Jenkins, Richard A PhD*†‡; Thapinta, Darawan RN, PhD§; Morgan, Patricia A PA-C, MS†‡; Wongkamhaeng, Siriluck RN†‡; Sornsathapornkul, Pornchai RN∥; Bussaratid, Valai MSc, MD¶; Sontirat, Auchara MSc#; Pitisuttithum, Punnee MBBS, DTM&H, FRCPT¶; Thongchareoen, Prasert MD, PhD∥; Khamboonruang, Chirasak MD, PhD#**; Suriyanon, Vinai MD#**; Nitayaphan, Sorachai MD, PhD††; Brown, Arthur E MD, MPH‡; and the Thai AIDS Vaccine Evaluation Group
Social and behavioral aspects of HIV-1 vaccine trials have received substantial attention over the past decade, particularly issues related to the conduct of trials outside North America and Europe.1-8 Concerns have been raised about behavioral side effects (increased risk behavior in response to a perception of protection), the potential for unblinding, and the possibility of employment or insurance discrimination on the basis of HIV-1 test results derived from commercial kits typically used in hospital or community settings.3,4,8-10 In addition, concerns have been raised regarding the potential for social discrimination because of volunteers' association with an HIV-1 vaccine trial.4,7 Studies of attitudes toward vaccine trial participation in Thailand have indicated that the greatest concerns were related to vaccine product safety, although concerns about social discrimination also have been common.11-13
Despite the many commentaries on the conduct of HIV-1 vaccine trials, a body of data-based literature addressing behavioral issues during trials is only beginning to accumulate. Much of this literature has come from trials conducted in Thailand, including 3 published investigations of longitudinal changes in behavior during HIV-1 vaccine trials.14-16 Declines in sexual and drug use risk behaviors were seen in a phase 3 trial among injection drug users (IDUs) in Bangkok, and no interactions were observed between risk behaviors and factors such as perceived experimental status (vaccine vs. placebo) or perceived protection from the vaccine, although a short-term increase in injecting frequency was observed early in the trial.14 Phase 1/II trials in broader populations “not at highest risk” found no behavior change across the vaccine trial period15 or declines in sexual risk behavior.16 Two other early stage trials reported relatively low risk behavior that was unchanged across trials but did not provide specific data.17,18
Outside Thailand, the largest study was based on a multisite phase III trial in the United States, which found high levels of sexual risk behavior at baseline. Risk behavior, in general, did not increase across the course of this trial, although increases were noted among several important subgroups of men who have sex with men (MSM): those less than 30 years of age, those who perceived themselves as being assigned to the vaccine condition, and those who were not of African-American race.19 McElrath et al20 reported no increase in risk behavior among participants in a large phase II trial, whereas Chesney et al21 found that risk behavior increased among MSM in 2 small trials.
The purpose of the present investigation was to examine behavior at entry into a preventive HIV-1 vaccine trial and across the course of the trial. This built on a prior investigation15 from a smaller trial (n = 50) in a similar population with a shorter follow-up period (8 months). In that study, relatively low levels of risk behavior were present at baseline and at the follow-up visits. Volunteers in that trial reported relatively few negative social reactions to their participation and no evidence of frank discrimination in employment, health care, or insurance. Because of the absence of increased risk behavior in this previous trial, we were interested in assessing the experience of vaccine trial volunteers in a larger sample to understand better the generalizability of these findings, particularly in light of other data from Thailand that obtained declines14,16 or little change17,18 in risk behavior and the more mixed findings from the United States.19-21 Attention was given to HIV-1 risk behavior, psychologic distress, HIV-1 knowledge, and trial participants' social experiences (including any exposure to social harms).
Data were collected during a phase I/II, double-blind, placebo-controlled trial of Chiron Biocine Thai E gp120/MF59 vaccine (Chiron/Vaccines, Emeryville, CA) alone and in combination with Chiron Biocine SF2 gp120 vaccine (subtype B) at varied dosages.22 Both vaccines were recombinant envelope products. Evidence of safety for the Chiron Biocine SF2 gp120 has been documented in North American23 and Thai samples.24 Evidence of safety for the Chiron Biocine Thai E gp120/MF59 vaccine also has been documented.22 The trial was conducted at 4 medical center-related clinics in Thailand: (1) the Joint Clinical Research Center (JCRC) at the Armed Forces Research Institute of Medical Science (AFRIMS) and Phramongkutlao Hospital in Bangkok, (2) the Faculty of Medicine at Siriraj Hospital of Mahidol University in Bangkok, (3) the Vaccine Trial Center (VTC) of the Faculty of Tropical Medicine at Mahidol University, and (4) the Research Institute for Health Sciences (RIHES) at Chiang Mai University in Chiang Mai.
The screening and vaccine trial protocols were reviewed and approved by Institutional Review Boards at the RIHES; Chiang Mai University; US Army Research and Materiel Command; Royal Thai Army Medical Department; Faculty of Medicine of Siriraj Hospital, Mahidol University; Faculty of Tropical Medicine of Mahidol University; and Medical School of Mahidol University as well as by the Ethical Review Committee for Research in Human Subjects of the Thai Ministry of Public Health. The vaccine trial also was required to be reviewed and approved by the HIV/AIDS Vaccine Development Review Subcommittee of the National AIDS Commission of Thailand.
The trial was completed by 363 volunteers (of 370 enrolled). Recruitment procedures have been described elsewhere.25 In brief, sites used a variety of community outreach strategies with Buddhist temples, universities and vocational schools, hospitals, community leaders, and direct contacts with community organizations. Newspaper advertisements and interviews with study investigators in various media also were used to publicize the trial. In addition, the RIHES contacted individuals who had participated in its earlier vaccine cohort preparation studies. At initial contact with the study, prospective volunteers were asked their reasons for interest (through an open-ended question) and information was orally provided regarding the trial (eg, screening process, inclusion/exclusion criteria, use of placebo, need for maintaining risk reduction) and supplemented with a brochure. Demographic information was collected from all prospective volunteers. Those who wished to have a screening appointment were asked for contact information (unless they had reported conditions that were obvious exclusion criteria).
The screening procedures have been detailed elsewhere22,25 and were similar to those used in a previous smaller trial.26 Pretrial medical and behavioral screening involved a minimum of 2 visits and completion of a prescreening questionnaire and an informed consent form separate from that used for the vaccine trial. A description of the trial and the vaccine trial consent form were provided during the first visit, although the consent form was not signed until the close of the final screening visit. Thus, volunteers could read the form at their leisure, formulate questions, and, if desired, discuss it with family members or friends. Enrollment was contingent on meeting a variety of medical and laboratory criteria as well as availability for follow-up. Enrollment also required the absence of serious psychiatric symptoms (based on clinical interview, questionnaire measures, and consultation with mental health professionals associated with the trial) and the absence of “highest risk” behavior (ie, active injection drug use, commercial sex work, active sexual relationship with a known HIV+ partner). Enrollment was contingent on passing a 37-item “test of understanding” regarding the vaccine product (eg, “The vaccine that is being tested in this study contains the HIV virus-true or false”), trial procedures (eg, “If I enter this study, I may receive an inactive medicine [placebo]-true or false”), management of HIV risk (eg, “People who join this study will no longer need to use condoms when they have sex with new partners-true or false”), and consent procedures (eg, “I can withdraw from the study at any time, without prejudice-true or false”), which was administered at the end of the final screening visit.
Behavioral data were collected through structured questionnaire interviews conducted by specially trained study nurses. Structured questionnaires were administered at the second screening interview and at the 4-, 8-, and 12-month (last visit of the trial) follow-up visits. Nursing staff conducting the interviews received training in behavioral interviewing with the measures used in this study as well as periodic supervision and consultation by 2 doctoral level psychologists (R.A.J. and D.T.). Female staff interacted with male and female volunteers, but male staff only interacted with male volunteers.
Baseline collection of behavioral data occurred as the last step during the screening but before administration of the test of understanding. Volunteers who were excluded on the basis of laboratory findings, medical history, or physical examination were not asked to provide baseline behavioral data, nor were volunteers who self-selected out of consideration for the trial during screening.
Core behavioral questionnaire content included lifetime and recent exposure risk behavior, sexually transmitted disease (STD) symptoms and diagnoses (which were to be followed up with clinical examination), HIV-1 transmission knowledge, and current distress. Items regarding recent and past risk exposure as well as HIV-1 knowledge were adapted from an interview protocol used in Thai vaccine preparatory cohort studies.27 Distress was measured by the relative frequency of depressive and anxiety symptoms during the past week and used Likert-type items developed for a previous study.15 The depressive symptom scale included 12 items (eg, suicidal thinking, difficulties with sleep and appetite, feeling unloved, lacking hope), and the anxiety scale included 5 items (eg, feeling nervous, being unable to relax, panic, worry about health). The coefficient alpha,28 a measure of a scale's internal consistency, was 0.67 and 0.62 for depressive symptoms and anxiety, respectively, in the current study sample. These items were part of the baseline assessment of psychiatric status (along with data collected in the medical history and systems interview). We collected these data in subsequent visits to see if trial participation might be related to changes in overall psychological functioning.
At the follow-up visits only, core interview content was supplemented by items regarding social experiences related to the trial (ie, had volunteers noticed changes in the way they were treated by friends, family, or employers or evidence of discrimination in health care, insurance, or employment, followed up by open-ended prompts to ascertain trial relevance), perceptions of experimental assignment (ie, vaccine, placebo, or unsure, followed up by open-ended prompts to explain their choice), and volunteers' general attitudes about their participation in the trial (measured with Likert-type items, followed up by open-ended prompts). These items, which addressed social harms, experimental assignment, and attitudes about the trial, had been developed for a previous trial15 and had been pretested with clinic populations.
HIV-1 education and risk behavior counseling were provided at all study visits (including those where no behavioral data were collected) by study nurses, with consultation provided by doctoral level psychologists who oversaw the behavioral assessment. The psychologists were available for consultation if significant behavioral problems occurred (eg, suicide attempts), along with senior medical staff. The Thai psychologist was available for direct assessment of volunteers. During data collection visits, counseling took place after behavioral data had been collected, although at baseline, pretest counseling had occurred at the blood draw, which preceded behavioral data collection. Nurses had received training in the following areas: rapport building, active listening, nondirective elicitation of information, collaborative problem solving, role definition and limit setting, discussion of sexual matters, and topics relevant to pre- and posttest counseling.
Data were analyzed using SPSS Windows, version 12.0 (SPSS, Chicago, IL). Within-sample comparisons were made using Pearson χ2 tests, Fisher exact tests, Pearson product-moment correlations, and t tests. The Goodman-Kruskal gamma statistic was used to evaluate correlations between pairs of ordinal variables, and the Cochran Q was computed to evaluate temporal changes in dichotomous variables. The Kruskal-Wallis test was used to evaluate site differences on continuous variables. The Kendall W statistic and Friedman test were used to evaluate changes in continuous variables between study visits.
A total of 363 (of 370) volunteers completed the trial. The mean age was 32.1 years (SD = 7.5, median = 30.5 years, range: 21-47 years). Table 1 summarizes other demographic characteristics of the sample. Overall, the sample was predominantly male and relatively well educated. Approximately one half of volunteers were single. There were no gender differences with respect to age or education; however, men were more likely to be single than women (Fisher exact test, P < 0.006). Bangkok volunteers were nonsignificantly more likely to be male, single, and more educated. Detailed site comparisons are noted in the article by Thapinta et al.25
Data on lifetime history of HIV-1 relevant risk behaviors, collected at baseline, are summarized in Table 2, along with data collected at the 3 follow-up visits. The 4 sites did not differ on baseline or follow-up levels of risk behavior, history of STDs, or past commercial sex worker (CSW) contacts. History of CSW contacts (66.9%) and STD history (4.3% of the sexually experienced men) were consistent with data from national surveys of Thai men.29,30 Of the sexually experienced men, had a history of CSW contacts; however, relatively few sexually experienced men reported having had sex with a CSW in the 6 months before enrollment (5.4%), and a similar level of CSW contact continued through the trial. Consistent condom use with CSWs was high at baseline and throughout the trial, and no seroconversions occurred during the trial or during the 18-month medical follow-up that took place after the trial had ended. Consistent with general population norms,31 casual partners were reported only by male volunteers (4.8%). Only 1 male participant reported sex with other men. At baseline, 1.4% reported having a recent HIV-seropositive partner, whereas 0.6% to 1.7% reported HIV-seropositive partners during follow-up visits. Condom use with non-CSW partners generally was low. At baseline, 9.7% reported consistent condom use with wives, 27.3% with girlfriends, and 100.0% with casual partners, and little change was reported over time.
Male participants accounted for all STD diagnosis histories, and 82.4% of STD symptom histories. In addition to the questionnaire data, it should be noted that no STDs were detected clinically or serologically at baseline and that none were clinically detected during the trial. Male participants also reported an earlier mean age at first sex (t = 2.69, df = 16.00, P < 0.001), consistent with population norms.
No injection drug use was reported. IDU partners rarely were reported (see Table 2). Recent use of other drugs was relatively infrequent (see Table 2), and marijuana was the most frequently reported drug. Alcohol use showed 2 somewhat divergent trends. Abstention from alcohol increased across visits (Cochran Q [3, 346] = 61.18; P < 0.001) from 33.2% to 48.5%. The frequency of drinking among those reporting alcohol use differed significantly among the visits (Kendall W [3, 369] = 0.03; P < 0.003; see Table 2), although most respondents reported using alcohol once per week or less. At baseline, alcohol use was significantly more common among men (72.2%) than women (59.1%) (χ2 [1, 369] = 6.20; P < 0.01). It also was significantly more common among married (76.2%) than unmarried (60.7% of widowed to 59.7% of single) volunteers (χ2 [3, 369] = 10.67; P < 0.005) but was unrelated to age and education. Use of alcohol showed site differences at baseline (χ2 [3, 369] = 26.25; P < 0.001) and was less common at one of the sites that recruited Buddhist monks (VTC [44.2%]) than elsewhere (77.2%-69.6%). Among alcohol users at baseline, frequency of drinking was unrelated to site or age but was significantly related to education (r = −0.23; P < 0.006), which suggested that drinking occurred more often among less educated volunteers. There also was significant association with marital status (Kruskal-Wallis test = 12.12; P < 0.006). Married volunteers drank less frequently than all categories of unmarried volunteers. These baseline associations held at subsequent visits.
At baseline, 99.7% of volunteers recognized all the commonly accepted modes of HIV-1 transmission (eg, sexual, perinatal, sharing of needles), and this level was maintained at each follow-up visit (98.9% recognized all modes of transmission at the 4-month follow-up, 98.6% at 8 months, and 98.1% at 12 months). No site or demographic differences could be detected given the high level of knowledge. Approximately one third of the sample (35.5%) indicated that HIV-1 transmission could occur through 1 or more types of casual contact (eg, shared utensils, mosquitoes) at baseline, and this finding differed significantly by site (χ2 [3, 369] = 10.73; P < 0.01), with a range from 25.0% (VTC) to 47.8% (Siriraj). Over time, there was a significant decline in the tendency to see HIV-1 transmission as possibly occurring through casual social contact (Cochran Q [3, 369] = 21.16), and only 24.3% endorsed (χ2 [3, 369] = 10.73; P < 0.01) 1 or more of these at the 12-month follow-up, although the relative site differences remained. There were no gender, age, or education differences.
Impact of Participation in the Trial
We present the distress data transformed to item-means rather than means for the total score. The summed scores across items were divided by the total number of items; thus, the means reflect the scaling that was present for any individual item (where 1 meant the absence of symptoms and 5 indicated that they were always present). Depressive symptoms were largely absent at baseline and at the follow-up visits, although there were significant differences in reported symptoms between visits (Table 3; Kendall W [3, 369] = 0.17; P < 0.001). There were significant baseline differences for depressive symptoms by site (Kruskal-Wallis H [3, 369] = 9.91; P < 0.02), and item-means ranged from 1.37 to 1.53. The item-means here indicated that respondents generally reported no or infrequent symptoms of depression, even when significant differences between groups were present.
Anxiety symptoms also tended to be low at all visits, but statistically significant differences were present between visits (Kendall W [3, 369] = 0.20; P < 0.001; see Table 3). There were statistically significant baseline differences in anxiety symptoms by site (item-means range: 1.29-1.52; Kruskal-Wallis H [3, 369] = 17.77; P < 0.001). Similar differences by site were present at the follow-up visits. No gender or age differences were present at baseline or follow-up for either scale. The item-means here indicate that respondents generally reported no or infrequent symptoms of anxiety, even when significant differences between groups were present.
There was a significant correlation between education and depressive symptoms at the 4-month follow-up visit (r = 0.10, P < 0.02) and with anxiety symptoms at all visits (r = 0.10-0.23). This reflected slightly greater reporting of symptoms by college-educated respondents. The single notable event associated with symptoms of distress (unrelated to the trial itself) was a suicide attempt that occurred after a participant experienced the end of a long-term relationship.
Comfort in Discussing the Trial With Other People
Most of volunteers were comfortable discussing the trial, and this level of comfort significantly increased across the duration of the trial (Table 4). We did not specifically ask who they told about their trial participation, but family members were most often mentioned. There were no site, gender, or age differences in comfort; however, there was a significant correlation with education at the 8-month (r = 0.11, P < 0.03) and 12-month (r = 0.13, P < 0.007) follow-up visits, which suggested that more educated volunteers were slightly more comfortable discussing the trial with others. A small number (approximately 1%-2% at each follow-up visit) reported that they did not discuss the trial with anyone. Others noted that they were selective in disclosing their participation with others.
Social Harms to the Volunteers
No reported instances of discrimination in employment, health care, or insurance occurred. One volunteer reported having been refused sex by a long-term partner. Twenty-two volunteers (5.9%) reported being treated “differently” in some way by others (all of these being negative reactions), however. Most of this occurred in the first 4 months after the initial vaccination (n = 15 [4.1%]). Six volunteers reported being treated differently at the 8-month follow-up visit, and 2 reported this at 12 months. Only 1 volunteer reported being treated differently during more than 1 follow-up period. The most common concerns involved family or friends being afraid of what might happen to the volunteer during the trial; however, 2 volunteers stated that family or friends had tried to dissuade them from taking part. Only 1 volunteer had family or friends visit the clinic. Others preferred discussing the trial with concerned family and friends on their own and used materials prepared by the clinics.
Perception of Experimental Assignment (vaccine vs. placebo)
The proportion of volunteers who believed they were on active vaccine ranged between 41.0% and 44.4% across the 3 follow-up visits (Table 5A). Perceptions at the 4-, 8-, and 12-month follow-up visits did not significantly differ from each other. Although volunteers were more likely to have been assigned to a vaccine condition than placebo, 45.2% to 50.4% believed they were in the placebo group. There was a nonsignificant trend toward perceiving oneself to be on placebo as the trial progressed. Perceptions of experimental assignment were unrelated to gender, age, education, or risk behavior. These perceptions were significantly related to site at each follow-up visit because of a tendency for more volunteers at the VTC to say they were unsure of the condition to which they were assigned (22%-24% vs. 3%-13% at the other sites).
The factor that seemed most decisive in the perceptions regarding vaccine or placebo assignment was the presence or absence of physical side effects (eg, pain around the injection site, tiredness, headache) during or after the injection. Depending on the visit, 24.1% to 34.7% of those who believed they had received vaccine explained their perception by mentioning physical symptoms. Only 1 person who believed he or she had received placebo attributed it to the presence of some side effect. In contrast, 57.8% to 71.4% (depending on the visit) of those who thought they had received placebo attributed this to the absence of any physical symptoms after their injections. Even so, 8.0% to 11.5% (depending on the visit) of those who perceived themselves to be receiving vaccine used the absence of physical symptoms to explain this perception. Between 1% and 2% at each visit attributed being on the vaccine to feeling stronger and healthier. A few others thought that their immunity had improved but did not elaborate. Depending on the visit, 6% to 8% noted that they thought they were in the vaccine group simply because most people were receiving vaccine. Two volunteers reported that they had donated blood (a deviation from the advice given to volunteers) without problems and thought themselves to be receiving placebo.
Approximately 40% to 45% of those actually assigned to the vaccine group perceived that they had received vaccine or placebo (see Table 5B), depending on the visit, a proportion that nearly approximates the odds of a coin flip. The ability of volunteers to guess the condition to which they had been assigned correctly did not change across the trial. Despite the importance of physical side effects as a determinant of perceived assignment, perceptions among volunteers seemed to approximate the odds associated with a coin toss. As noted previously, only a small percentage recognized that the odds of assignment favored being in the vaccine group. Most (87%-92%) of those who said they did not know to which group they had been assigned actually had been assigned to the vaccine condition.
Benefits of Trial Participation
At each follow-up visit, most of volunteers (98.1%-98.3%) were able to name some benefit from participation, whereas 4 to 5 volunteers specifically indicated that they did not receive any benefit from the trial. Approximately two fifths of the volunteers (38.0%-42.0% depending on the visit) named more than 1 benefit, and a small number of people named as many as 4 benefits. The most commonly reported benefits were knowledge about HIV/AIDS (36.2%-43.8% depending on the visit), monitoring of their health (30.9%-36.3%), and helping society (29.5%-30.2%). Compensation was mentioned by 2 people at each visit. Small numbers of people indicated that knowing the staff, receiving counseling for a personal problem, or making friends with other volunteers had been benefits.
When asked at the close of the trial if they would join again if asked, 95.9% indicated that they would. Their reasons largely mirrored those that they had provided as “benefits” of their participation in the trial, although some also noted that “nothing bad happened.” Dissenting opinions focused on the inconveniences of the trial. Three volunteers indicated that they might consider joining another trial depending on the circumstances (ie, personal schedule, potential effectiveness of the vaccine).
During this phase I/II HIV-1 vaccine trial in Thailand, low levels of current risk behavior were evident at baseline and follow-up. No new HIV infections occurred during the trial. Similar findings were reported in smaller previous trials in Thailand that had demographically similar samples.16-18,25 Data from phase II trials conducted with a variety of high-risk populations in the United States have reported no evidence of increased risk behavior.20 In contrast, a study of 2 small phase I/II preventive HIV-1 vaccine trials in the United States targeting individuals with high-risk sexual behavior21 found significant increases in some risk behavior during the trials.
The nonoccurrence of increased risk behavior in our trial may reflect different screening criteria, which involved excluding individuals with evidence of high-risk behavior. Bartholow et al19 found no overall increase in risk behavior among phase III trial participants in the US trial overall but did find increases among those who believed that they were receiving vaccine rather than placebo. Chesney et al21 found that increased risk behavior was associated with a history of high-risk behavior and hope that the vaccine would confer protection. van Griensven et al14 found decreases in risk behavior in their phase III trial with a high-risk population (IDUs) in Bangkok, with no interactions between risk behavior and perceptions of protection. Maek-A-Nantawat et al16 reported decreases in risk behavior in a population similar to that presented in this report. The frequency of counseling in the Chesney et al21 and McElrath et al20 studies was not reported, although Chesney et al21 indicated a high level of knowledge retention regarding directives for protection. The phase III trials in Bangkok14 and in the United States19 included counseling at periodic 6-month visits in conjunction with HIV-1 testing at those visits. Volunteers in the present investigation received counseling at all visits, as was also the case in Maek-A-Nantawat et al's study.16 Follow-up periods in these studies have varied widely from 8 months15 to 36 months,19,20 although most have followed volunteers for a total of 12 months.14,16-18,21 The relatively low levels of relevant risk behavior (and social harms) did not permit us to use the more powerful multivariate longitudinal statistics used in some other investigations.14,19
There was a much lower level of recent CSW contact in the current study than in a trial that had occurred 3 years earlier15; however, a continuing decline in CSW patronage has been noted in Thailand since the early 1990s.32 High levels of condom use with CSW partners have been found in recent research conducted in the communities in which our trial occurred32,33; however, 2 cautions should be considered in evaluating the risk behavior data. The use of face-to-face interviews may have constrained disclosure of risk behavior relative to other methodologies.34 Interview methodologies are widely used in Thailand, including the recent phase III trial among IDUs.14 Work with Thai populations has suggested that high rates of disclosure are possible when interviewers are well trained and rapport has been developed.35 Interviews also were used in the trials presented by McElrath et al20 and Chesney et al.21 Interviews do mitigate problems with comprehension, inconsistent response sets, and missing data as well as permitting complex questionnaire structures; however, all these advantages can be incorporated into more up-to-date methods such as audio computer-assisted self-interviewing (ACASI), which provides the anonymity of self-administration36 and can be adapted to handheld computers. This technology began to be used in HIV-1 prevention research in Thailand after the initiation of this trial.37 In this study, data collection interviews took place before counseling was provided (except in the case of the baseline, where pretest counseling took place when blood was drawn for screening), limiting the effect of counseling on responses. Gender matching (male staff were matched only with male volunteers, but female staff saw male and female volunteers) also may have mitigated some reluctance to disclose.
Volunteers generally were quite knowledgeable about HIV-1 at enrollment. As in most other studies conducted in Thailand, knowledge about HIV-1 transmission was high, although concerns have persisted regarding the potential for transmission through casual contact.38 Statistically significant site differences existed for beliefs regarding acquisition of HIV through casual contact. These may reflect differences in the recruitment of volunteers by site, although expression of 1 or more of these beliefs was relatively common (25%-50%) at all sites initially and declined over time. High levels of scientifically correct knowledge about modes of HIV transmission have been reported in other preventive HIV-1 vaccine trials conducted in Thailand.17,18,25
The potential for social harm has been raised as a concern in investigations of willingness to participate in trials conducted in Thailand.12,39 In the present investigation, no problems related to employment, insurance, or access to extratrial health care were reported. Overtly negative social reactions to study participation were reported by a few volunteers, although some participants were reticent about disclosing their participation in the trial to other people. Family and friends accounted for most of the negative reactions to a volunteer's participation in the trial. These situations were infrequent and occasionally required assistance from the study staff. Clinic nurses provided educational materials and, in a single case, met with the family of a volunteer to explain the study design and the vaccine product that was being evaluated. Overall, the negative social consequences of participation were less frequent and less severe than those reported by Sheon et al10 in a phase II trial conducted in the United States. Levels of depressive and anxiety-related symptoms were low at baseline and at all 3 follow-up visits. There were statistically significant differences in depressive and anxiety symptoms by site and by visit, although these were of small magnitude and their statistically significance may have been an artifact of sample size. There were site differences in alcohol use that may have reflected the recruitment of Buddhist monks.
It would seem that the multistage recruitment and screening process allowed individuals uncomfortable with participation to opt out before a full commitment to the trial was necessary.15 Only 7 volunteers dropped out of the full trial, whereas approximately one fifth of the sample recruited for screening formally dropped out or were lost to follow-up by not returning telephone or mail messages (see Thapinta et al15 for a fuller description of attrition during screening and prescreening phases).
There were anecdotal reports of unblinding through donation of blood (a practice that was discouraged by the staff), but there was no evidence of systematic unblinding. Concern has been raised that vaccine volunteers might unblind themselves, assume protection if vaccine-induced seropositivity was present, and engage in exposure risk behavior.3,4,8-10 As in US trials,40 volunteers often inferred that they had received vaccine because of the presence of side effects at or after injection. In general, these were common side effects of injection and could probably be attributed to the adjuvant that was present in the vaccine and the placebo. Nonetheless, these beliefs persisted despite efforts by staff to remind volunteers that that they, too, were blinded to assignment and that the participants' symptoms could result from placebo as well as from vaccine. This would have been a greater concern if there was evidence of increased risk behavior among the volunteers.
The social and behavioral consequences of participation were infrequent and minor. Their frequency was similar to that observed in an earlier smaller trial with a similar population.25 There was some evidence of social and behavioral benefits to participation. Education regarding vaccines and HIV was the most commonly cited benefit of participation. Smaller numbers of volunteers cited counseling and assistance with personal problems. Some volunteers maintained contact with study staff (on their own initiative) and helped to recruit participants for subsequent trials. Overall, data from this trial provide further evidence that HIV-1 preventive vaccine trials are feasible in Thailand.
The authors acknowledge the study volunteers who have made this report possible. They also express appreciation for the contributions made by the study physicians, Narongrid Sirisophana and Rudiwilai Samkoses; the clinical coordinators, Catherine Chaddic, Joseph Chiu, and Benjaluck Phonrat; the study nurses, Wantanee Boenim, Puangmalee Buapunth, Wipada Cheewawat, Suchada Chinawporapong, Supa Naksrisook, Somsri Ruckphaopunt, Nampueng Sirijongdee, Siriporn Suwankiti, Vanida Tubtong, Tasanai Vonchak, Chanmet Suwanarach, Attaporn Triampon, and Yupadee Yutabootr; the data analyst, Prasong Srisaengchai; and the support staff, Wisut Lokpichat, Boontrika Boonchoom, and Chuancheun Plavooth. Helpful editorial feedback was provided by Lorrie Gavin, Brad Bartholow, Jim Carey, Tim Dondero, and Ron Stall. Nipapat Upom assisted with the language translation of qualitative data. Institutional support for the study was provided by Pricha Singaraj, Thomas Brewer, and Choo-Chat Khambu na Ayuddhya from AFRIMS; Deborah Birx, John McNeil, and Donald Burke from Walter Reed Army Institute of Research; and Anne-Marie Duliege and colleagues at Chiron Vaccines.
1. Bloom BR. Ethical issues in AIDS vaccine development. AIDS Patient Care STDs. 1999;13:273-279.
2. Cohen J. Shots in the Dark. The Wayward Search for an AIDS Vaccine. New York: WW Norton; 2001.
3. Frey SE. Unique risks to volunteers in HIV vaccine trials. J Investig Med. 2003;51(Suppl 1):S18-S20.
4. Grady C. The Search for an AIDS Vaccine: Ethical Issues in the Development and Testing of a Preventive HIV Vaccine. Bloomington, IN: Indiana University Press; 1995.
5. Gray JN, Lyons PM, Melton GB. Ethical and Legal Issues in AIDS Research. Baltimore: Johns Hopkins University Press; 1995.
6. Kilmarx PH, Ramjee G, Kitayaporn D, et al. Protection of human subjects' rights in HIV-preventive clinical trials in Africa and Asia: experiences and recommendations. AIDS. 2001;15(Suppl 5):73-79.
7. Loue S. Legal and Ethical Aspects of HIV-Related Research. New York: Plenum; 1995.
8. Lurie P, Bishaw M, Chesney MA, et al. Ethical, behavioral, and social aspects of HIV vaccine trials in developing countries. JAMA. 1994;271:295-301.
9. Ackers ML, Parekh B, Evans TG, et al. Human immunodeficiency virus (HIV) seropositivity among uninfected HIV vaccine recipients. J Infect Dis. 2003;187:879-886.
10. Sheon AR, Wagner L, McElrath MJ, et al. Preventing discrimination against volunteers in prophylactic HIV vaccine trials: lessons from a phase II trial. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;19:519-526.
11. Celentano DD, Nelson KE, Beyrer C, et al. Decreasing incidence of HIV and sexually transmitted diseases in young Thai men: evidence for the success of the HIV/AIDS control and prevention program. AIDS. 1998;12:F29-F36.
12. Jenkins RA, Temoshok LR, Virochsiri K, et al. Incentives and disincentives to participate in prophylactic HIV vaccine research. J Acquir Immune Defic Syndr Hum Retrovirol. 1995;9:36-42.
13. Jenkins RA, Torugsa K, Markowitz LE, et al. Willingness to participate in HIV vaccine trials among young Thai men. Sex Transm Infect. 2000;76:386-392.
14. van Griensven F, Keawkungwal J, Tappero JW, et al. Lack of increased HIV risk behavior among injection drug users participating in the AIDSVAX® B/E HIV vaccine trial in Bangkok, Thailand. AIDS. 2004;18:295-301.
15. Thapinta D, Jenkins RA, Morgan PA, et al. Recruiting volunteers for a large phase I/II HIV preventive vaccine trial in Thailand. J Acquir Immune Defic Syndr. 2002;30:503-513.
16. Maek-A-Nantawat W, Pitisuttithum P, Phonrat B, et al. Evaluation of attitude, risk behavior and expectations among Thai participants in phase I/II HIV/AIDS vaccine trials. J Med Assoc Thai. 2003;86:299-307.
17. Phanuphak P, Teeratakulpixarn S, Sarangbin S, et al. International clinical trials of vaccines: I. Phase I trial of an HIV-1 synthetic peptide vaccine in Bangkok, Thailand. Asian Pac J Allergy Immunol. 1997;15:41-48.
18. Pitisuttithum P, Migasena S, Laothai A, et al. Risk behaviours and comprehension among intravenous drug users volunteered for HIV vaccine trial. J Med Assoc Thai. 1997;80:47-50.
19. Bartholow BN, Buchbinder S, Celum C, et al. HIV sexual risk behavior over 36 months of follow-up in the world's first HIV vaccine efficacy trial. J Acquir Immune Defic Syndr. 2005;39:90-101.
20. McElrath MJ, Montefiori D, Wolff M, et al. Safety, immunity, and risk behavior in HIV-1 uninfected volunteers representing diverse risk populations following recombinant envelope vaccinations: a three-year follow-up [abstract A284]. Presented at: XI International Conference on AIDS; 1996; Vancouver.
21. Chesney MA, Chambers DB, Kahn JO. Risk behavior for HIV infection in participants in preventive HIV vaccine trials: a cautionary note. J Acquir Immune Defic Syndr Hum Retrovirol. 1997;16:266-271.
22. Pitisuttithum P, Nitayaphan S, Thongcharoen P, et al. Safety and immunogenicity of combinations of recombinant subtype E and B human immunodeficiency virus type 1 envelope glycoprotein 120 vaccines in healthy Thai adults. J Infect Dis. 2003;188:219-227.
23. Keefer MC, Wolff M, Gorse GJ, et al. Safety profile of phase I and II preventive HIV type 1 envelope vaccination: experience of the NIAID AIDS Vaccine Development Group. AIDS Res Hum Retroviruses. 1997;13:1163-1177.
24. Nitayaphan S, Khamboonruang C, Sirisophana N, et al. A phase I/II trial of HIV SF2 gp120/MF59 vaccine in seronegative Thais. Vaccine. 1999;18:1448-1455.
25. Thapinta D, Jenkins RA, Celentano DD, et al. Evaluation of behavioral and social issues in Thai HIV vaccine trial volunteers. J Acquir Immune Defic Syndr. 1999;20:308-314.
26. Jenkins RA, Chinaworapong S, Morgan PA, et al. Motivation, recruitment, and screening of volunteers for a phase I/II HIV preventive vaccine trial in Thailand. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;18:171-177.
27. Nelson KE, Celentano DD, Suprasert S, et al. Risk factors for HIV infection among young adult men in Northern Thailand. JAMA. 1993;270:955-960.
28. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;16:297-334.
29. Sittitrai W, Phanuphak P, Barry J, et al. Thai Sexual Behavior and Risk of HIV Infection: a Report of the 1990 Survey of Partner Relations and Risk of HIV Infection in Thailand. Bangkok: Program on AIDS, Thai Red Cross Society; 1992.
30. Thongthai V, Guest P. Thai sexual attitudes and behaviour: results from a recent national survey. Presented at: Conference on Gender and Sexuality in Modern Thailand; 1995; Canberra, Australia.
31. Knodel J, Vanlandingham M, Saengtienchai C, et al. Thai views of sexuality and sexual behaviour. Health Transit Rev. 1996;6:179-201.
32. Nelson KE, Eiumtrakul S, Celentano DD, et al. HIV infection in young men in northern Thailand, 1991-98: increasing role of injection drug use. J Acquir Immune Defic Syndr. 2002;29:62-68.
33. Mills S, Benjarattanaporn P, Bennett A, et al. HIV risk behavioral surveillance in Bangkok, Thailand: sexual behavior trends among eight population groups. AIDS. 1997;11(Suppl 1):S43-S51.
34. Catania JA, Gibson DR, Chitwood DD, et al. Methodological problems in AIDS behavioral research: influences on measurement error and participation in studies of sexual behavior. Psychol Bull. 1990;108:339-362.
35. Vanlandingham M, Knodel J, Saengtienchai C, et al. Aren't sexual issues supposed to be sensitive? Health Transit Rev. 1994;4:85-90.
36. Turner CF, Ku L, Rogers SM, et al. Adolescent sexual behavior, drug use, and violence: increased reporting with computer survey technology. Science. 1998;28:867-873.
37. van Griensven F, Supawitkul S, Kilmarx PH, et al. Rapid assessment of sexual behavior, drug use, HIV, and sexually transmitted diseases in Northern Thai youth using audio-computer-assisted self-interviewing and noninvasive specimen collection. Pediatrics
. 2001;108:E13. Available at: http://pediatrics.aappublications.org/cgi/content/full/108/1/e13
38. Knodel J, Soottipong R, Saengtienchai C. AIDS Awareness and Knowledge in Thailand: Quantitative and Qualitative Evidence (Population Study Center Publication No. 97-380). Ann Arbor: Population Study Center, University of Michigan; 1997.
39. Celentano DD, Beyrer C, Natpratan C, et al. Willingness to participate in AIDS vaccine trials among high risk populations in northern Thailand. J Acquir Immune Defic Syndr Hum Retrovirol. 1995;9:1079-1083.
40. Johnson MO. Belief in vaccine receipt in HIV vaccine trials: further cautions. J Acquir Immune Defic Syndr. 1999;21:413-416.
© 2005 Lippincott Williams & Wilkins, Inc.