevention. Implementation of named HIV reporting: New York City, 2001. MMWR Morb Mortal Wkly Rep