AIDS and Behavior

, Volume 18, Supplement 3, pp 248–255

Hepatitis B Virus (HBV) Infection, Immunity and Susceptibility Among Men Who Have Sex with Men (MSM), Los Angeles County, USA

  • Marc A. Pitasi
  • Trista A. Bingham
  • Ekow Kwa Sey
  • Amanda J. Smith
  • Eyasu H. Teshale
Original Paper

DOI: 10.1007/s10461-013-0670-2

Cite this article as:
Pitasi, M.A., Bingham, T.A., Sey, E.K. et al. AIDS Behav (2014) 18(Suppl 3): 248. doi:10.1007/s10461-013-0670-2

Abstract

Men who have sex with men (MSM) bear a disproportionate burden of hepatitis B virus (HBV) infections. We used serologic data from the National HIV Behavioral Surveillance (NHBS) system to determine the prevalence and correlates of HBV infection, immunization, and susceptibility in a sample of Los Angeles County MSM. Approximately 19 % (95 % CI 15–24 %) had serologic evidence of current or past infection, while 35 % (95 % CI 30–40 %) were susceptible. Compared with the youngest age group, MSM ages 40–49 years had a lower prevalence of immunization (aPR 0.28, 95 % CI 0.17–0.45) and a higher prevalence of infection (aPR 8.53, 95 % CI 3.95–18.4) and susceptibility (aPR 2.02, 95 % CI 1.13–3.63). We also observed poor concordance between self-reported and serologic measures of vaccination. Our results indicate the possibility of missed opportunities to vaccinate MSM. Gaps in implementing existing vaccination strategies must be addressed to increase hepatitis B vaccination coverage for MSM, especially in older age groups.

Keywords

Hepatitis B virus (HBV)Men who have sex with men (MSM)Gay menImmunizationSusceptibility

Introduction

Infection with hepatitis B virus (HBV) is a major cause of acute and chronic liver disease in the United States [1]. Despite a continued decline in the incidence of acute HBV infection during the past decade [2], an estimated 804,000–1.5 million people are infected with HBV in the US, and 35,000 new infections were estimated in 2010 [2]. Accordingly, chronic HBV infection and its sequelae remain a critical public health problem [3], contributing to an estimated 2,000–4,000 deaths and $1 billion in hospitalizations each year in the US [4]. Immunization is the most effective way to prevent HBV transmission [5]. In 1982, the Advisory Committee on Immunization Practices (ACIP) recommended the first US-licensed hepatitis B vaccine for targeted immunization of adults with risk behaviors for HBV transmission, such as men who have sex with men (MSM) [6]. In 1991, ACIP recommended universal vaccination of all infants [7]; these guidelines were expanded in 1997 to cover all unvaccinated infants and children 18 years or younger [8]. Because the incidence of acute hepatitis B remains highest in adults with specific behavioral risk factors, ACIP recommended in 2006 the universal vaccination of adults attending healthcare settings serving high-risk adults, such as HIV or STD clinics, substance use treatment centers, correctional facilities, and settings with a high proportion of MSM or injection drug-using clients [9].

Despite the availability of hepatitis B vaccine and the implementation of comprehensive national guidelines, immunization coverage remains suboptimal in certain demographic and risk groups. Integration of hepatitis B vaccine into routine childhood vaccination schedules has dramatically increased immunization coverage among younger age groups, but similar gains in coverage have not been demonstrated among high-risk adults [10], a population that accounts for an estimated 75–95 % of all incident HBV cases in the US [9, 11]. Hepatitis B immunization coverage among adults and high-risk adults is estimated to be 33 and 42 %, respectively [12], indicating missed opportunities in successfully targeting high-risk adult populations, including MSM.

Hepatitis B vaccine efficacy trials relied heavily on MSM study populations because of their historically elevated risk of HBV infection and ease of accessibility [13, 14]. It is therefore an ethical imperative to ensure that MSM populations adequately reap the benefits of the research that they helped enable. Furthermore, gay, bisexual, and other MSM are particularly vulnerable to HBV infection due to the continued high prevalence of HBV infection in the MSM community and the increased risk of transmission associated with unprotected anal sex [9]. While MSM represent only 2 % of the adult population [15], nearly a quarter of all new HBV infections in adults in 2005 were among MSM [9]. Despite this continued disparity in hepatitis B incidence among MSM in the US, there is a paucity of recent data describing the prevalence of hepatitis B infection and immunization in this population. Moreover, many previous investigations have relied on self-reported measures [1619], which may yield prevalence estimates of questionable validity [20] compared with serologic indicators of HBV infection and immunization.

To gain a current understanding of the epidemiology and burden of HBV infection in a sample of MSM in Los Angeles County (LAC), California, we analyzed serologic data collected for the first time during the third round of the National HIV Behavioral Surveillance (NHBS) system. These data provide seroprevalence estimates and correlates of hepatitis B infection, immunization and susceptibility among MSM 18 years and older living in LAC in 2011, and may guide efforts to identify and understand MSM subgroups that may be targeted for HBV immunization efforts.

Methods

Sampling

NHBS is a repeated, cross-sectional survey designed to monitor HIV prevalence, risk behaviors, and use of prevention services among adults at increased risk for HIV infection, such as MSM. NHBS methodology has been described elsewhere in further detail [21]. Eligible participants were asked to complete an anonymous, interviewer-administered questionnaire. All interviewees were offered anonymous HIV testing regardless of their HIV status; those who consented to HIV testing were also offered hepatitis B testing. Blood specimens were collected and transferred to the Los Angeles County Department of Public Health (LACDPH) Public Health Laboratory for hepatitis B surface antibody (HBsAb), hepatitis B surface antigen (HBsAg), and hepatitis B core antibody (HBcAb) testing using chemiluminescent immunoassay (ADVIA Centaur, Bayer Healthcare). Participants were encouraged to retrieve their hepatitis B test results by calling the study team within 1–2 weeks of their screening test. Participants with results indicating active/current infection were referred to their private provider or to a public facility for follow up. Participants who obtained results indicating HBV susceptibility were similarly referred for vaccination.

Measures

The NHBS questionnaire collected information on basic demographic factors as well as self-reported sexual and substance-use behaviors occurring in the 12 months preceding the survey. Demographic items included age, race, ethnicity, country of birth, and education; behavioral items included disclosure of same-sex attraction or behavior to a healthcare provider and history of injection drug use. In addition, the questionnaire assessed participants’ access to and use of health services (e.g., insurance status and type, usual source of healthcare) and other clinical variables, such as self-reported HIV status, STD diagnoses in the past 12 months, and self-reported hepatitis B immunization status.

Outcome Variables

HBV infection, hepatitis B immunization, and susceptibility to HBV were the primary outcome variables for this analysis. HBV infection included all participants with a positive HBcAb test result, which we used as a proxy indicator of current or past infection. We did not conduct immunoglobulin M (IgM) testing, which would have differentiated between acute and chronic HBV infection for participants testing positive for HBcAb. HBV immunization was defined as a positive (≥10 mIU/mL) HBsAb result in combination with negative HBcAb and HBsAg results. Susceptibility to HBV infection was defined as negative test results on all three (HBsAb, HBsAg, and HBcAb) markers. We present this classification, as well as serologic results from our sample, in Table 1.
Table 1

Hepatitis B serologic outcomes among 345 MSM in Los Angeles County, 2011

Serologic results

Interpretation

Outcomea

Count

%

HBsAg−, HBcAb−, HBsAb−

Susceptible

Susceptible

121

35.1

HBsAg−, HBcAb−, HBsAb+

Immune from vaccination

Immunized

157

45.5

HBsAg−, HBcAb+, HBsAb+

Immune from natural infection

Infected

55

15.9

HBsAg+, HBcAb+, HBsAb−

Current infection (chronic or acute)

Infected

5

1.5

HBsAg−, HBcAb+, HBsAb−

Unclear. May be resolved or resolving infection, low-level chronic infection, or false-positive HBcAb test result

Infected

7

2.0

MSM men who have sex with men

aDenotes outcome classification used in regression analyses

Analysis

We limited all analyses to men who reported at least one male sex partner in the past 12 months for whom complete HBV serology (HBsAb, HBsAg, and HBcAb) test results were available. Frequencies of hepatitis B infection, immunization, and susceptibility are reported overall and by selected covariates. We present unweighted estimates that do not account for the potential clustering within sampling venues that may occur with time-location sampling. We calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and corresponding 95 % confidence intervals (CI) of the self-reported measure of hepatitis B vaccination, using serologic test results as the “gold standard” indicator of vaccination. Respondents who did not know their vaccination status (n = 39) were excluded from these analyses. We then used log-binomial regression to estimate measures of association as crude and adjusted prevalence ratios (PR) and 95 % CI. We selected a predefined set of covariates for our adjusted models based on a priori knowledge. Since our sample sizes were relatively small, and because log-binomial models are prone to numerical instability and convergence failures [22], we removed covariates from certain models to facilitate convergence. Decisions to remove specific covariates were based on the relative effect of each covariate’s removal on standard convergence, relative Hessian convergence, and the regression coefficients of the remaining covariates in the model. Covariates were removed if (1) they were the primary contributors to failed or questionable convergence as defined by standard and relative Hessian convergence criteria exceeding the limits of 0.0001, and (2) removal led to no more than 10 % change in the regression coefficients of the remaining covariates. Critical values for convergence criteria and percent change were determined a priori. After this process, we ranked the remaining converged models for each outcome according to corrected Akaike information criteria (AICc) values and the model with the lowest value was selected as the “best” model [23]. In the event that the lowest-ranked models had similar AICc values (i.e., ∆i ≤ 2 [24]; this criterion was determined post hoc), we chose the model containing the higher number of covariates to prioritize bias reduction over precision. All analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC, USA).

Results

Participant Characteristics

A total of 535 men completed the questionnaire. Of the 421 participants who were recruited when a study phlebotomist was available, 410 (97 %) accepted HIV testing and were offered hepatitis B testing. Three hundred sixty-five (89 %) of these eligible men consented to hepatitis B testing; however, an additional 20 men were excluded from all analyses due to missing or invalid lab results (n = 9) or because they reported no male sex partners in the 12 months preceding the survey (n = 11). Table 2 displays the demographic, behavioral, and clinical characteristics of the study population. Nearly half (49 %, n = 169) were under the age of 30 years. Participants were predominantly Latino (39 %, n = 135), white (34 %, n = 117), and born in the US (82 %, n = 281). A majority had received at least some college education (78 %, n = 268) and disclosed attraction to or sex with men to their healthcare provider (84 %, n = 288). Although most participants had a usual source of healthcare (78 %, n = 270), 42 % (n = 144) did not currently have health insurance. Eighteen percent (95 % CI 14–22 %) disclosed a positive HIV status, 38 % (95 % CI 33–43 %) reported never receiving a dose of HBV vaccine, 12 % (95 % CI 8–15 %) reported diagnosis with an STD (i.e., either syphilis, gonorrhea or Chlamydia) in the past 12 months, and 7 % (95 % CI 4–10 %) reported ever injecting drugs. Nearly 22 % (95 % CI 18–26 %) of our sample tested HIV positive at the interview and 1 % (95 % CI 0–2 %) had evidence of HIV/HBV co-infection (data not shown in table).
Table 2

Hepatitis B virus infection, immunity, and susceptibility by demographic, behavioral, and clinical characteristics among 345 MSM in Los Angeles County, 2011

 

Total no. (%)

Infected no.a (%)

Immunized no.b (%)

Susceptible no.c (%)

Total

345 (100.0)

67 (19.4)

157 (45.5)

121 (35.1)

Age (years)

 18–24

85 (24.6)

1 (1.2)

67 (78.8)

17 (20.0)

 25–29

84 (24.4)

7 (8.3)

42 (50.0)

35 (41.7)

 30–39

80 (23.2)

18 (22.5)

25 (31.3)

37 (46.3)

 40–49

63 (18.3)

27 (42.9)

15 (23.8)

21 (33.3)

 50+

33 (9.6)

14 (42.4)

8 (24.2)

11 (33.3)

Race/ethnicity

 White

117 (33.9)

29 (24.8)

42 (35.9)

46 (39.3)

 Black

58 (16.8)

13 (22.4)

30 (51.7)

15 (25.9)

 Latino

135 (39.1)

17 (12.6)

67 (49.6)

51 (37.8)

 Asian/Pacific Islander

25 (7.3)

5 (20.0)

14 (56.0)

6 (24.0)

 American Indian/Alaska native

2 (0.6)

2 (100.0)

0 (0.0)

0 (0.0)

 Multiracial

7 (2.0)

0 (0.0)

4 (57.1)

3 (42.9)

US born

 Yesd

281 (81.5)

54 (19.2)

125 (44.5)

102 (36.3)

 No

63 (18.3)

13 (20.6)

31 (49.2)

19 (30.2)

Education

 High school or below

77 (22.3)

13 (16.9)

35 (45.5)

29 (37.7)

 Some college

119 (34.5)

20 (16.8)

61 (51.3)

38 (31.9)

 Bachelor’s degree or above

149 (43.2)

34 (22.8)

61 (40.9)

54 (36.2)

Out to healthcare providere

 Yes

288 (83.5)

59 (20.5)

128 (44.4)

101 (35.1)

 No

57 (16.5)

8 (14.0)

29 (50.9)

20 (35.1)

Health insurance

 None

144 (41.7)

25 (17.4)

65 (45.1)

54 (37.5)

 Private

165 (47.8)

28 (17.0)

75 (45.5)

62 (37.6)

 Public

35 (10.1)

14 (40.0)

16 (45.7)

5 (14.3)

Usual source of healthcare

 Yes

270 (78.3)

55 (20.4)

125 (46.3)

90 (33.3)

 No

75 (21.7)

12 (16.0)

32 (42.7)

31 (41.3)

Self-reported HIV status

 Negative/unknown

283 (82.0)

45 (15.9)

129 (45.6)

109 (38.5)

 Positive

62 (18.0)

22 (35.5)

28 (45.2)

12 (19.4)

Self-reported hepatitis B vaccination

 Not vaccinated

131 (38.0)

31 (23.7)

42 (32.1)

58 (44.3)

 Vaccinated

175 (50.7)

32 (18.3)

97 (55.4)

46 (26.3)

 Don’t know

39 (11.3)

4 (10.3)

18 (46.2)

17 (43.6)

STD in past 12 monthsf

 No

304 (88.1)

61 (20.1)

139 (45.7)

104 (34.2)

 Yes

41 (11.9)

6 (14.6)

18 (43.9)

17 (41.5)

Ever injected drugs

 No

320 (92.8)

56 (17.5)

150 (46.9)

114 (35.6)

 Yes

25 (7.3)

11 (44.0)

7 (28.0)

7 (28.0)

MSM men who have sex with men, HIV human immunodeficiency virus, STD sexually transmitted disease

aDefined as ever infected with HBV according to positive hepatitis B core antibody

bDefined as positive (≥10 mIU/mL) hepatitis B surface antibody in combination with negative hepatitis B core antibody and hepatitis B surface antigen

cDefined as negative hepatitis B core antibody, negative hepatitis B surface antibody, and negative hepatitis B surface antigen

dExcludes Puerto Rico

eDefined as disclosure of same-sex attraction or behavior to healthcare provider

fIncludes syphilis, gonorrhea, and Chlamydia only

HBV Infection

Overall, 67 (19 %) participants (95 % CI 15–24 %) had positive HBcAb test results, indicating that HBV infection had occurred. Most (88 %, n = 59) of these men were age 30 years or older, and 33 % (n = 22) reported infection with HIV. Older age and self-reported HIV-positive status were independently associated with a positive HBcAb test result after adjusting for race, education, disclosure of same-sex attraction or behavior to healthcare provider, source of healthcare, and injection drug use (Table 3). Current HBV infection (defined as positive HBsAg and HBcAb) was detected in 5 (1.5 %, 95 % CI 0.2–2.7 %) participants. All except one were 32 years of age or older.
Table 3

Characteristics associated with hepatitis B virus infection, immunity, and susceptibility in multivariate regression analyses among 345 MSM in Los Angeles County, 2011

 

Infecteda

Immunizedb

Susceptiblec

PR (95 % CI)

aPR (95 % CI)

PR (95 % CI)

aPR (95 % CI)

PR (95 % CI)

aPR (95 % CI)

Age (years)

 18–24

Ref.

Ref.d

Ref.

Ref.

Ref.

Ref.

 25–29

7.08 (0.89, 56.3)

0.63 (0.50, 0.81)

0.59 (0.45, 0.77)

2.08 (1.27, 3.42)

2.31 (1.38, 3.87)

 30–39

19.1 (2.61, 140)

4.22 (1.90, 9.34)

0.40 (0.28, 0.56)

0.37 (0.26, 0.54)

2.31 (1.42, 3.76)

2.50 (1.51, 4.15)

 40–49

36.4 (5.08, 261)

8.53 (3.95, 18.4)

0.30 (0.19, 0.48)

0.28 (0.17, 0.45)

1.67 (0.96, 2.89)

2.02 (1.13, 3.63)

 50+

36.1 (4.94, 263)

7.12 (3.17, 16.0)

0.31 (0.17, 0.57)

0.32 (0.17, 0.59)

1.67 (0.88, 3.17)

1.83 (0.89, 3.76)

Race/ethnicitye

 White

Ref.

Ref.

Ref.

Ref.

Ref.

Ref.

 Black

0.90 (0.51, 1.60)

1.15 (0.66, 2.02)

1.44 (1.02, 2.04)

1.00 (0.72, 1.39)

0.66 (0.40, 1.07)

0.71 (0.44, 1.15)

 Latino

0.51 (0.29, 0.88)

0.65 (0.38, 1.10)

1.38 (1.03, 1.86)

0.90 (0.65, 1.24)

0.96 (0.70, 1.31)

0.95 (0.68, 1.31)

 Asian/Pacific Islander

0.81 (0.35, 1.88)

1.20 (0.53, 2.75)

1.56 (1.02, 2.38)

0.99 (0.67, 1.48)

0.61 (0.29, 1.27)

 

US born

 Yesf

Ref.

Ref.

Ref.

Ref.

Ref.

 No

1.07 (0.63, 1.84)

1.11 (0.83, 1.47)

1.27 (0.90, 1.78)

0.83 (0.55, 1.25)

0.83 (0.55, 1.25)

Education

 High school or below

Ref.

Ref.

Ref.

Ref.

Ref.

 Some college

1.00 (0.53, 1.88)

1.22 (0.64, 2.32)

1.13 (0.83, 1.52)

0.85 (0.57, 1.25)

0.81 (0.55, 1.18)

 Bachelor’s degree or above

1.35 (0.76, 2.41)

1.00 (0.56, 1.80)

0.90 (0.66, 1.23)

0.96 (0.67, 1.38)

0.82 (0.57, 1.20)

Out to healthcare providerg

 Yes

Ref.

Ref.

Ref.

Ref.

Ref.

Ref.

 No

0.69 (0.35, 1.35)

1.11 (0.56, 2.20)

1.14 (0.86, 1.52)

0.89 (0.64, 1.23)

1.00 (0.68, 1.47)

1.03 (0.70, 1.49)

Health insurance

 None

Ref.

Ref.

Ref.

Ref.

Ref.

 Private

0.98 (0.60, 1.60)

1.01 (0.79, 1.29)

1.00 (0.77, 1.29)

1.00 (0.75, 1.34)

1.06 (0.76, 1.49)

 Public

2.30 (1.34, 3.95)

1.01 (0.68, 1.52)

1.01 (0.67, 1.53)

0.38 (0.16, 0.88)

0.53 (0.23, 1.25)

Usual source of healthcare

 Yes

Ref.

Ref.

Ref.

Ref.

Ref.

Ref.

 No

0.79 (0.44, 1.39)

1.32 (0.75, 2.34)

0.92 (0.69, 1.23)

0.86 (0.62, 1.20)

1.24 (0.90, 1.70)

1.06 (0.75, 1.50)

Self-reported HIV status

 Negative/unknown

Ref.

Ref.

Ref.

Ref.

Ref.

 Positive

2.23 (1.45, 3.43)

1.61 (1.00, 2.61)

0.99 (0.73, 1.34)

0.50 (0.30, 0.85)

0.62 (0.36, 1.07)

Ever injected drugs

 No

Ref.

Ref.

Ref.

Ref.

Ref.

Ref.

 Yes

2.51 (1.52, 4.15)

1.31 (0.74, 2.34)

0.60 (0.32, 1.13)

0.94 (0.50, 1.77)

0.79 (0.41, 1.50)

0.91 (0.47, 1.76)

MSM men who have sex with men, PR prevalence ratio, CI confidence interval, aPR adjusted prevalence ratio, Ref. reference group, HIV human immunodeficiency virus, STD sexually transmitted disease

aDefined as positive hepatitis B core antibody

bDefined as positive (≥10 mIU/mL) hepatitis B surface antibody in combination with negative hepatitis B core antibody and hepatitis B surface antigen

cDefined as negative hepatitis B core antibody, negative hepatitis B surface antibody, and negative hepatitis B surface antigen

dReference group is 18–29 years due to small count of young infected MSM

eAmerican Indian/Alaska native and multiracial participants excluded from all multivariate models and all bivariate race models but included in all other bivariate models

fExcludes Puerto Rico

gDefined as disclosure of same-sex attraction or behavior to healthcare provider

Hepatitis B Immunization

One hundred fifty-seven (46 %) participants (95 % CI 40–51 %) had serologic evidence of HBV immunization, 62 % (n = 97) of whom reported receiving one or more doses of hepatitis B vaccine. Of the 175 MSM who reported receiving at least one dose of vaccine, 55 % (n = 97) had serologic evidence of immunization, while 32 % (n = 42) of those who reported never receiving a hepatitis B vaccine had serologic evidence of immunization. Overall, the measure of self-reported hepatitis B vaccination in our sample had a sensitivity and specificity of 69.8 % (95 % CI 62.1–77.4 %) and 53.3 % (95 % CI 45.7–60.9 %) when compared against serologic evidence. This corresponded to positive and negative predictive values of 55.4 % (95 % CI 48.1–62.8 %) and 67.9 % (95 % CI 60.0–75.9 %).

Variables associated with hepatitis B immunization are shown in Table 3. No difference in immunization prevalence was found by self-reported HIV-positive status or STD history (Table 2). Similarly, nearly half of the participants who reported ever receiving a hepatitis B blood test (47 %) or an HIV test (45 %) had evidence of hepatitis B immunization. Only age was independently associated with hepatitis B immunization in multivariate analyses. The prevalence of immunization observed among men ages 18–24 years was three times greater than that observed among men 50 years or older (adjusted PR 0.31, 95 % CI 0.17–0.57).

Susceptibility to HBV Infection

Thirty-five percent (95 % CI 30–40 %) of participants (n = 121) tested negative for all HBV serologic markers and were therefore classified as susceptible to HBV. Older age was associated with HBV susceptibility in both bivariate and multivariate analyses with the highest relative prevalence of susceptibility observed among participants ages 30–39 years compared with those ages 18–24 years (adjusted PR 2.50, 95 % CI 1.51–4.15). Although not statistically significant, participants with public health insurance had lower relative prevalence of HBV susceptibility compared with participants reporting no insurance or private health insurance.

Discussion

Compared with earlier serologic estimates of 9–17 % immunization coverage among young MSM in the US [25, 26], our study demonstrated a dramatically higher prevalence of hepatitis B immunization among all MSM in our sample (46 %) and especially among MSM ages 18–24 years (79 %).We observed a similar overall HBcAb prevalence (19 %) compared to earlier prevalence estimates of 11–21 % [25, 26], although the prevalence of HBcAb positivity in our sample varied from 1 to 42 % with increasing age. These findings underscore the significant public health achievement resulting from universal infant vaccination initiated in 1991 and guidelines for adolescent catch-up immunization adopted in 1997. Nevertheless, our study findings showed that (1) HBcAb prevalence in our study population was roughly four to five times higher than in the general population [3] and (2) more than one-third of the MSM participants were most likely still susceptible to HBV infection.

The results of our investigation suggest that there may be missed opportunities to vaccinate MSM, particularly among those reporting additional indications for hepatitis B vaccination. For example, the prevalence of a serologic marker of immunization among MSM who had additional risk factors such as HIV-positive status or recent STD diagnosis as well as for those who had ever received HIV or hepatitis B testing was nearly identical compared with the total study population. Thus, MSM who likely accessed HIV/STD preventive, diagnostic, or treatment services appeared no more likely to be immunized for hepatitis B than our total sample. This finding highlights the obstacles faced by past and current efforts to implement universal risk-based immunization and emphasizes the need for more successful integration of hepatitis B testing and vaccination services into existing HIV, STD, and other sexual health programs serving MSM. A better understanding of the relative importance of resource allocation, funding limitations, patient characteristics, and provider characteristics as potential barriers to hepatitis B immunization in MSM is crucial to achieving these goals.

Our study is not without limitations. First, NHBS uses a cross-sectional, venue-based sampling design to recruit men from primarily gay-identified public settings. To the extent that men who do not attend these recruitment venues have different hepatitis B prevalence and incidence, our findings may not be generalizable to all MSM in LAC. Second, there is the potential that immunized MSM were misclassified as susceptible. Prior research suggests that 15–45 % of infants and young children immunized against HBV will have low or undetectable HBsAb titers within 5–22 years following immunization [3]. Although these individuals are believed to be protected even after HbsAb titers fall to undetectable levels, such participants may have been misclassified as susceptible, resulting in underestimated hepatitis B immunization prevalence. Third, HBcAb-positive and HBsAb-positive participants categorized as HBV infected may include persons who were infected after vaccination.

In spite of these limitations, the strength of this investigation lies in the serologic assessment of hepatitis B status over studies that rely solely on self-reported measures [1619]. Previous research [20] has shown the limitations of self-reported prevalence measures without validation with serologic data. Moreover, our results show poor concordance between self-reported and serologic measures of hepatitis B vaccination status. The self-reported measure had moderate to low sensitivity (70 %) and specificity (53 %) based on serologic evidence as the gold standard. This indicates that use of the self-reported measure in the absence of serologic data could have potentially resulted in the misclassification of a sizeable proportion of vaccinated MSM as unvaccinated (NPV = 68 %) and, perhaps more importantly, the misclassification of an even larger proportion of unvaccinated MSM as vaccinated (PPV = 55 %). Furthermore, more than 10 % of our sample (n = 39) reported not knowing their vaccination status, adding another layer of doubt to the validity of this self-reported measure. Nevertheless, the results of this validity analysis must be interpreted with caution since we operated under the assumption that serologic markers perfectly indicate true vaccination status, which is not likely the case. Although the HBsAb serologic marker may be limited in its ability to indicate hepatitis B immunity and susceptibility long after vaccination, our overall results provide support for its use above self-reported measures alone. Future studies should therefore strive to implement serologic testing to estimate HBV prevalence and immunization coverage whenever possible, especially given the feasibility shown in our sample in which the vast majority (89 %) consented to screening. Past research [27] has also demonstrated the potential utility of medical records in the validation of self-reported vaccination measures, which may lead to more accurate prevalence estimates and form the basis of a more ideal gold standard for validity testing.

Our study results provide current estimates of hepatitis B infection, immunization, and susceptibility among MSM in LAC and may serve as valuable baseline estimates for use in HBV surveillance and the assessment of future immunization efforts. We found that a significant proportion of MSM in this study did not have evidence of protection from HBV infection. Most of these MSM reported recent encounters with the healthcare system, and many received services from facilities where universal vaccination of high-risk adults is recommended by ACIP. Given the apparent low adherence to ACIP recommendations among private health providers as reported by our sample, we suggest formative research to characterize the differences between public and private immunization practices for MSM. Once these barriers are identified, implementation of targeted programs to increase ACIP adherence among private providers may then contribute to additional reductions of HBV transmission in this high-risk population.

Acknowledgments

The authors would like to acknowledge the assistance and support of Deborah Emlein and Nicole Green of the LACDPH Public Health Laboratory, Douglas M. Frye, and the Epi Scholars Program. We also acknowledge the funding support from Centers for Disease Control and Prevention, Grant No. 1U62/PS000975-01.

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • Marc A. Pitasi
    • 1
  • Trista A. Bingham
    • 2
  • Ekow Kwa Sey
    • 2
  • Amanda J. Smith
    • 3
  • Eyasu H. Teshale
    • 4
  1. 1.Department of Epidemiology, Rollins School of Public HealthEmory UniversityAtlantaUSA
  2. 2.Division of HIV and STD ProgramsLos Angeles County Department of Public HealthLos AngelesUSA
  3. 3.Division of HIV/AIDS Prevention, NCHHSTPCDCAtlantaUSA
  4. 4.Division of Viral Hepatitis, NCHHSTPCDCAtlantaUSA