Bonney, Loida E. MD, MPH*†; Rose, Jennifer S. PhD†‡; Clarke, Jennifer G. MD, MPH†‡; Hebert, Megan R. MA†; Rosengard, Cynthia PhD, MPH†‡; Stein, Michael MD†‡
SEXUALLY TRANSMITTED INFECTIONS (STIs) impose a significant health burden on society with an estimated 18.9 million new infections per year.1 Behavioral prevention efforts have presented a challenge to the medical and public health communities and have had limited success.2 A vaccine against human papillomavirus (HPV) has recently become available, and vaccines are in development for prevention of herpes simplex virus-2 (HSV-2), Chlamydia trachomatis (CT), and human immunodeficiency virus (HIV).3–6 However, the availability of vaccines will not guarantee success in preventing STIs if at-risk segments of the population, like those who enter the correctional systems,7 do not agree to be immunized.
There is developmental research on a vaccine against Neisseria gonorrhoeae (GC), though no effective vaccine is expected in the next 5–8 years. When considering the acceptability of a putative GC vaccine, attention to the literature on hepatitis B vaccine (HBV) (1 of 2 currently available STI vaccines), and other hypothetical STI vaccine acceptability in high-risk groups may provide insight. Much of this research has been conducted within the framework of the health belief model (HBM) and the theory of reasoned action (TRA).8–13 The HBM is a theory of behavior change, which posits that health behaviors are influenced by perceived susceptibility and severity of an illness, and benefits and barriers to a protective behavior.9 The TRA posits that behaviors are determined by a person's beliefs about the outcomes of the behavior as well as the person's belief about how salient-others view the behavior.8 Other predictors of STI vaccine acceptance have been vaccine-related factors such as efficacy, affordability, and recommendation by a physician.11,14–17 Among adolescents, parental intent to vaccinate and having a sexually experienced friend predicted intent to accept STI vaccines.18,19
It is imperative to understand the associations between personal beliefs and vaccine acceptability among incarcerated populations. In such settings, vaccination programs may have substantial impact due to the high prevalence of unsafe sexual behaviors in this population.20 To date, constructs that are associated with incarcerated women's acceptance of STI vaccines in development have not been explored. In the current study, we hypothesize that acceptance of hypothetical GC vaccine among incarcerated women will correlate with elements of the HBM and TRA among incarcerated women participating in an HIV risk-reduction study.
Materials and Methods
Sample and Procedure
Data for this analysis were collected from February 2004 through May 2005 during the baseline assessment of an ongoing clinical trial at the Women's Facility of the Rhode Island Adult Correctional Institute (ACI), a unified correction system serving as a combined prison and jail. Women entering the ACI, who were awaiting their court appearance, were recruited to participate in a randomized trial of an experimental intervention to reduce alcohol use and HIV risk behaviors. Eligibility criteria included: English speaking, hazardous alcohol use (had 4 or more drinks on 1 occasion 3 or more times in the 3 months before incarceration, or consumed 8 or more drinks per week on average in the 3 months before incarceration), housed in general prison population, age 18 or older, reside in or near the city of Providence, reported sexual activity with male partner(s), and competent to provide written informed consent. HIV risk criteria were determined by a response of “yes” to the question “During the last 3 months, have you had sex without a condom or shared any part of injection drug equipment at least once a month?”
Study approval was obtained from the Miriam Hospital Institutional Review Board, the Office for Human Research Protection and the Medical Research Advisory Group at the ACI before starting the study. A Certificate of Confidentiality was obtained from the federal government to further ensure participant privacy. The warden of the women's facility agreed to help guarantee participant confidentiality and granted permission for all interactions with the women to occur one-on-one with trained female research assistants in unmonitored rooms.
Research assistants reviewed “traffic sheets” (daily printouts of all female inmates committed to the facility) on a daily-basis, Monday (which included weekend traffic) through Friday and attempted to contact each recently detained woman. If a woman was unable to be screened secondary to being in segregation, ill, or in acute withdrawal from drugs and/or alcohol, her status was followed until she was released or could be approached for participation.
Within 4 days of incarceration, women were approached by a research assistant who obtained verbal consent before screening for study eligibility. Following screening, a detailed written informed consent was read aloud to each participant. It was emphasized that study participation was voluntary and would not affect status or services received in the ACI or in the community. Eligible women completed the baseline interview that was conducted within the ACI facility in a confidential setting. Interviews lasted between 45 and 60 minutes on average. Participants were compensated $20 for their time and were paid in the form of a money order at the time of release.
Of the 1402 women who were approached, 1120 agreed to be screened (79.9%). Of the 1120, 979 did not qualify (Fig. 1). Two hundred forty-seven did not meet alcohol nor HIV risk criteria, 363 did not meet alcohol criteria, and 177 did not meet HIV risk criteria. One hundred ninety-two women either did not speak English, were incompetent, lived out of state, were sentenced, did not have reliable contacts, were released before the baseline assessment could be completed, or had participated in the study previously during incarceration. Of the 141 eligible women, 106 participated (75%). Thirty-five women declined participation.
The outcome variable of interest was gonorrhea vaccine acceptability. Using a 12-point linear scale (0 = “I would never get this vaccine” to 11 = “I would definitely get this vaccine”), this variable was assessed based on participant responses to the following question: “Please tell us which number best describes your willingness to receive a vaccine for gonorrhea if a safe and effective vaccine for gonorrhea was available.” “Gonorrhea” was not defined for participants. Intermediate response scale anchors were placed along the numerical continuum at values of 3 (“probably would not”), 5 (“may or may not”), and 8 (“probably would”). For analysis, the responses were dichotomized such that participants who answered that they would probably to definitely get the vaccine (response options ranging from 8 to 11) were considered as “yes” answers and all other responses were coded as “no” answers. The outcome variable was dichotomized because the distributions of responses on the original 12-point scales were extremely nonnormal.
Primary Independent Variables
Primary independent variables consisted of 6 subscales from a gonorrhea vaccine attitudes scale developed for use in this study. Development of the scale was guided by the Health Belief Model and by the Theory of Reasoned Action. A preliminary analysis of the GC vaccine battery used here was conducted based on feedback from interviewers concerning battery length and participants' lack of understanding of some items after data were collected from 37 initial participants. The analysis consisted of an examination of the distributional properties of the items, as well as correlations among items and coefficient alphas for each of the theoretical subscales. Results suggested that 10 items could be deleted due to lack of variability, low correlations (or correlations in an unexpected direction) with other subscale items, and/or markedly improved subscale coefficient alphas when these items were deleted. The final scale consisted of 29 items.
The 6 subscales, derived from the final 29-item scale, assessed perceived severity (5 items, e.g., “If I got infected with GC, it would mean I could get very sick,” α = 0.77), perceived vulnerability (7 items, e.g., “There is a chance that I will be exposed to GC sometime in the future,” α = 0.81), vaccine benefits (2 items, e.g., “Getting a GC vaccine would be a good way for me to stay healthy,” α = 0.77), negative behavioral expectancies (3 items, e.g., “If I got a GC vaccine, I might use condoms less,” α = 0.65), subjective norms (3 items, e.g., “If I knew lots of people were getting the GC vaccine, I would too,” α = 0.58), and vaccine fear (3 items, e.g., “It could be dangerous for me to get vaccinated for GC,” α = 0.63). Response options ranged from 1 = strongly disagree to 6 = strongly agree. Six composite scores were created by averaging the items within each subscale.
Other Independent Variables
Other independent variables included age, race, education, health insurance status, history of an STI (categorized as none, positive history of an STI other than GC, and positive history of GC), health status (a single item measuring perceived general health on a 5-point response scale ranging from 1 = poor to 5 = excellent), and self-reported history of hepatitis B vaccination receipt.
Logistic regression analysis was used to test the relations of demographic characteristics and the 6 vaccine attitudes composite scores to the dichotomous vaccine acceptance outcome variable. Initially, bivariate relations were tested to obtain unadjusted odds ratios and confidence intervals for each correlate. Next, a multivariate logistic regression was performed that included demographic correlates of age, race, education, and STI history, as well as psychosocial vaccine scale correlates related to vaccine acceptance.
Study participants were primarily white (64%) with a low education level (69% did not complete high school) (Table 1). The average age was 34, and more than half reported that they had health insurance. The majority reported having a positive STI history, of which 31% reported GC. In addition, the majority reported having had a Hepatitis B vaccine in the past. Most women (79%) indicated that they would accept a GC vaccine if it was available. Bivariate analyses indicated that demographic correlates were not significantly related to vaccine acceptance (Table 2). In addition, bivariate analyses indicated that all of the psychosocial correlates, with the exception of negative behavioral expectancies, were significantly related to vaccine acceptance at P <0.05. Results from the bivariate analyses suggested that women who perceived greater GC severity, greater GC vulnerability, and greater benefits of a GC vaccine, were significantly more likely to accept a GC vaccine. Subjective norms also correlated with increased acceptance. Women who had greater fear of a GC vaccine were significantly less likely to accept a GC vaccine.
The multivariate logistic regression analysis included the 5 psychosocial constructs found to be significant in the bivariate analyses (Table 2). Although not significant in the bivariate analyses, age, race, education, and STI history were also included in the multivariate analysis to control for potential confounding by these demographic characteristics. Demographic relationships remained nonsignificant after adjustment. Results indicated that greater perceived severity, greater vulnerability, and lower fear were significant independent correlates of vaccine acceptance.
GC is widespread and its sequelae are grave. Seventy-nine percent of the detained women indicated intent to be vaccinated if a safe and effective vaccine for gonorrhea were available. In support of our hypotheses, certain domains of the HBM and TRA, namely perceived severity, perceived susceptibility and fear, were associated with vaccine acceptability. Age, race, education, history of STI, and general health status were not associated with acceptance of this hypothetical vaccine. Interestingly, the proportion of incarcerated women who would accept hypothetical GC vaccine was similar to the proportion of college students, other inmates, and HSV-2 vaccine trials participants willing to accept HIV (70%), hepatitis B (93%), and HSV-2 (75%) vaccines, respectively, although these surveys included men and women while our study included women only.11,21,22
Race was not a correlate of hypothetical GC vaccine acceptance in our sample of incarcerated women and this is consistent with studies examining male and female college students' acceptability of HIV immunization.16,22 In HSV-2 vaccine trial participants, being Hispanic was associated with greater acceptance of HSV-2 immunization.11
We found no association of acceptability with age or education. However, Zimet et al. surveying an STI clinic mixed gender sample reported younger age and lower level of education as demographic variables significantly predictive of hepatitis B vaccine acceptance.15 A previous study from a Rhode Island mixed gender sample found hepatitis B vaccine acceptance was significantly associated with lower education levels among incarcerated persons.22 The differences between sample compositions may explain why we found no association with age and education.
Incarcerated women who perceived themselves vulnerable to GC were more likely to indicate they would accept a GC vaccine (AOR = 2.85), as predicted by the HBM. Vulnerability to infection has been an important construct in predicting STI vaccine acceptance in other research.18,21 Further studies are needed to reveal the degree to which high-risk detainees view themselves susceptible to gonorrhea.
Increased perceived severity was a significant predictor (AOR = 3.33) of GC vaccine acceptance. Perceived severity of infection was not predictive of acceptance of HSV-2, HIV, nor HPV vaccines in prior studies.11,21,22 History of GC was not reported in the cited articles, but a large proportion of our sample had had GC and experienced the medical risk and discomfort associated with this infection.
Our study found perceived benefits of vaccination not to be a significant correlate of GC vaccine acceptance after multivariate logistic regression. Perceived benefits of vaccination were independently predictive of HSV-2 and were independently predictive of HIV vaccine acceptability in college students.11,22 Wording of the benefit items may account for lack of association of this variable with vaccine acceptance in our study. Participants were presented with the statements “Getting a GC vaccine would be a good way for me to stay healthy” and “One way for me to keep in good health would be to get a gonorrhea vaccine.” Some women may not have perceived themselves as “healthy” or “in a state of good health” thereby driving them to disagree with the statements.
Fear of the vaccine was negatively correlated with the acceptance of hypothetical GC vaccine in our incarcerated women sample. We remain uncertain if study participants fear vaccine side effects, pain from inoculation, or infection with GC from a vaccination; and further studies are needed. Unspecified fear of the vaccine, but not fear of needles, was important in the college student group's refusal of HIV vaccine.16 Understanding and directing messages at vaccine fears will be needed to optimize large-scale vaccination programs.
Clearly, components of the HBM and TRA are not universally salient in predicting acceptability of vaccines against distinct STIs in diverse study groups. Furthermore, there are additional factors that potentially influence decision making. Cost and vaccine efficacy have been significant in studies of hypothetical HPV and HIV vaccines,17,23,24 but our participants were not presented with a cost of the GC vaccine, and their responses were based on presumed availability of an “effective” vaccine for GC without numerically specifying its rate of effectiveness.
The current study has several limitations. First, intent to receive a hypothetical vaccine may not predict actual vaccine acceptance if a GC vaccine were available. In the absence of a ready product, this limitation could not be overcome.
Second, our sample was unique in that it included incarcerated women with heavy drinking patterns before incarceration. The findings may not generalize to nondrinking women, nor to nonincarcerated women, or to men. In addition, 60% of our sample had had a prior STI; populations with less prevalent STI histories may be less accepting. Also, this study is about 1 specific STI. Other STIs might have been viewed differently. We cannot assume that GC acceptability data are generalizable to acceptability of other STI vaccines.
Third, despite the informed consent procedure and our protocol's attention to confidentiality, incarcerated women may feel coerced to give socially appropriate answers. Fourth, we dichotomized the outcome variable because the distribution of responses on the original 12-point scales was extremely non-normal. However, the decision to dichotomize our outcome variable may have led to an overestimate of the acceptance of hypothetical GC vaccine. We chose to assign anyone who would “probably” to “definitely” get the vaccine to the “accept vaccine” category. However, answers other than “definitely” may be less predictive of actual vaccine acceptance.
Finally, “gonorrhea” was not defined for participants. Misconceived notions of cost, treatment, or outcome of infection may have influenced results.
Our results suggest that STI vaccines could be offered in prisons to take advantage of the willingness of high-risk incarcerated women to accept such a prevention intervention. For those who are less certain of their decision to accept vaccination, information regarding the severity of specific STIs and one's susceptibility to an STI may be influential. Intervention efforts in women's jails and prisons can rationally target resources towards educating incarcerated women about vaccine utilization as STI vaccines become available.
1. Weinstock H, Berman S, Cates W. Sexually transmitted diseases among American youth: Incidence and prevalence estimates, 2000. Perspect Sex Reprod Health 2004; 36:6–10.
2. Ethier KA, Kershaw T, Niccolai L, et al. Adolescent women underestimate their susceptibility to sexually transmitted infections. Sex Transm Dis 2003; 79:408–411.
3. Blake M, Wetzler L. Vaccines for gonorrhea: Where are we on the curve? Trends Microbiol 1995; 3:469–474.
4. Koutsky LA, Ault KA, Wheeler CM, et al. A controlled trial of a human papillomavirus type 16 vaccine. N Engl J Med 2002; 347:1645–1651.
5. Eko FO, He Q, Brown T, et al. A novel recombinant multisubunit vaccine against chlamydia. J Immunol 2004; 173:3375–3382.
6. Stratov I, DeRose R, Purcell D, et al. Vaccines and vaccine strategies against HIV. Curr Drug Targets 2004; 5:71–88.
7. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveilance 2004. Atlanta: CDC; 2005.
8. Fishbein D, Ajzen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Don Mills, Ontario: Addison-Wesley; 1975.
9. Janz NK, Becker MH. The health belief model: A decade later. Health Educ Q 1984; 11:1–47.
10. Liau A, Zimet GD. Undergraduates' perception of HIV immunization: Attitudes and behaviours as determining factors. Int J STD AIDS 2000; 11:445–450.
11. Zimet G, Fortenbury J, Fife K, et al. Acceptability of genital herpes immunization: The role of health beliefs and health behaviors. Sex Transm Dis 1997; 24:555–560.
12. Zimmerman R, Tabbarah M, Nowalk MP, et al. Predictors of pneumococcal polysaccharide vaccination among patients at three inner-city neighborhood health centers. Am J Geriatr Pharmacother 2005; 3:149–159.
13. Rhodes S, Hergenrather K. Exploring hepatitis B vaccination acceptance among young men who have sex with men: Facilitators and barriers. Prev Med 2002; 35:128–134.
14. Samoff E, Dunn A, VanDevanter N, et al. Predictors of acceptance of hepatitis B vaccination in an urban sexually transmitted diseases clinic. Sex Transm Dis 2004; 31:415–420.
15. Zimet GD, Kee R, Winston Y, et al. Acceptance of hepatitis B vaccination among adult patients with sexually transmitted diseases. Sex Transm Dis 2001; 28:678–680.
16. Liau A, Zimet GD, Fortenberry JD. Attitudes about human immunodeficiency virus immunization. Sex Transm Dis 1998; 25:76–81.
17. Zimet G, Mays R, Winston Y, et al. Acceptability of human papillomavirus immunization. J Womens Health Gend Based Med 2000; 9: p. 47–50.
18. Zimet G, Perkins S, Sturm L, et al. Predictors of STI vaccine acceptability among parents and their adolescent children. J Adolesc Health 2005; 37:179–186.
19. Mays R, Sturm L, Zimet GD. Parental perspectives on vaccinating children against sexually transmitted infections. Soc Sci Med 2004; 58:1405–1413.
20. Clarke JG, Hebert MR, Rosengard C, et al. Reproductive health care and family planning needs among incarcerated women. Am J Public Health 2006; 96:834–839.
21. Vallabhaneni S, Macalino GE, Reinert SE, et al. Prisoners' attitudes toward hepatitis B vaccination. Prev Med 2004; 38:828–833.
22. Zimet GD, Liau A, Fortenberry JD. Health beliefs and intention to get immunized for HIV. J Adolesc Health 1997; 20:354–359.
23. Webb PM, Zimet GD, Mays RM, et al. HIV immunization: acceptability and anticipated effects on sexual behavior among adolescents. J Adolesc Health 1999; 25:320–322.
24. Liau A, Zimet GD. The acceptability of HIV immunization: examining vaccine characteristics as determining factors. AIDS Care 2001; 13:643–650.