Knowledge of cervical cancer prevention and human papillomavirus among women with HIV

Results Among 1588 women, HIV seropositive women better understood facts about cervical cancer prevention and HPV than seronegative women, but both had substantial knowledge deficits. Almost all women considered Pap testing important, although 53% of HIV seropositive and 48% of seronegative women considered cervical cancer not preventable ( P = 0.21). Only 44% of HIV seropositive women knew Paps assess the cervix, versus 42% of HIV seronegative women ( P = 0.57). Both groups understood that HPV causes genital warts and cervical cancer (67% of HIV seropositive vs. 55% of seronegative women, P = 0.002). About half of both groups considered HPV vaccination extremely important for cervical cancer prevention. HIV seronegative women were more likely to report learning of HPV vaccination through advertising than from clinicians (81% vs. 64%, P < 0.0001). Conclusion High risk women need effective education about cervical cancer prevention, HPV, and HPV vaccination. Keywords HPV Cervical cancer prevention Pap test Health education HIV in women Introduction Women infected with the human immunodeficiency virus (HIV) have high rates of coinfection with human papillomavirus (HPV) [1] . Persistent infection with carcinogenic types of HPV can lead to the development of cervical intraepithelial lesions (CIN) and cervical cancer, and women with HIV face a high risk of abnormal Pap test results and CIN [2-4] . Population based registry studies have shown that women with HIV are at higher risk for invasive cervical cancer than HIV-uninfected women [5,6] , though their risk approaches that of the general population when they participate in regular cervical cancer screening and prevention programs [7] . Such participation may be enhanced when women consider themselves at risk for cervical cancer and when they understand the course of HPV infection and the cervical cancer prevention process. Understanding cervical oncogenesis can be difficult, since it involves multistep carcinogenesis, beginning with sexual acquisition of HPV infection, failure of immune-mediated HPV clearance, and the progression of preinvasive lesions to cancer. The mechanics of cervical cancer prevention can be similarly confusing, requiring an often arduous program of cytologic screening, colposcopy triage, and treatment. Among women with HIV, failure rates for treatment of cancer precursors are high [8] . For these women, prevention may involve rounds of cytology, colposcopy, and treatment, with multiple opportunities for discouragement and default that may allow cancer precursors to progress. Some 35% of women with HIV default from colposcopy referral [9] . In other populations, educational interventions to address misunderstandings about cervical cancer prevention have improved compliance with follow-up [10-13] . These have not been tested in HIV-infected individuals, and understanding what HIV infected women know about cervical cancer and what contributes to misunderstanding might help guide effective interventions. Among U.S. women, knowledge of HPV and its consequences is quite limited [14-21] . Women least likely to know about HPV and its relationship to cervical cancer are those from lower socio-economic strata, those with lower educational attainment, and those who do not obtain regular Pap testing [15,16,19] . These in turn are risks for cervical cancer [22] . Despite the particular threat of cervical cancer for women with HIV, little is known about what HIV-infected women understand about HPV and cervical disease. To provide a more complete understanding, we administered a questionnaire to women with HIV and to comparison women uninfected with HIV. We inquired about their knowledge of HPV, HPV vaccination, and the cervical cancer prevention process. We asked about women's sources for knowledge about HPV vaccination. We attempted to identify characteristics of women who knew little about these areas as a basis for interventions. Methods This investigation was part of the Women's Interagency HIV Study (WIHS), an ongoing multicenter prospective cohort study of the natural history of HIV infection and related health conditions among HIV seropositive women and at-risk HIV uninfected comparison women. The protocols, recruitment processes, procedures, and baseline results of the WIHS have been described [23,24] ; seropositive WIHS participants are representative of U.S. women with HIV [23] . Enrollment began with 2,623 women in 1994 at 6 study consortia (Bronx, Brooklyn, Chicago, Los Angeles, San Francisco, and Washington, D.C.). The cohort was expanded to 3,766 women during 2001–2002 to recruit younger, AIDS-free, and therapy naïve HIV seropositive women, along with HIV-uninfected women with similar socio-demographic and sexual risk characteristics [24] . Comparisons by WIHS administrators to statistics from the Centers for Disease Control and Prevention have shown that the demographics and HIV risk characteristics of the cohort are broadly similar to those of U.S. women with HIV, though WIHS does not include Southern women and so has marginally greater representation of Latinas from the New York and Los Angeles areas than the U.S. population. Adolescents and young women are also underrepresented. Written informed consent was obtained after local human subjects committees approval. Follow up continues, but this analysis reports information from a cross sectional questionnaire on knowledge of and attitudes toward cervical cancer prevention and HPV administered between April and September, 2006. Reading level and a neuropsychological screen for attention and cognitive dysfunction were assessed between October 2004 and September 2005. HIV status was determined by Western blot at study entry for all participants and annually thereafter for those initially seronegative. Ethnicity and years of education were self-reported. The English version of this questionnaire has been previously described [16] ; it was translated into Spanish for this study. Questions asked about HPV, Pap tests, cancer risks, and HPV vaccination. The WIHS National Community Advisory Board reviewed a draft of the questionnaire and provided feedback prior to field implementation. Multiple choice questions and response options were read by participants or to participants by trained interviewers, and responses were recorded. Interviewers were instructed to clarify questions as needed but to defer requests for information until after the questionnaires had been completed. On completion, participants were given written explanations of the correct answers with background, and further information was supplied if requested. Responses to the 44-item questionnaire were tabulated and compared by HIV status using a global chi-square test. Responses then were coded as correct or incorrect where applicable and subjected to a principal component analysis for item reduction. The principal axis method was used to extract the components, and this was followed by a varimax (orthogonal) rotation [25] . A single summary factor-based score was computed for each subject based on the remaining 26 questionnaire items from the principal component analysis (Chronbach's alpha = 0.88). It included items related to knowledge of HPV, risk factors for cervical cancer, the HPV vaccine, and care following abnormal Pap smears. Scores were correlated with demographic variables, including age at questionnaire administration, ethnicity, education attained by study entry, reading level, and household income; medical factors, including HIV serostatus, abnormal Pap history, prior colposcopy, and cervical disease treatment; and measures of attention, depressive symptoms, and reading level as a proxy for educational attainment. To explore links between study responses and general cognition, we used information gathered during the Neurocognition Substudy in WIHS. The Wide Range Achievement Test-Version 3 (WRAT) for English speakers and the Word Accentuation Test for Spanish Speakers [26,27] were used to assess basic academic skills. A cognitive task, the Symbol Digit Modalities Test, was used to assess information processing and attentiveness, including visual scanning and mental and motor speed, and immediate paired recall of the same test was used to assess short-term memory [28] . Clinically significant depressive symptoms were screened for using the Center for Epidemiologic Studies Depression (CES-D) scale, with a cutoff score of 16 considered as positive [29] . Multivariable analysis was carried out with the knowledge factor-based score as the outcome. Linear regression was used to assess characteristics associated with knowledge score. For the initial model, each independent variable was evaluated for fit using the Type III SS value and P -value and were included in the analyses if they had a P -value < 0.05. Raw Symbol Digit and WRAT score were added to subsequent models. Due to minor but potentially confounding connotative differences between English and Spanish speakers, 137 women who completed their questionnaires in Spanish were excluded from multivariable analyses. All final regression models were created using the PROC Generalized Linear Models (GLM) procedure in SAS software [30] . Results Of the 2,091 women seen at WIHS visit 26, a total of 1,597 (76%) completed questionnaires on cervical cancer and HPV, while 156 (7%) did not receive questionnaires, 167 (8%) refused or did not return questionnaires, and 171 (8%) returned substantially incomplete questionnaires. No significant differences were seen between those who were excluded because of missing data from questionnaire and those who were not except for site and age; those missing data were slightly older 44.5 vs. 43.1 ( P = 0.05) and more likely to be from the Washington, Los Angeles, and Chicago sites compared to other sites. Nine additional women were excluded because of HIV seroconversion during the years of study, a group too small for analysis. This left 1588 women for analysis. Women who were excluded were more likely to come from the District of Columbia, Los Angeles, and Chicago sites and were marginally older (44.5 vs. 43.1 years, P = 0.05). There were no differences between included and excluded women by CES-D score; HIV serostatus income; alcohol, smoking and drug use; ethnicity; education level; history of abnormal Pap, or most recent Pap grade. As shown in Table 1 , when compared to HIV seronegative women, HIV seropositive women in our study group were older (median age 43.9 vs. 40.5 years, P < 0.0001 by Wilcoxon two-sample test) and less likely to use alcohol and tobacco currently but more likely to have a history of injected drug use. There were also differences between HIV seropositive and seronegative women by site and income, though not in overall depressive symptoms or reading level achieved. Table 2 shows that HIV seropositive women were more likely than seronegative women to have prior abnormal Paps, more severe abnormalities, and more colposcopies and cervical disease treatments. Tables 3 – 6 present questionnaire results. As shown in Table 3 , a minority of women were knowledgeable about cervical cancer prevention processes. In most cases where differences were significant, HIV seropositive women had a better understanding of facts about cervical cancer prevention. Tables 4 and 5 similarly show that HIV seropositive women understood aspects of HPV infection and vaccination better than HIV seronegative women, although substantial minorities in both groups were unaware of facts concerning HPV and vaccination. Table 6 shows how women differed in their attitudes toward Pap testing and HPV vaccination. Although many women had responded that cervical cancer is not preventable, almost all women surveyed considered Pap testing at least somewhat important. Despite lack of knowledge about many aspects of HPV, about half the women studied considered HPV vaccination extremely important for cervical cancer prevention, and more than half would recommend vaccination to female relatives and friends, though 30% of HIV seropositive women and 37% of seronegative women believed they needed additional information before doing so. We surveyed women to learn where they had received information about HPV vaccination. Among women who had heard about it, 19% had learned of it from doctors, 11% from nurses, 17% from WIHS staff, 61% from news reports, and 68% through advertising, while 8% could not remember their source of information. HIV seronegative women were more likely to report learning of HPV vaccination through advertising than from clinicians (81% vs. 64%, P < 0.0001), but other sources of information did not differ by serostatus. The factor-based score computed using the final 26 items ranged from − 1.98 to 1.78 (median = 0.27), with negative scores showing worse knowledge and increasing positive scores showing greater knowledge. HIV seropositive women had a higher median knowledge score than seronegative women (0.37 vs. − 0.03, Two independent sample t -test P < 0.0001). Results of our main multivariable models are presented in Table 7 , which show correlates of the factor-based score across all knowledge fields. In the first model, better knowledge was associated with being HIV seropositive and of white or other ethnicity, as well as with having more education, higher income, and a prior abnormal Pap, while depressive symptoms were associated with lower knowledge score. The second model included a measure of sustained attention and perceptual speed by adding the Symbol Digit Modalities Test score; this increased the model's predictive value ( R 2 ) and demonstrated that greater attentiveness was linked to better scores, displacing age and Hispanic ethnicity. In the third model, controlling for reading achievement as a proxy for educational attainment in addition to number of years of education by adding the WRAT reading recognition score further improved predictive value, while screening positively for depressive symptoms became nonsignificant. Discussion For many women with HIV and their HIV uninfected peers, knowledge gaps can pose a barrier to engaging in cervical cancer prevention programs. Most women in our study did not know correct answers to questions about several fundamental aspects of cervical cancer prevention, including the concept that Pap testing evaluates the cervix. This result was unanticipated. On the one hand, our participants came predominantly from lower socioeconomic and educational backgrounds [23] , factors that have predicted lower awareness of HPV and cervical cancer prevention processes [15,16,19] . On the other hand, WIHS participants had personal experience of semiannual Pap testing. Most had abnormal Pap results. All had opportunities to learn about HPV and cervical cancer prevention through newsletters, peer education, and staff contact after abnormal Pap results. Women with prior abnormal Pap tests knew more about cervical cancer prevention, but only marginally so. In high-risk populations like ours, unstructured encounter-based education may be insufficient to raise understanding of cervical cancer risks and prevention strategies. Culturally tailored educational interventions designed to improve compliance with screening, treatment, and vaccination among women like those we studied will need to incorporate basic information about genital anatomy and the natural history of cervical disease. Women with less than a high school education have the greatest knowledge deficits and merit particular outreach. Most participants learned about HPV vaccination from advertising and news, not WIHS researchers or clinicians, but the substantial knowledge gaps we found suggest that media may communicate messages incorrectly or incompletely to low-income women of color. The importance of ethnicity, income, and quality of education in predicting knowledge suggests that educational messages should be culturally specific. Research is needed to determine whether more tailored education from clinicians, such as multimedia approaches incorporating visual and auditory aids, might improve women's understanding of cancer prevention and if so whether better understanding leads to better compliance. Our study was novel in incorporating psychometric assessments. These included measures of sustained attention, mental speed, reading as a proxy for education, and depressive symptoms. Multivariable analysis showed that all but depressive symptoms were significant contributors to the level of knowledge about cervical cancer prevention and HPV, and future studies in these areas should incorporate them. Unfortunately, models combining these factors with nominal years of education and proxies for cultural factors like income and ethnicity failed to explain much of the variability in knowledge. Unmeasured factors, such as the perceived reliability of the information source, may be important and deserve further exploration. Nevertheless, some women, such as those who do not know what a cervix, a cell, or a cancer is, may have knowledge deficits that cannot be addressed readily in brief clinical encounters or educational campaigns. In fact such efforts may be counterproductive if exposing knowledge deficits erodes women's self-worth and desire to pursue cancer prevention. For these women, efforts focused on developing trust may be more effective in improving compliance with prevention measures than educational outreach. Appropriately educated HIV seropositive women may make effective peer counselors for women needing such support, as participants frequently indicated that they considered HPV vaccination an important measure against cervical cancer and would recommend vaccination to female relatives and friends. Whether vaccination is safe or effective for HIV-infected women is the subject of ongoing trials. Results from our study were broadly congruent with recent research on knowledge and attitudes regarding cervical cancer prevention, HPV, and HPV vaccination. For example, a recent review found that 8–68% of women asked closed-ended questions could identify the link between HPV and cervical cancer [22] . A focus group study conducted by the Centers for Disease Control in 2002–3 found that women preferred to receive information about HPV from sources that were trustworthy, accessible, convenient, and confidential; while they preferred clinicians as information sources, we found that many of our participants had received their information from media and advertising [31] . Our study was limited by several factors. First, women from similar socioeconomic backgrounds but irregularly screened may have even lower levels of understanding of cervical cancer prevention, HPV, and HPV vaccination than our participants. Second, since this study was nested in a larger study of other health outcomes, restricting time availability, we used multiple choice testing. Knowledge may be lower when measured without prompting [21] and using open-ended questions [20] . Third, because WIHS is a comprehensive study of multiple health outcomes with limited time at each visit, measures of vocabulary and cognitive function were administered at different visits, potentially limiting the strength of correlations. Fourth, because we excluded women who spoke only Spanish from analyses, conclusions may not apply to less acculturated Latina women. Fifth, our findings may not reflect those of young women or those from the South, who are underrepresented in WIHS. Finally, our study was conducted as HPV vaccine marketing was initiated; ongoing marketing of HPV vaccines has likely increased awareness of HPV and cervical cancer prevention [32] , and we recently completed a follow-up survey to assess how knowledge is evolving. In addition to education about cervical cancer prevention processes, HPV education is important because an HPV diagnosis can induce feelings of anxiety, shame, and stigmatization, which actually may be stronger among women who are knowledgeable about HPV [33] . Understanding the near-ubiquity of HPV infection may reduce these reactions [33] . However, improving knowledge may not lead to behavior change. For example, among parents with vaccine-eligible daughters, an HPV education sheet improved knowledge about HPV but did not alter willingness to consider vaccination [34] . We plan follow-up studies to assess the impact of an HPV-related educational intervention on knowledge scores and colposcopy compliance among women with abnormal Pap tests. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgments Data in this manuscript were collected by the Women's Interagency HIV Study (WIHS) Collaborative Study Group with centers (Principal Investigators) at New York City/Bronx Consortium (Kathryn Anastos); Brooklyn, NY (Howard Minkoff); Washington DC Metropolitan Consortium (Mary Young); The Connie Wofsy Study Consortium of Northern California (Ruth Greenblatt); Los Angeles County/Southern California Consortium (Alexandra Levine); Chicago Consortium (Mardge Cohen); Data Coordinating Center (Stephen Gange). The WIHS is funded by the National Institute of Allergy and Infectious Diseases ( UO1-AI-35004 , UO1-AI-31834 , UO1-AI-34994 , UO1-AI-34989 , UO1-AI-34993 , and UO1-AI-42590 ) and by the Eunice Kennedy Shriver National Institute of Child Health and Human Development ( UO1-HD-32632 ). The study is co-funded by the National Cancer Institute, the National Institute on Drug Abuse, and the National Institute on Deafness and Other Communication Disorders. Funding is also provided by the National Center for Research Resources (UCSF-CTSI Grant Number UL1 RR024131 ). H.D. Strickler was supported by NCI R01 CA85178-01 . The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. References [1] H.D. Strickler R.D. Burk M. Fazzari K. Anastos H. Minkoff L.S. Massad Natural history and possible reactivation of human papillomavirus in human immunodeficiency virus (HIV) positive women J. Natl. Cancer Inst. 97 2005 577 586 [2] T.C. Wright T.V. Ellerbrock M.A. Chiasson N. Van Devanter X.W. Sun and the New York Cervical Disease Study Cervical intraepithelial neoplasia in women infected with human immunodeficiency virus: prevalence, risk factors, and validity of Papanicolaou smears Obstet. Gynecol. 84 1994 591 597 [3] L.S. Massad K.A. Riester K.M. Anastos R.G. Fruchter J.M. Palefsky R.D. Burk Prevalence and predictors of squamous cell abnormalities in Papanicolaou smears from women infected with human immunodeficiency virus-1 J. Acquire. Immun. Deficiency Syndromes Hum. Retrovirol. 21 1999 33 41 [4] L.S. Massad L. Ahdieh L. Benning H. Minkoff R.M. Greenblatt H. Watts Evolution of cervical abnormalities among women with HIV-1: evidence from surveillance cytology in the Women's Interagency HIV Study J. Acquir. Immunodef. Human Retrovirol. 27 2001 432 442 [5] M. Frisch R.J. Biggar J.J. Goedert Human papillomavirus-associated cancers in patients with human immunodeficiency virus infection and acquired immunodeficiency syndrome J. Natl. Cancer Inst. 92 2000 1500 1510 [6] B. Gallagher Z. Wang M.J. Schymura A. Kahn E.J. Fordyce Cancer incidence in New York State acquired immunodeficiency syndrome patients Am. J. Epidemiol. 154 2001 544 556 [7] L.S. Massad E.C. Seaberg D.H. Watts N.A. Hessol S. Melnick P. Bitterman Low incidence of invasive cervical cancer among HIV-infected US women in a prevention program AIDS 18 2004 109 113 [8] L.S. Massad M.J. Fazzari K. Anastos R.S. Klein H. Minkoff D.J. Jamieson Outcomes after treatment of cervical intraepithelial neoplasia among women with human immunodeficiency virus J. Lower Genital Tract. Dis. 11 2007 90 97 [9] H. Cejtin E. Komanoff L.S. Massad A. Korn J.B. Schmidt D. Eisenberger-Matityahu Adherence to colposcopy among women with HIV infection J. Acquir. Immun. Deficiency Syndromes Hum. Retrovirol. 22 1999 247 252 [10] D.E. Stewart P.M. Buchegger G.M. Lickrish S. Sierra Effect of educational brochures on follow-up compliance in women with abnormal Papanicolaou smears Obstet. Gynecol. 83 1994 583 585 [11] S.M. Miller K.K. Siejak C.M. Schroeder C. Lerman E. Hernandez C.W. Helm Enhancing adherence following abnormal Pap smears among low income minority women: a preventive telephone counseling strategy J. Natl. Cancer Inst. 89 1997 703 708 [12] K. Ell B. Vourlekis L. Muderspach J. Nissly D. Padgett D. Pineda Abnormal cervical screen follow-up among low-income Latinas: Project SAFe J. Women's Health 11 2002 639 651 [13] L. Lacey J. Whitfield W. DeWhite D. Ansell S. Whitman E. Chen Referral adherence in an inner city breast and cervical cancer screening program Cancer 72 1993 950 955 [14] L.S. Massad P.M. Meyer J. Hobbs Knowledge of cervical cancer screening among women attending urban colposcopy clinics Cancer Prev. Detect 21 1997 103 109 [15] L.S. Massad S.J. Verhulst M. Hagemeyer P. Brady Knowledge of the cervical cancer screening process among rural and urban Illinois women undergoing colposcopy J. Lower Genital Tract. Dis. 10 2006 252 255 [16] K.A. Hild-Mosley D.M. Patel S. Markwell L.S. Massad Knowledge of cervical cancer screening, human papillomavirus, and HPV vaccine among Midwestern gynecology patients J. Lower Genital Tract. Dis. 13 2009 200 206 [17] S.L. Pruitt P.A. Parker S.K. Peterson T. Le M. Follen K. Basen-Engquist Knowledge of cervical dysplasia and human papillomavirus among women seen in a colposcopy clinic Gynecol. Oncol. 99 2005 S236 244 [18] B. Holcomb J.M. Bailey K. Crawford M.T. Ruffin Adults' knowledge and behaviors related to human papillomavirus infection J. Am. Board Fam. Pract. 17 2004 26 31 [19] J. Tiro H.I. Meissner S. Kobrin V. Chollette What do women in the U.S. know about human papillomavirus and cervical cancer Cancer Epidemiol. Biomarkers Prev. 16 2007 288 294 [20] S.J. Klug M. Hukelmann M. Blettner Knowledge about infection with human papillomavirus: a systemic review Prev. Med. 46 2008 87 98 [21] L.A.V. Marlow J. Waller J. Wardle Public awareness that HPV is a risk factor for cervical cancer Br. J. Cancer 97 2007 691 694 [22] Cervical Cancer. NIH Consensus Statement 1996 Apr 1–3;14(1):1–38. [23] S.E. Barkan S.L. Melnick S. Martin-Preston K. Weber L.A. Kalish P. Miotti The Women's Interagency HIV Study Epidemiol 9 1998 117 125 [24] M. Bacon V. von Wyl C. Alden G. Sharp E. Robison N. Hessol The Women's Interagency HIV Study: an observational cohort brings clinical sciences to the bench Clin. Diag. Lab. Immunol. 12 2005 1013 [25] L. Hatcher A step-by-step approach to using SAS for factor analysis in structural analysis and structural equation modeling. 1994 SAS Press Cary, NC [26] G.S. Wilkinson Wide Range Achievement Test 3 - Administration Manual 1993 Jastak Associates, Inc. Wilimington,DE [27] T. Del Ser J.I. Gonzalez-Montalvo S. Martinez-Espinosa C. Delgado-Villapalos F. Bermejo Estimation of premorbid intelligence in Spanish people with the Word Accentuation Test and its application to the diagnosis of dementia Brain and Cognition 33 1997 343 356 [28] A. Smith The Symbol-Digit Modalities Test: a neuropsychologic test for economic screening of learning and other cerebral disorders Learning Disorders 3 1968 83 91 [29] L.S. Radloff The CES-D scale: a self-report depression scale for research in the general population Applied Psychol. Measurment 1 1977 385 401 [30] D.G. Kleinbaum L.L. Kupper K.E. Muller A. Nizam Applied regression analysis and other multivariable methods 3 rd ed 1998 Brooks/Cole Publishing Pacific Grove CA [31] D.H. McCree P.A. Sharpe H.M. Brandt R. Robertson Preferences for sources of information about abnormal Pap tests and HPV in women tested for HPV Prev. Med. 43 2006 165 170 [32] G.G.G. Donders G. Bellen A. Declerq J. Berger T. Van Den Bosch I. Riphagen M. Verjans Change in knowledge of women about cervix cancer, human papillomavirus (HPV) and HPV vaccination due to introduction of HPV vaccines Eur. J. Obstet. Gynecol. Reprod Biol. 145 2009 93 95 [33] J. Waller L.A. Marlow J. Wardle The association between knowledge of HPV and feelings of stigma, shame, and anxiety Sex Transm Infect 83 2007 155 159 [34] A.F. Dempsey G.D. Zimet R.L. Davis L. Koutsky Factors that are associated with parental acceptance of human papillomavirus vaccines: a randomized intervention study of written information about HPV Pediatrics 117 2006 1486 1493