TOPICAL MICROBICIDES WILL BE female-controlled products that could be used within the vagina to reduce the risk of acquisition or transmission of sexually transmitted infections (STIs). Currently, topical microbicides are in development, and once they are available, they will need to be used correctly and consistently if they are to impact the STI epidemic. Thus, they need to be acceptable to potential users. Although the outcome of acceptability is correct and consistent use, individuals' views on product characteristics are multidimensional and incorporate many features of the product and of its use. The results of previous research have suggested that key characteristics might be ease of insertion, messiness, level and type of side effects (e.g., itching, burning, leakage), formulation, and effects on sexual pleasure.1–12 Our previous research focusing specifically on acceptability of use for adolescents suggested that there were aspects of acceptability that could be grouped into 5 concepts. These were pragmatic/convenience, ancillary effects, pleasure, efficacy, and interpersonal/intrapersonal relationships.
The importance of these dimensions of acceptability will vary based on cultural and developmental characteristics of the individual. For example, cultural attitudes about vaginal wetness vary. In a study examining attitudes of women in the United States and South Africa, the women enjoyed the lubricating properties of the gel, but it was especially enjoyable to the women of South Africa.10 Even within one country such as the United States, women of different race/ethnicities differ in their level of acceptability for these products.13 Non-Hispanic white women had a higher level of acceptability than all other races/ethnicities. Within one cultural group, individuals of differing ages or experiences (i.e., STI risk) may vary in the acceptability of products. For example, a previous study found that younger girls were less accepting of messiness and a visible residue left on the penis as a result of a microbicide compared with older women.8 Another study found that women who were at risk for HIV were more accepting of a microbicide product, even if there were numerous side effects.2
The development of a psychometrically sound scale measuring acceptability could extend previous research by providing a method of assessing these dimensions and furthering our understanding of how women make decisions. The scale then could be used across studies and across microbicide candidates.
Adolescent girls in the United States represent a group that may have unique cultural and developmental needs and are a target group for microbicide use given their rates of STIs.13 Thus, the purpose of the current study was to use these key concepts to develop a scale that could be used to examine adolescent girls' initial perceptions of microbicides and to determine if these initial perceptions varied based on age, race/ethnicity, and STI history.
Adolescent girls, ages 14 through 21 years, were recruited through U.S. school-based health clinics and local colleges and through snowball sampling (participants referring other girls) to participate in a 6-month study of acceptability and use of a microbicide surrogate. During recruitment to the study, the girls were told that we were enrolling girls who had had sexual intercourse, but they did not need to be planning to have intercourse in the next 6 months. After the girls and parents signed the consent form, the girls were interviewed face-to-face at intake, 3 months, and 6 months and completed weekly phone interviews. The girls received $30 for each interview completed and $3 for every phone interview. The current study is based on responses to the intake interview.
During the intake, the participants were told that scientists were developing products that would be used to protect against STIs. They were shown 2 vaginal moisturizers, a gel, Replens Vaginal Moisturizer (Warner Wellcome, Morris Plains, NJ), and a suppository, Lubrin Inserts (Bradley Pharmaceuticals, Fairfield, NJ), and given directions on how to use the products. The girls then were told that they could choose one or both formulation(s). After making their decision, the participants were asked to respond to a 22-item questionnaire for which they indicated likelihood (or degree) on a 5-point Likert scale for the product they had chosen. When given the questionnaire, the participants were told to think about one of the products, either the gel or the suppository. For the first 160 participants, if the girl chose both products, the interviewer alternated to which product the girl responded. Preliminary analyses indicated no difference in responses based on product formulation, so the remaining participants were allowed to choose the product about which they responded.
The study was approved by the Institutional Review Board at the University of Texas Medical Branch in Galveston. Consent was obtained from those girls who were over 18 years of age, and parental consent and child assent were obtained from girls under 18 years of age.
Focus groups were conducted with adolescent girls (7 groups), mothers of adolescent girls (6 groups), healthcare providers (3 groups), and medical students (3 groups).10,11 Participants' perceptions of topical microbicides were elicited through focus groups, which were used to promote the interaction and exchange of ideas on a topic that was a novelty for most participants. Based on the analysis of the focus group data, questions for each of the 5 salient concepts described were developed. These were as follows: pragmatic/convenience (ease of putting it in, affordability, messiness, ease of use in comparison to condoms); ancillary effects (“wetness” of vagina, penile irritation, vaginal irritation, getting stuck in one place), pleasure (intercourse interruption, pleasure of intercourse for user and partner, feels better than condoms); efficacy (sexually transmitted disease prevention, vaginal coverage and protection, vaginal health); and interpersonal/intrapersonal relationships (partner's attitude, control of health, perceived as sexually easy, or cheating).
Based on these concepts, a 22-item questionnaire, with items designed to match these categories, was developed (3–4 items per scale). The questionnaire then was given to 6 colleagues who were asked to classify each item according to the 5 concepts. In general, there was good concordance between the experts and our a priori classifications. However, the experts often placed an item in more than one category, suggesting that there might be some difficulty discriminating these concepts. Given the consensus, we left the scale intact. The ratings by each expert are presented in Table 1.
One of the 208 girls (14–21 years) enrolled in the study had to leave the interview in the middle before completing this questionnaire; thus, the final sample size for this analysis is 207 girls. The 207 adolescents had a mean age of 18.18 years (standard deviation = 1.93) and a median age of 18.3 years. The sample was 41% black, 29% non-Hispanic white/other, and 30% Hispanic. Sixty adolescents (29%) reported a history of an STI.
One hundred sixty girls responded to the questions thinking of the gel, and 47 girls responded to the questions thinking of the suppository. Using general linear models and a modified Bonferroni approach14 to maintain the type I error rate at 0.05 across 22 tests, none of the items met criteria for significance. All further analyses were conducted using a data set that combined answers to the gel and suppository.
The 22 items were placed in a principal components analysis with a varimax rotation.15 An item was treated as relevant to a factor if it had a loading in the rotated factor pattern above 0.3. For those items with loadings >.30 on more than one factor, the primary (largest) factor loading had to be greater than 0.5 and a secondary loading less than 0.31 for the item to be assigned to a factor. Items yielding less separation between the primary and largest secondary loading were considered factorially ambiguous and thus not assigned to any factor. A combination of criteria based on the magnitude of eigenvalues and interpretability of resulting factors was used to determine how many factors to retain.16 The initial solution, based on the Guttman-Kaiser criterion of eigenvalues exceeding 1.0,15 was an 8-factor solution. Two items (the gel/suppository would make me seem “easy” and the gel/suppository will get stuck in one place) did not have clear factor loadings; the eighth factor consisted of only one item (the gel/suppository will interrupt intercourse). In addition, our experience with the interview process suggested that the girls were not relating to the hypothetical nature of the question regarding affording the gel/suppository, and in this factor analysis, it loaded on the seventh factor with an item that did not appear to be conceptually related (the gel/suppository covering and protecting your whole vagina). Based on this review of the interpretability of the results of this first factor analysis, a decision was made to drop 4 items (the gel/suppository would make me seem “easy,” would get “stuck in one place,” and would interrupt intercourse, and “I could afford the gel/suppository would be) and to rerun the factor analysis using 18 items. The factor analysis with 18 items resulted in a 6-factor solution based on both the eigenvalue criterion and factor interpretability (see Table 2). The final 6 factors and the standardized Cronbach alphas for the items they comprised were as follows: factor 1: comparison to condoms (0.77); factor 2: negative impact on interpersonal relationships (0.69); factor 3: health benefits (0.56); factor 4: negative effects (0.63); factor 5: positive impact on pleasure (0.85), and factor 6: comfort while using (0.65).
The mean scores for each of the 6 factors were created by summing the scores on each item in that factor and then dividing by the number of items for that factor. These results suggested that overall, the girls viewed the product as more favorable than condoms (4.06; 4 = “likely”) as unlikely to have a noteworthy negative impact on their interpersonal relationships (1.81; 2 = “unlikely”), to be pleasurable (4.05; 4 = “likely”), and to be comfortable (4.38; 4 = “likely”). The girls were more neutral in the expected impact on their health (3.67; 4 = “likely”) and negative effects (2.60; 3 = “neither unlikely nor likely”).
To evaluate how demographic characteristics of the girls were related to their perceptions of the products, we examined the relationship of chronologic age, race/ethnicity, and history of STI to mean factor scores. Chronologic age was evaluated by correlation coefficients, and race/ethnicity and history of STI was evaluated by analysis of variance. The results indicated that age was positively correlated with the mean comparison to condom factor score (r = 0.30, P <0.01) and the comfort while using factor score (r = 0.14, P = 0.04) and negatively correlated with negative impact on interpersonal relationships factor (r = − .21, P< 0.01). There was no association between age and the other factor scores.
Using analysis of variance, race/ethnicity was significantly related to mean comparison to condom scale score (F = 3.59, P = 0.03) and the negative impact on interpersonal relationships scale score (F = 4.21, P = 0.02). Follow-up analysis (Student-Newman-Keuls tests) showed that white girls had significantly higher mean scores on the comparison to condom factor score than both black and Hispanic girls. Black and Hispanic girls had higher mean scores on the negative impact on interpersonal relationships factor than did the white girls. Race/ethnicity was not related to any other factor scores.
In this sample, there were significant differences in age based on race/ethnicity (F = 5.18, P <0.01) with those adolescents who were white/other being significantly older than the blacks and Hispanic adolescents. Because exactly the same 2 scale scores were significant for both age and race/ethnicity, the analyses were rerun with both predictors. When age was placed first in the model, race/ethnicity no longer accounted for a significant amount of the variance in either the comparison to condom score (F = 2.00, P = 0.14) or to negative impact on interpersonal relationships score (F = 2.49, P = 0.09).
Using analysis of variance, those with a history of an STI had significantly higher mean scores on the health benefits scale (F = 4.49, P = 0.04), positive impact on pleasure factor (F = 4.91, P = 0.03), and the comfort while using factor (F = 5.17, P = 02). A history of STI was not related to any other factor scores. Because mean comfort while using score also was related to age, the relationship between a history of STI and the mean score on the comfort while using scale was reanalyzed controlling for the age effects. When age was placed first in the model, a history of an STI no longer accounted for a significant amount of the variance (F = 3.36, P = 0.07).
If microbicides are to make a difference in the STI epidemic for adolescent girls, the girls must find the product acceptable for use. For girls to use the product the first time, they must perceive the product in a favorable manner before use. The results of this study demonstrated that, in general, adolescent girls felt positively about the product and the potential interpersonal and physical effects of use.
The results indicated that when responding to the scale based on a product formulation they had chosen, adolescent girls responded similarly across formulations. This is consistent with previous acceptability research that demonstrated that individuals will value similar qualities when choosing a product, but these individuals ascribe those characteristics to different formulations.4–7,17,18 This may suggest that to reach the highest number of users, a variety of formulations eventually will be needed. However, given the fact that more girls chose the gel, a gel formulation may be the most appropriate first step.
Healthcare practitioners can use the 6 concepts identified by the factor structure to initiate conversations with potential adolescent users. The factor loadings were not exactly as originally planned, suggesting there may be more to learn about how adolescent girls will approach potential microbicide use. In the meantime, these concepts (comparison to condoms, negative impact on interpersonal relationships, health benefits, negative effects, positive impact on pleasure, comfort while using) could provide a guide for ensuring that salient concepts are reviewed with potential users. Concerns then could be addressed either by providing additional information or strategies for management. It is possible that perceptions will differ postuse and that the factor composition may change. For example, literature supports that individuals from some societies may find the lubricating quality of a microbicide to be more beneficial. However, the amount of lubrication during sex varies also by individual preference as well as cultural context.2,8,9,19
For example, in this preuse analysis, increased wetness appeared to be viewed as impacting comfort. However, qualitative data suggest that after girls use the product, they view the additional lubrication as having an impact on the comfort of intercourse and therefore, increasing their pleasure.20 Similarly, “messy” loaded on the factor with irritation of the penis and vagina. Although “messy” is a word used often in the assessment of microbicide acceptability, it is not clear exactly what participants mean by this word, and it is possible that it represents a variety of experiences.
Age of the girl was related to responses on some of the factors. The older girls in this study were more likely to see these products in a more positive light than condoms, to have comfort regarding the product, and to be less likely to view it as having a negative impact on interpersonal relationships. Younger girls may have viewed the product more negatively, especially related to these areas, because they lack experience in both sexual behavior and partner communication. Further studies will be needed to identify the underlying developmental issues leading to these perceptions; however, it does suggest that younger adolescent girls may need some additional preuse guidance and counseling if they are going to be willing to try microbicides.
There are a number of limitations to the study design that should be considered when evaluating the results. This was a convenience sample of adolescent girls who were willing to participate in a study using a surrogate topical microbicide that was not effective in STI or pregnancy prevention. The sample was unique in that all of those under 18 years of age had parental consent, and all of the participants were willing to partake in a study in which they were asked very explicit questions about their sexual behaviors. Thus, these girls are likely to be those who are relatively comfortable with their sexuality. Among some groups of potential users, an important aspect of the product may be the possibility covert use. This may be most relevant aspect of the product in cultures in which male partners would not want their partners to use any methods.4 A question specifically addressing covert use was not included in this measure, because among adolescent girls in the United States, the concerns appear to focus on the relationship aspects (the likelihood that the partner would think poorly of them) rather than specific covert use. An item related to covert use might be important to include in future development of this scale.
It also is difficult to know how using the product “to help the scientists” impacted perceptions. In addition, the data were collected from face-to-face interviews. The adolescents might have been less willing to share negative perceptions of the microbicides and might have been less than forthcoming with regard to information such as their STI history. Furthermore, the participants were asked to complete the scale thinking of a hypothetical product. This may have been difficult for some of the girls as a result of the developmental and experiential level of the girls. Furthermore, the scale was not designed to assess differences in perceptions across formulations.
The results of this study suggested that it should be possible to develop a scale that will provide information regarding critical attitudes toward microbicide acceptability. To be able to determine which products are most acceptable and to develop targeted interventions to enhance use, it will be important to have a scale with sound psychometric properties that could be used to distinguish across products and between potential users. Future research should refine the measure further to determine if it predicts actual use. In addition, the scale will need to be examined for the appropriateness of its use with individuals from other cultural contexts, including older women and those from developing countries.
1. Bentley ME, Fullem AM, Tolley EE, et al. Acceptability of microbicide among women and their partners in a 4-country phase I trial. Am J Public Health 2004; 94:1159–1164.
2. Bentley ME, Morrow KM, Fullem A, et al. Acceptability of a novel vaginal microbicide during a safety trial among low-risk women. Fam Plann Perspect 2000; 32:184–188.
3. Darroch JE, Frost JJ. Women's interest in vaginal microbicides. Fam Plann Perspect 1999; 31:16.
4. Green G, Pool R, Harrison S, et al. Female control of sexuality: Illusion or reality? Use of vaginal products in south west Uganda. Soc Sci Med 2001; 52:585–598.
5. Hammett TM, Mason TH, Joanis CL, et al. Acceptability of formulations and application methods for vaginal microbicides among drug-involved women: Results of product trials in three cities. Sex Transm Dis 2000; 27:119–126.
6. Hardy E, Jimenez AL, de Padua KS, Zaneveld LJD. Women's preference for vaginal antimicrobial contraceptives III. Contraception 1998; 58:245–249.
7. Hardy E, de Padue KS, Osis MJD, Jimenez AJ, Zaneveld LJD. Women's preferences for vaginal antimicrobial contraceptives IV. Contraception 1998; 58:251–255.
8. Mason TH, Foster SE, Finlinson HA, et al. Perspectives related to the potential use of vaginal microbicides among drug-involved women: Focus groups in three cities in the United States and Puerto Rico. AIDS Behav 2003; 7:339–351.
9. Morrow K, Rosen R, Richter L, et al. The acceptability of an investigational vaginal microbicides, PRO 2000 gel, among women in a phase I clinical trial. J Women Health 2003; 12:655–666.
10. Short MB, Mills L, Majkowski JM, Stanberry LR, Rosenthal SL. Topical microbicide use by adolescent girls: Concerns about timing, efficacy, and safety. Sex Transm Dis 2003; 30:854–858.
11. Short MB, Mills L, Majkowski JM, Stanberry LR, Rosenthal SL. Adolescent issues associated with knowledge of and access to topical microbicides. J Women Health 2004; 13:1129–1138.
12. Weeks MR, Mosack KE, Abbott M, Sylla LN, Valdes B, Prince M. Microbicide acceptability among high-risk urban US women: Experiences and perceptions of sexually transmitted HIV prevention. Sex Transm Dis 2004; 31:682–690.
13. Tracking the Hidden Epidemics: Trends in STDs in the United States 2000. Atlanta: Centers for Disease Control and Prevention, 2000.
14. Holland BS, Copenhaver M. Improved Bonferroni-type multiple testing procedures. Psychol Bull 1988; 104:145–149.
15. Nunnally JC, Bernstein IH. Psychometric Theory, 3rd ed. New York: McGraw-Hill, 1994.
16. Schuster C, Yuan K. Factor analysis. In: Kempf-Leonard K, ed. Encyclopedia of Social Measurement. Boston: Elsevier, 2004:1–8.
17. Hardy E, de Padua KS, Jimenez AL, Zaneveld LJD. Women's preferences for vaginal antimicrobial contraceptives II. Contraception 1998; 58:239–244.
18. Minnis AM, Shiboski SC, Padian NS. Barrier contraceptive method acceptability and choice are not reliable indicators of use. Sex Transm Dis 2003; 30:556–561.
19. Smith EJ. Will vaginal microbicides be acceptable. Network 2002; 22:24–26.
20. Zubowicz EA, Oakes JA, Short MB, Perfect MM, Succop PA, Rosenthal SL. Adolescents' descriptions of the physical characteristics of microbicide surrogates and experiences of use. J Womens Health. In press.