We evaluated data from Project RESPECT, a multicenter, randomized, controlled trial conducted from 1993 through 1996. Project RESPECT evaluated the efficacy of 3 HIV prevention counseling models in increasing condom use and preventing incident HIV and STDs. 1 Participants were recruited from STD clinics in Baltimore, Denver, Long Beach, Newark, and San Francisco. Heterosexual, HIV-negative men and women aged 14 or older who visited one of the clinics for an STD examination and agreed to an HIV test were eligible to participate. Participants were randomly assigned to 1 of the 3 HIV prevention strategies: a brief 2-session client-centered model, an enhanced 4-session client-centered model, and 2 sessions of brief informational messages.
Once enrolled and randomly assigned a prevention strategy, participants met with the next available counselor. Clients were not matched with counselors according to gender or race. For participants assigned to brief client-centered counseling, at the initial visit, the counselor and client assessed the client’s risk and perception of risk, identified barriers to risk reduction, and negotiated a risk-reduction plan. At this visit, counselors helped clients identify and commit to one step they could take to reduce their risk. At the last session, scheduled for 7 to 10 days later when HIV test results would be available, the counselor delivered test results and reviewed progress with the plan set forth at the initial visit and, if appropriate, negotiated a new plan.
Participants assigned to enhanced prevention counseling were asked to attend a 4-session intervention based on behavioral prediction and change theory. 12 The first session was identical to the brief counseling intervention. The remaining 3 sessions took place over the next 3 weeks and lasted approximately 1 hour each. These sessions were designed to address issues around attitude change, self-efficacy, and perceived norms regarding safe sex and included such skills as latex condom use, attitudes toward condom use, self-efficacy toward condom use, and perceived norms concerning condom use. 12 At each session, an incremental risk-reduction plan was developed.
Incentives were offered to participants in Project RESPECT. Free condoms were offered at every visit and $15 for each session attended after the first session (ie, for enhanced prevention counseling, a maximum of $45; and for brief prevention counseling and informational messages, a maximum of $15). 1
To ensure quality counseling sessions, Project RESPECT counselors attended a required intervention training course; were provided detailed, partially scripted protocols; and were regularly observed and received feedback. Additionally, participants completed semistructured postintervention surveys regarding their perceptions of the experience and activities. 12
Project RESPECT demonstrated that both brief and enhanced counseling increased condom use and prevented new STDs compared with a third intervention group that received informational messages. 1 STD reduction in brief and enhanced counseling intervention groups were similar. Our analysis includes participants who were randomly assigned to either brief or enhanced counseling and completed at least one follow-up STD examination over the ensuing 12 months.
Intervention completion was defined as clients who attended all counseling sessions to which they were assigned (ie, enhanced prevention counseling: one 20-minute session and 3 hourly sessions; brief prevention counseling: two 20-minute sessions).
The incident STDs measured for this study included chlamydia, gonorrhea, syphilis, trichomoniasis (women only), and HIV. Incident STDs were defined by laboratory tests, with chlamydia defined as a positive Chlamydia trachomatis polymerase chain reaction test from an endocervical specimen for women or a urine specimen for men. Gonorrhea was defined as a positive culture for Neisseria gonorrhoeae or, for men, Gram-negative intracellular diplococci seen with Gram stain of a urethral swab. Syphilis was defined by a suggestive history and physical examination with supportive treponemal and nontreponemal antibody test results. Trichomoniasis (women only) was defined by detection of typical motile trichomonads on microscopic examination of a saline-suspended vaginal sample. HIV was defined by a repeatedly reactive enzyme immune assay for HIV antibody with a positive confirmatory test result. Each participant was counted only once in the dataset, and only the first infection was included in this analysis.
After univariate analysis, we computed adjusted odds ratios using generalized estimating equations (GEE) using Proc GENMOD in SAS Version 8.02 (SAS Institute, Inc., Cary, NC). Our computations allowed for correlated outcomes because each counselor could affect multiple outcomes. A GEE model allows for fixed and time-varying covariates. Fixed covariates in this analysis included the variables age, study site, gender, and having an STD at baseline. Time-varying covariates included counselor experience. This technique also was used to examine groups of clients and counselors that shared the same demographics (gender, race, and both gender and race). All multivariate analyses were adjusted for study site and for demographics and characteristics of the client (race, gender, age, type of counseling, STD at baseline, education); counselor (degree, race, gender, and experience); and counselor–client dyad (same counselor, same race, same gender, and same race and gender). Results were described as statistically significant when P values were less than or equal to 0.05.
Demographics and Characteristics of the Sample
The 32 counselors were racially and ethnically diverse (41% white, 41% black, and 13% Hispanic). Most were female (69%). Most (59%) of the counselors had a college education, an additional 19% had attended graduate school, and the remaining 19% had completed high school. Each counselor conducted between 2 and 731 HIV/STD risk reduction counseling sessions in Project RESPECT, with a mean of 223.6 and median of 188. Counseling experience was defined as having conducted at least 75 sessions in Project RESPECT; 72% (n = 23) of the counselors were classified as experienced.
Of the 2885 participants randomized to receive brief or enhanced counseling, 2239 (78%) had at least one STD examination over 12 months and were eligible for inclusion in this analysis. Of these clients, 58% were black, 19% were white, and 16% were Hispanic. A little over half (56%) of the participants were male. Nearly half (47%) of the participants were between the ages of 20 and 29 years, 16% were between the ages of 19 and 14, and 37% were 30 or older. Only 8% of the clients had graduated from college, 24% had some college education, 41% had a high school diploma, and 27% had not finished high school. Most clients (97%) had the same counselor for all prevention counseling sessions.
Demographics and Characteristics Associated With Intervention Completion
Most clients (88%; n = 2239) completed the intervention. Failure to complete the intervention was slightly higher for black clients (13%) than Hispanic (11%) or white (8%) clients. Clients younger than 20 were less likely to complete the intervention (22%) compared with clients between the ages of 20 and 29 (12%) or older than 30 (6%). Clients who had an STD at the baseline examination were less likely to finish the intervention (17%) than were clients who did not have an STD at baseline (9%). Clients who did not complete the intervention were more likely to acquire a new STD infection (31%) by the 12-month follow-up visit than were clients who completed the intervention (19%).
Counselor demographics and characteristics were not associated with client failure to complete the intervention. A similar number of clients failed to complete the intervention whether their counselor was black 12%, Hispanic 11%, white 11%, or other 6%. Compared with white counselors, black counselors were slightly more likely to have clients fail to complete the intervention (adjusted odds ratio [ORa], 1.6; 95% confidence interval [CI], 0.9–2.9). Male and female counselors had similar numbers of clients who did not complete the intervention. Counselor education and experience were not associated with client failure to complete the intervention (Table 1).
In 54% of all counseling sessions, the client and the counselor were of the same race. There were 889 black dyads, 49 Hispanic dyads, and 212 white dyads. In 47% of the sessions, the client and the counselor were the same gender, with 724 female dyads and 289 male dyads. Client and counselor were of the same race and gender in 26% of the counseling sessions. No dyad demographics (same race, same gender, same gender and race) were associated with client failure to complete the intervention (Table 1).
Multivariate analysis found most predictors of client failure to complete the intervention were demographics and characteristics of clients rather than counselor characteristics. Client demographics and characteristics associated with failure to complete the intervention included intervention type, age, STD at baseline examination, and education. Failure to complete the intervention was more likely for clients who were randomly assigned to enhance 4-session counseling (ORa, 3.0; CI, 2.2–4.0 compared with brief 2-session); aged 14 to 19 (ORa, 3.8; CI, .2.4–5.9 compared with 30 and older) or 20 to 29 (ORa, 2.4; CI, .1.6–3.5 compared with 30 and older); had an STD at the baseline examination (ORa, .2.2; CI, .1.6–2.9 compared with no STD at baseline examination); and had less than a high school education (ORa, .1.8; CI, .1.2–2.9 compared with some college) or a high school education (ORa, .1.7; CI, .1.2–2.6 compared with some college).
Demographics and Characteristics Associated With New Sexually Transmitted Diseases
Nearly one third of the clients (31%; n = 700) had an STD at enrollment, and 20% of the clients (n = 448) had received a diagnosis of a new STD by the 12-month visit. New STDs were more commonly diagnosed among black (27%) clients compared with Hispanic (13%) or white (9%) clients. More female than male clients returned with a new STD (25% vs. 16%). Clients with less education than high school were more likely to return with a new STD (27%) than were clients with a high school education (22%), some college education (14%), or a college degree (5%).
Black counselors had 1 in 4 of their clients return with a new STD within 12 months, slightly more than counselors of other races. However, this difference did not exist after multivariate analysis (ORa, 1.1; 95% CI, 0.6–1.9) (Table 2). Female counselors had 19% of their clients return with a new STD, whereas male counselors had 25% return with a new STD (ORa, 0.9; CI, 0.7–1.2) (Table 2). A majority of the counselors had a college education (n = 19; 59%). Clients of counselors with a high school education (ORa, 0.8; CI, 0.4–1.6) or college education (ORa, 0.9; CI, 0.5–1.6) were no more likely to acquire a new STD infection than were clients of counselors that had a graduate education. Inexperienced counselors were no more likely to have a client develop a new STD than counselors who had conducted at least 75 sessions (ORa, 1.1; CI, 0.8–1.5). Most clients had the same counselor for all sessions (97%) so we could not assess the effect of changing counselors on the likelihood of becoming infected with a new STD. No dyad characteristic was associated with clients acquiring a new STD in Project RESPECT (Table 2). Clients who did not complete the intervention were more likely to develop a new STD (ORa, 1.7; CI, 1.2–2.4).
Multivariate analysis found most predictors of new STDs were demographics and characteristics of the client rather than counselor. 14 Client demographics and characteristics associated with a new STD included study site, race, gender, education, and having an STD at the baseline examination. Persons were more likely to develop a new STD if they lived in Newark (ORa, 2.7; CI, 1.4–4.9, compared with San Francisco), were black (ORa, 2.0; CI, 1.3–3.2 compared with all other races), were female (ORa, 2.2; CI, 1.7–2.9), had less than a high school education (ORa, 1.7; CI, 1.2–2.4 compared with some college) or a high school education (ORa, 1.5; CI, 1.1–2.0 compared with some college), and had an STD at baseline (ORa, 2.4; CI, 1.9–3.0).
We did not identify specific counselor or counselor–client dyad demographics or characteristics that influenced the effectiveness of prevention counseling. Variables associated with failure to complete the intervention and acquiring new STDs were more client-related than counselor-related, including client race, gender, education, and having an STD at baseline. 14 Clients who completed the intervention were less likely to acquire a new STD infection, as would be expected because the original study 1 showed the intervention was effective. Efforts to encourage counseling completion such as client incentives, follow-up reminders, and convenient hours should be explored. However, HIV prevention counseling should focus on risk reduction and not on asking the patient to return for the next session.
Matching clients and counselors for HIV prevention counseling sessions could be more likely to appeal to a client’s preference than to result in a direct client benefit. Other studies in general counseling research have found counselor characteristics associated with positive client perception of the client–counselor relationship to include similar personality, attitude, and socioeconomic background; similar age or older; and more education. 8,10,13,15,18 In other studies, clients most strongly preferred counselors with similar gender and ethnicity. 16 Given this similarity, clients like their counselors more, perceive the counselor–client relationship as more helpful, and perceive their counselor as more understanding and empathic. 8,10,13,16 Ethnic similarity is especially preferred by black and Mexican-American clients and has been associated with commitment to remain in an intervention. 10,13,15,16 Although clients could use similarities to establish trust with their counselor, these similarities are not directly related to outcome. 13 Counselors could be perceived as more credible when they exhibit crosscultural knowledge, and this would likely be an important variable for intervention effectiveness as opposed to variables such as shared counselor and client ethnicity. 17 Matching clients and counselors based on race/ethnicity could still result in cultural mismatches if a client and counselor from the same ethnic group have different values, lifestyles, and experiences.
Our study had some limitations. First, clients were randomly assigned to intervention groups and not to counselors. Once enrolled and assigned to an intervention, participants met with the next available counselor without concern for the client or counselor’s gender or race. Second, Project RESPECT was limited to HIV-negative heterosexual clients attending STD clinics. Although this population is at high risk, findings might not be generalizable to other populations such as men who have sex with men, intravenous drug users, commercial sex workers, or persons who are HIV-positive where peers influence outcomes. Third, Project RESPECT studied only one possible approach to HIV prevention counseling: client-centered HIV/STD prevention counseling models. In other approaches to HIV prevention counseling, such as small group counseling sessions or psychotherapy, counselor demographics could be more salient to the client. Fourth, the small number of counselors in the study limited our power to detect a statistically significant difference. Fifth, other characteristics not measured as part of Project RESPECT could contribute to HIV prevention counseling outcomes. Sixth, Project RESPECT did not systematically collect client satisfaction data. Seventh, we do not know about STDs in persons who did not return to the clinic.
Finally, quality assurance was an important part of Project RESPECT, but because it was so important and such an integral part of the prevention counseling models, it possibly made the counseling more homogenous than it would have been otherwise.
The client-centered counseling studied by Project RESPECT can prevent STD. 1 Although we did not identify specific counselor or dyad demographics or characteristics associated with acquiring a new STD or intervention completion, these results should not be interpreted as showing that other counselor characteristics are unimportant. There are several characteristics thought to be associated with counseling that we did not study, the most prominent of which are the abilities of the counselor. 4 HIV/STD prevention counseling programs should strive to hire and retain skilled counselors. A good prevention counselor believes counseling can make a difference and has a genuine interest in the counseling process. Good counselors are active listeners, able to use open-ended questions, negotiate rather than persuade, and foster a supportive atmosphere for the client. Good counselors feel comfortable discussing specific HIV risk behaviors such as explicit sex or drug behaviors. Lastly, good counselors remain focused on risk-reduction goals with the client. 4 When hiring and training counselors, we recommend focusing on these skills and not on matching counselors and clients by gender or race.
1. Kamb ML, Fishbein M, Douglas JM, et al. Efficacy of risk-reduction counseling to prevent human immunodeficiency virus and sexually transmitted diseases. JAMA 1998; 280: 1161–1167.
2. Shain RN, Piper JM, Newton ER, et al. A randomized controlled trial of a behavioral intervention to prevent sexually transmitted disease among minority women. N Engl J Med 1999; 340: 93–100.
3. Wolitski RJ, MacGowan RJ, Higgins DL, Jorgensen CM. The effects of HIV counseling and testing on risk-related practices and help-seeking behavior. AIDS Educ Prev 1997; 9: 52–67.
4. Centers for Disease Control and Prevention. Revised guidelines for HIV counseling, testing, and referral. MMWR Morb Mortal Wkly Rep 2001; 50: 1–53.
5. Centers for Disease Control and Prevention. Technical guidance on HIV counseling. MMWR Morb Mortal Wkly Rep 1993; 42( No. RR-2): 5–9.
6. Carkhuff RR, Berenson BG. Beyond Counseling and Therapy. New York: Rinehart and Winston, 1967.
7. Flaskerud JH. Matching client and therapist ethnicity, language, and gender: a review of the research. Issues in Mental Health Nursing 1990; 11: 321–336.
8. Atkinson DR. Ethnic similarity in counseling psychology: a review of research. Counseling Psychologist 1983; 11: 79–92.
9. Simons NW, Berkowitz NM, Moyer RJ. Similarity, credibility, and attitude change: a review and a theory. Psychol Bull 1970; 73: 1–6.
10. Atkinson DR, Furlong MJ, Poston WC. Afro-American preferences for counselor characteristics. Journal of Counseling Psychology 1986; 33: 326–330.
11. Jemmott JB, Jemmott LS, Fong GT, McCaffree K. Reducing HIV risk-associated sexual behavior among African American adolescents: testing the generality of intervention effects. Am J Community Psychol 1999; 27: 161–187.
12. Kamb ML, Dillon BA, Fishbein M, Willis KL, Project RESPECT study group. Quality assurance of HIV prevention-counseling in a multi-center randomized controlled trial. Public Health Rep 1996; 111( suppl 1): 99–107.
13. Beutler LE, Clarkin J, Crago M, Bergan J. Client–therapist matching. In: Handbook of Social and Clinical Psychology. Elmsford, NY: Pergamon Press, Inc, 1991: 699–716.
14. Peterman TA, Lin LA, Newman DR, et al. Does measured behavior reflect STD risk? An analysis of data from a randomized controlled behavioral intervention study. Sex Transm Dis 2000; 27: 446–451.
15. Atkinson DR, Poston C, Furlong MJ, Mercado P. Ethnic group preferences for counselor characteristics. Journal of Counseling Psychology 1989; 36: 68–72.
16. Lopez SR, Lopez AA, Fong KT. Mexican Americans 61 initial preferences for counselors: the role of ethnic factors. Journal of Counseling Psychology 1991; 38: 487–496.
17. Sue S. Psychotherapeutic services for ethnic minorities. Two decades of research findings. Am Psychol 1988; 43: 301–308.
18. Tall KM, Ross MJ. The effects of therapist-client age similarity on pretherapy expectations. Journal of Contemporary Psychotherapy 1991; 21: 197–209.