This study was conducted according to the principles expressed in the Declaration of Helsinki, and was approved by the Institutional Review Boards of the Population Council, Centers for Disease Control and Prevention, and the Ethical Review Committee of the Thailand Ministry of Public Health. Written informed consent was obtained from all study participants. The protocol for this trial is available as supporting information; see Protocol S1.

Acceptability data were collected as part of a 3-month, randomized, placebo-controlled crossover trial conducted to evaluate the safety and acceptability of Carraguard among HIV-infected women (hereafter referred to as “crossover trial”). Each woman participated in each of 3 study interventions (Carraguard, methylcellulose placebo, and no product) over three consecutive months with the order of interventions randomly assigned. When using study gels, women were blinded to the type of gel they received. Carraguard and placebo are both clear and odorless gels packaged in a single-use Microlax®-type applicator (Norden Pac International AB, Kalmar, Sweden). This applicator was shown in previous studies to dispense about 5 ml of gel [11]. Participants attended the clinic at day 0, 7, and 14 of each study month for interview, examination and specimen collection. They were asked to apply study gel vaginally once daily for 7 days starting after menses were completed (day 0) during the two gel months. Full methods and safety findings are reported in detail elsewhere [12].

Women were recruited from HIV care clinics and support groups in five districts of Chiang Rai province in northern Thailand. Eligibility criteria included aged 18-50 years; CD4+ cell count 51–500/mm³; not currently on antiretroviral therapy (ART); and either being sexually abstinent or having a single male partner who was confirmed HIV-infected and aged 18 years or older. At study initiation, eligibility for ART in public clinics was limited; women were referred for ART at the time of study screening and encouraged to initiate ART as soon as available, regardless of study participation. During the study, eligibility expanded substantially and all eligible women accessed ART during or shortly after the study.

After completing both gel-use periods, women completed an interviewer-administered questionnaire that included both structured and open-ended questions about the participants' views on product attributes and use of the gels. Most of these questions addressed both gels simultaneously, as previous studies had shown no difference in acceptability responses between Carraguard and placebo groups [13], [14]. However, the previous studies were traditional parallel arm studies in which participants were randomized to either Carraguard or placebo. Since this crossover trial was the first time that individual women used both Carraguard and placebo, additional questions asked women to compare the two gels they had used. At the time these questions were asked, the women did not yet know which of the two gels was Carraguard and which was placebo. Adherence was assessed by participant self-report in face-to-face interviews and by counting returned used and unused applicators (returned in separate bags).

Data were double-entered into an Access database (Microsoft, Inc., Redmond, Washington, USA). Codes for open-ended questions were initially developed by trial staff and were reviewed and revised by staff members at the Population Council in New York City. Data were analyzed using SAS statistical software package (SAS version 9.1, SAS Institute, Cary, North Carolina, USA). Chi-square tests were used to evaluate differences between groups.

Some acceptability questions in the crossover trial were asked in an identical manner in two previous studies of Carraguard safety and acceptability at the same site in Thailand; for these questions, comparisons are presented here. The first study was conducted to assess safety and acceptability among 165 low-risk, HIV-uninfected women followed for one year (hereafter referred to as “phase II trial”)[14], [15]. Participants were asked to apply Carraguard or placebo once per day in the evening, and if sex took place, within one hour before sex together with condoms; 98% of participants were married and sexually active. The second study enrolled 55 low-risk, HIV-uninfected, sexually active couples for 6 months, primarily to assess safety of exposure during intercourse and acceptability among male partners; safety and acceptability among female participants were also evaluated [13], [16]. Couples were asked to apply Carraguard or placebo within one hour before sex; condoms and counseling were provided but condoms were rarely used in the trial. We present acceptability responses from female participants in the phase II and couples trials concurrently with results of the crossover trial for comparison.

Between 2002 and 2004, 222 women were screened and 60 were enrolled in the trial; one woman was lost to follow up after she completed the acceptability assessment. Enrolled women had a median age of 34 years, had been diagnosed with HIV a median of 44 months previously, and had a median CD4+ lymphocyte count of 296 cells/mL (Table 1). Seventeen (28%) women were married, with the remainder either widowed (60%) or divorced (12%). The 15 (25%) women who were sexually active at the time of the trial reported having had sex a median of 4 times in the previous month, and median condom use in the previous month was 4 times. No participants reported anal or oral sex in the previous month.

Reported adherence to gel use was high, with 59 (98%) women using at least 6 of 7 scheduled doses during each gel use period. Only one woman interrupted gel use for reasons unrelated to adverse effects.

Overall satisfaction rating of the gels was high, with 22 (37%) and 30 (50%) participants reporting they liked the gel “very much” and “somewhat,” respectively, and 60 (100%) stating they would recommend it to a friend (Table 2). Most participants described the applicator as appealing (72%) or neutral (25%), and found the gel easy to apply (83%). The gel volume was assessed as “just right” by 46 (77%) women and “too much” by 14 (23%), while 55 (92%) women considered the extra lubrication provided by gel an advantage. Only 17 (28%) women believed that it would be possible to use gel without their partner being aware. In response to open-ended questions about how they would change the product if they could, 45 women had no suggested changes, 12 suggested softening the applicator tip, and 10 suggested a smaller gel volume.

Acceptability responses varied among the three Thai safety trial populations for most characteristics (Table 2). For example, there was a significant difference among groups in the overall satisfaction rating of gel; more women in the crossover trial said they “liked the gel very much” compared to women in the other two studies. There were also differences in the proportion of women considering covert use possible, and judging the volume of gel as too much. For most parameters, participants in the crossover trial evaluated the gels more favorably than those in the other two trials.

Crossover trial participants rated the following characteristics as most important in a microbicide: the partner being unable to tell gel is there (77% very important, 15% somewhat important), being personally unable to tell gel is there (73% very important, 13% somewhat important), providing extra lubricant (58% very important, 37% somewhat important), and making the vagina feel tight (57% very important, 30% somewhat important) (Table 3). There were statistically significant differences among the 3 trial populations for only three of the ten microbicide characteristics assessed: extra lubricant, giving a warm feeling, and having a good taste.

Almost half of crossover trial participants (48%) reported that they noticed a difference between the Carraguard and placebo gels (Table 4). Among the 29 women who reported a difference, 15 reported on open-ended questioning that placebo gel leaked more, 6 that placebo felt more viscous, 5 that Carraguard felt more natural, and 4 that Carraguard felt more viscous. A total of 20 (33%) women preferred Carraguard while 7 (12%) preferred placebo (p = 0.01), with no significant difference by order of gel received. Most women who preferred Carraguard over placebo did so because they perceived that it was less messy (12 responses) and felt natural (9 responses) (Table 5). More women on placebo reported leakage or messiness (16 responses) as their least liked feature.

When asked what role microbicides might play for them as HIV-infected women, participants most commonly responded that microbicides could protect them from infection with another strain of HIV or other STI (51 responses), and that it might protect their partner from HIV acquisition (13 responses). On the other hand, when asked, “If microbicides provided the same protection as condoms, do you think men would be less willing to use condoms?” 57 (95%) women responded yes. A majority of women (68%) reported being willing to pay at least the equivalent price as condoms for a microbicide.