Barriers to participating in an HIV vaccine trial

Abstract

Introduction Human immunodeficiency virus (HIV) exacts a heavy toll in terms of morbidity and mortality. Education programs stressing preventative methods have managed to slow the spread of HIV infection in some regions [1]. However, an estimated 15 000 new individuals are infected each day. Highly active antiretroviral therapy (HAART), despite its proven benefits, remains too costly and complex for the majority of HIV-infected people, 90% of whom live in developing countries [2]. Under these circumstances, an effective preventative vaccine may represent the best hope for reversing the growing HIV pandemic. Phase I and II clinical trials [3] have evaluated more than 40 candidate HIV vaccines. Recently, investigators presented the disappointing results of the first Phase III clinical trials [4]. Efficiently implementing the additional large-scale vaccine studies necessary to test an effective preventative vaccine will require an understanding of the obstacles to enrolling sufficient numbers of people at high risk for HIV infection. Researchers have attempted to identify the barriers to enrolment by questioning individuals potentially eligible for or participating in preventative vaccine studies. We sought to synthesize the information from these studies by conducting a systematic review of this literature using content analysis techniques, particularly focusing on the currently existing qualitative data. In addition, we determined whether quantitative studies such as surveys were examining the domains identified by qualitative studies. Methods Search strategy We performed a systematic, all language literature search for qualitative studies and quantitative surveys addressing the anxieties and barriers to participation in HIV vaccine trials. Qualitative studies attempted to elucidate participants’ barriers through the use of open-ended questions. Quantitative studies limited the number of responses possible. Qualitative surveys asked questions in an open-ended manner and quantitative surveys asked questions in a structured manner. In duplicate we searched the following databases: AMED (1985–March 2004), Campbell Collaboration (2001–March 2004), CinAhl (1982–March 2004), Cochrane Library (1998–March 2004), Embase (1990–March 2004), E-Psyche (1993–March 2004), HTA (1988–March 2004), ERIC (1966–March 2004), MedLine (1966–March 2004), and NHS EED (1975–March 2004). Unpublished studies were also sought using Clinicaltrials.gov and the UK National Research Register. Our search strategy combined terms that represented attitudes, barriers and expectations. We supplemented this search by searching the bibliographies of key papers. Selection of abstracts Two of us (C.S. and K.W.) independently evaluated the abstracts of retrieved articles. Eligible studies met the following criteria: (1) report an original research study; (2) contain content addressing barriers or negative attitudes to participating in HIV vaccine trials; and (3) be either a focus group study, semi-structured interview or survey. Kappa scores reflected chance-adjusted inter-observer agreement in the study identification process. Data abstraction and validity assessment Two study group members (A.G. and B.R.), independently extracted data and appraised both validity and content. A third reviewer (E.M.) dealt with discrepancies. We characterized the content of the methodology using a modified checklist designed to examine internal validity [5]. A.G. and B.R. read and scored each qualitative study using the checklist (Table 1). We used content analysis to synthesize the data extracted. From an initial review of the papers A.G. and B.R. iteratively developed a coding template. This template consisted of mutually exclusive headings and subheadings to categorize barriers identified in the papers. A second review of the papers identified whether individual papers had identified a barrier that fell under one of the subheadings.Table 1: Characteristics of semi-structured qualitative interview studies on barriers to participation in HIV vaccine trials.Subgroup analysis This analysis compared the results of qualitative studies and quantitative surveys to determine whether the questions asked in quantitative surveys reflected the issues that were raised through the qualitative studies. Differences in the reporting rates of each issue were evaluated by Fisher's exact test. A total count of the number of themes mentioned in each category for each study was calculated and an Exact Wilcoxon rank-sum test (i.e. a non-parametric equivalent to a two-sample t-test, suitable for small datasets where the data follows a highly irregular distribution) evaluated whether associations between study type and outcomes were attributable to chance. A sensitivity analysis was performed by Mann–Whitney U-test to determine whether the types of issues examined in these studies changed over time (determined according to year of publication), and whether the nature of this change differed across qualitative and quantitative studies. We also examined if country of origin, urban versus rural, status of risk, and participation in a trial affected reported outcomes. The effect of reporting quality, as determined by the checklist outlined in Table 1 was assessed by Spearman's test for correlation. Results Study selection and characteristics The systematic literature search yielded 211 studies (Fig. 1). There was near-perfect agreement between reviewers in choosing the final 46 relevant studies (κ = 0.88). Of these 46 articles, 20 were excluded because they were not original, did not examine barriers to participation in vaccine studies, or were not focused on HIV vaccination. The remaining 26 studies were included in our analysis (κ for inclusion = 0.97).Fig. 1.: Flow diagram of included studies.Seven studies were qualitative in nature. Five conducted semi-structured qualitative interviews (n = 2157) [6–10] and two utilized open-ended questionnaires (n = 645) [11,12]. Nineteen studies proved to be structured quantitative questionnaires (total n = 23 785). Sixteen studies were conducted in the United States [6–8,10,12–23], five in Thailand [9,11,24–26], two in Brazil [27,28], and one each from Uganda [29], Canada [30] and Kenya [31]. All studies were published in English. Of the seven qualitative studies, two studies included individuals who were already enrolled in an HIV vaccine trial [8,9]. Characteristics of study methodology There was large variability in the quality of reporting of the five semi-structured qualitative interview studies; most studies did not describe the methodology in detail (mean score 3.2/8) (Table 1). Characteristics of study methodology were not associated with the total number of issues raised (0.73, P = 0.73) but did correlate with issues of concerns about study design (0.89, P = 0.04) and with issues of fear or mistrust (0.89, P = 0.04) (Table 2).Table 2: Characteristics of qualitative studies.Themes reported in qualitative studies We analyzed the themes reported in the five semi-structured interview studies and two open-ended questionnaires (Table 3). Themes fell into the following mutually exclusive categories: (1) safety concerns; (2) fear or mistrust; (3) concerns or misunderstandings about study design; (4) discrimination/social risk; (5) pragmatic obstacles; and (6) other. The agreement between reviews was, in general, high. The kappa ranged from 0.55 to 1.00.Table 3: Issues raised in seven qualitative studies and compared in 19 surveys addressing barriers to participating in an HIV vaccine trial.All qualitative studies reported safety concerns. Specific concerns included vaccine specific side effects (five of seven), vaccine safety and efficacy (five of seven), the risk of contracting HIV from the vaccine (five of seven), and the potential for increased high-risk behavior (three of seven). Three of seven qualitative studies reported concerns or misunderstandings about study design. Specific concerns included the issue of blinding and not being aware of treatment (two of seven), the possibility of receiving placebo vaccine (two of seven), general study design concerns (one of seven), and the belief that one could not participate if not considered a high enough risk or not HIV positive (one of seven). Three of the seven qualitative studies considered fear or mistrust and identified fear or mistrust of governments (three of seven), researchers and the research process (three of seven), and pharmaceutical companies (one of seven) as potential barriers to participation in vaccine trials. One open-ended questionnaire raised the concerns about being treated like a ‘guinea pig’ and that potential study candidates ‘Do not know what researchers are putting in the needle [7]'. All qualitative studies discussed discrimination and social stigma risk. Key themes raised included the concern of sero-conversion to positive status (five of seven), general discrimination against the participant and his/her family (four of seven), being mistakenly viewed as HIV infected (two of seven), that other people (friends, family, and partners) may be afraid of infection through contact with participant (two of seven), and refusal of sex by a partner because of involvement (one of seven). One semi-structured interview discussed inadvertent disclosure of trial participation as a barrier. In one specific case, a patient was hospitalized with pneumonia. After informing the staff of participation in a preventative HIV vaccine study, the patient chart was marked as HIV positive [8]. Five of the seven qualitative studies identified pragmatic obstacles. Participants described the inconvenience or personal limitations on day-to-day life that participation may bring were (four of seven), their travel or immigration concerns and insurance concerns (one of seven), and employment concerns (one of seven). One semi-structured interview study and one open-ended questionnaire described employment concerns. One semi-structured interview and one open-ended questionnaire described barriers or beliefs that we did not categorize under the previous headings. The semi-structured interview discussed ‘unpleasant experiences'. The open-ended questionnaire looked at variables such as dislike of needles or medical procedures, animal testing, not being a believer in modern medicine, and wanting to get advice from leading AIDS activist groups. Survey issues related to qualitative studies Quantitative survey questionnaires failed to incorporate several important issues at a significant or near-significant level (Table 4). Based on pooled analysis, barriers relating to safety (P ≤ 0.01) and discrimination/social risk (P = 0.02) are under-represented in quantitative studies when compared with the frequency in which these issues are mentioned in qualitative studies.Table 4: Qualitative studies versus quantitative studies reporting of barriers to participating in an HIV vaccine trial.Sensitivity analysis A sensitivity analysis revealed evidence of a time-effect in the reporting of pragmatic obstacles when issues were pooled (P = 0.01). These issues have been raised more frequently in recent studies than they were in earlier studies. This was true for both qualitative and quantitative studies. However no evidence of a differential time effect between quantitative and qualitative studies on any of the other individual issues was found. We found no effects of country of origin (P = 1.00), urban versus rural (P = 1.00), status of risk (P = 0.57) or participation in a trial (P = 0.14) on the reported barriers and obstacles to study participation. Discussion This is the first published systematic review examining the barriers pertaining to HIV vaccine trial participation. The original studies consistently identified key themes of safety, mistrust of those involved with clinical research, and risk of discrimination as a result of participation in HIV vaccine studies. Vaccine trialists can use the findings of this study to anticipate common concerns of participants and offer guidance to volunteers contemplating participation in a clinical trial by informing them of the documented social harms that have occurred in previous phase I/II trials [17]. The results also suggest that trialists should emphasize the required preclinical and clinical safety monitoring for all products tested. Despite the monumental efforts taken to develop candidate HIV vaccines that are safe, this is an issue of concern often raised within the scientific community and HIV advocates [32,33]. Our review demonstrates that vaccine safety is a foremost concern among those most likely to participate in preventative vaccine trials. This highlights the importance of addressing issues of safety in an open and thorough fashion. There is accumulating data pertaining to the safety and tolerability of several different HIV vaccine candidates [34]. This experience should be shared with individual participants and the communities in which these studies are conducted. In addition to the obvious benefits of easing participant anxieties, an approach such as this is likely to allay concerns related to mistrust of the scientific community and the research process [33]. Concern about study design also arose in several studies. Study designs can be complicated and the terms used to characterize them (e.g. blinding, randomized, placebo) may confuse or dissuade otherwise eligible candidates from participation. The process of informed consent is an ideal time to clearly define these terms and provide candidates with the opportunity to clarify the meaning of the technical terms used in clinical research. It is not difficult to understand why those participating in clinical research may have fears regarding the intent of the investigators. Different cultural backgrounds, languages, historical experiences, levels of education and technological development contribute to mistrust of researchers, governments and industry. History provides multiple examples that justify these concerns [35]. Successful conduct of future vaccine trials therefore depends on a concerted effort by those conducting research to clearly communicate intentions, involve local investigators and lay people in the conduct of research, and to assist in the development of long-term expertise in medical care and clinical research within the local population. Any trial of an HIV vaccine must take into account the history of exploitation and abuse of vulnerable people in clinical trials [36,37]. Discrimination and fear of isolation constitute additional barriers to participation in vaccine studies. HIV seropositivity resulting from HIV vaccination, whether protective or not, may have a devastating effect on participant lives in terms of medical care, employment opportunities, family and sexual relationships, and the ability to obtain insurance. Community participation in the conduct of vaccine studies and a broad effort to education the medical community and population at large will help to avoid at least some of these negative consequences. A comparison of the content of quantitative versus qualitative surveys suggests a discordance between what researchers anticipate as significant barriers to enrolment and what participants see as obstacles to study conduct. Specifically, content specific to safety concerns and fear of discrimination appear to be under-represented in quantitative evaluations. Furthermore, our findings suggest the advisability of paying greater attention to study participants’ mistrust of researchers and the research process. Perhaps the most important factor limiting enrolment, which may be underestimated by those conducting vaccine studies, is the pragmatic issue of participant inconvenience and disruption to the normal day-to-day routine resulting from participation in a clinical trial. Systematic review of qualitative studies is a relatively new pursuit and the methodology employed is experimental [38–40]. This work incorporates and expands upon previous systematic reviews of qualitative studies [41–44]. The major limitation of our methodology is that if a barrier is not identified in several studies, it may be a result of reporting bias of the authors and not actually represent the absence of the barrier [45]. However, if a barrier is identified in many studies, it provides strong support for addressing the issue. It should be noted that although our sensitivity analysis did not show differences by country of origin, this might be due to the small number of qualitative studies available for inclusion in our analysis. Therefore, HIV vaccine trialists should also specifically examine the results of the qualitative studies that were obtained in the countries where they are intending to conduct trials. Based on the results of our review, safety concerns and fear of social discrimination are key issues that may limit participation in future preventative vaccine trials. A well co-ordinated research plan that includes education, active participation of the local community and the development of local expertise and infrastructure is likely to help to address these concerns, assist in enrolment, and improve protocol compliance. Acknowledgements Sponsorship: This study was partially funded by a grant from The Ontario HIV Treatment Network. EM is supported by the Ontario HIV Treatment Network and KW is a CIHR new investigator. Author contributions The contributions of the individual authors were: concept: E.M., K.W., C.C.; design: E.M., K.W., G.G.; searching, data abstraction: E.M., B.R., A.G., C.S., K.W.; data analyis: E.M., K.W., B.R., A.G., C.S., G.G.; writing: E.M., K.W., G.G., C.C., C.S., B.R., A.G.; critical revisions: G.G., E.M., C.C., C.S.