The seroincidence of herpes simplex virus type 2 (HSV-2) infection was determined among 1766 patients attending sexually transmitted disease (STD) clinics and enrolled in a randomized, controlled trial of human immunodeficiency virus (HIV)/STD risk-reduction counseling (RRC). Arm 1 received enhanced RRC (4 sessions); arm 2, brief RRC (2 sessions); and arm 3, the control arm, brief informational messages. The overall incidence rate was 11.7 cases/100 person-years (py). Independent predictors of incidence of HSV-2 infection included female sex; black race; residence in Newark, New Jersey; <50% condom use with an occasional partner; and, in females, incident trichomoniasis and bacterial vaginosis. Only 10.8% of new HSV-2 infections were diagnosed clinically. Incidence rates were 12.9 cases/100 py in the control arm, 11.8 cases/100 py in arm 2, and 10.3 cases/100 py in arm 1 (hazard ratio, 0.8 [95% confidence interval, 0.6–1.1], vs. controls). The possible benefit of RRC in preventing acquisition of HSV-2 infection offers encouragement that interventions more specifically tailored to genital herpes may be useful and should be an important focus of future studies.
Herpes simplex virus (HSV) type 2, the main cause of genital herpes, is a growing public health concern [1, 2]. Serial serosurveys have established that the prevalence of HSV-2 has been increasing in the United States, with 22% of the US population 112 years old infected by the early 1990s . The prevalence in high-risk populations, such as patients attending sexually transmitted disease (STD) clinics, is even higher: >40% [4–6]. Furthermore, there is mounting evidence that HSV-2 facilitates both transmission and acquisition of HIV infection . Despite these concerns, the incidence and predictors of new HSV-2 infections are not well characterized, and, more importantly, public health interventions with proven effectiveness to prevent genital herpes are limited . Experts at an advisory meeting to develop a national agenda for prevention of genital herpes called for urgent evaluation of new strategies to prevent sexual transmission of HSV [1, 8].
Project RESPECT was a randomized, controlled trial of HIV/STD risk-reduction counseling (RRC) aimed at reducing new STDs by increasing safer sexual behaviors, particularly by increasing condom use . This trial, conducted among patients attending public STD clinics, demonstrated that RRC was effective in increasing self-reported condom use and decreasing new bacterial STDs by 20% over the course of 12 months. However, it is unclear whether RRC aimed at increasing condom use would be effective in preventing viral STDs, such as HSV-2 infections. HSV-2 can be shed from both skin and mucosal surfaces over a wide area of the genital region [5, 10], and, thus, transmission could occur outside the protective area of a condom. However, recent studies have shown for the first time that condoms can prevent transmission of genital herpes [11, 12]. The purpose of the present analysis was to use data from Project RESPECT to determine the incidence and predictors of HSV-2 infection in patients attending STD clinics and to evaluate the effectiveness of RRC in prevention of new HSV-2 infections.
PATIENTS AND METHODS
Study design and procedures. We performed a secondary analysis of prospective data from Project RESPECT, for which the complete methodology has been reported elsewhere . In brief, the trial was conducted from July 1993 through September 1996 at public STD clinics in Baltimore, Maryland; Denver, Colorado; Long Beach, California; Newark, New Jersey; and San Francisco, California. English-speaking patients ⩾14 years old who presented for STD examinations during the study period and who had had vaginal intercourse during the previous 3 months were eligible to participate in the study. Homosexual males were excluded from the study, as were patients found to be HIV positive at enrollment. Overall, 43% of eligible patients enrolled in the trial . The Project RESPECT protocol was reviewed and approved by the institutional review board at each participating site, and all participants gave written, informed consent. The human-experimentation guidelines of the US Department of Health and Human Services and those of participating institutions were followed in the conduct of this research.
Participants were randomly assigned to 1 of 3 counseling intervention arms. The interventions and their quality-assurance process evaluations have been described in detail elsewhere [9, 13]. Participants in arm 1 received enhanced RRC, consisting of 4 personalized interactive sessions based on the Theory of Reasoned Action and Social Cognitive Theory [14–16]. Session 1 was 20 min, and the remaining sessions were 60 min each. Participants in arm 2 received brief RRC, consistent with the Centers for Disease Control and Prevention's client-centered HIV Prevention Counseling model [17, 18]. Brief counseling consisted of two 20-min interactive sessions. Both the enhanced and the brief RRC interventions focused on safer sexual behaviors in general; neither was specifically tailored to prevention of genital herpes. Participants in arm 3, the control arm, received informational safe-sex messages during two 5-min sessions, which was typical of the care in most STD clinics at the time the study was designed  and is still typical in many settings today. Participants were scheduled for quarterly followup visits over the course of the next 12 months, which were calculated from the date of enrollment to occur at 3-month intervals .
At enrollment, participants were interviewed to obtain information on sociodemographic characteristics and lifetime sexual and STD history. At the baseline, 3−, 6−, 9−, and 12-month visits, participants were also interviewed about sexual behavior and STD history during the previous 3 months. No data were available with regard to previous symptoms suggestive of genital HSV (for example, undiagnosed genital blisters or ulcerations) or contact with a sex partner with genital herpes. Participants were asked at each visit, “In the past 3 months, have you been told by a doctor or nurse that you had genital herpes?” Genital examinations and laboratory testing for STDs, including serologic testing for HSV-1 and HSV-2, were performed for all participants at the baseline, 6−, and 12-month visits. In addition to scheduled study visits, all participants were encouraged to return for clinical examinations whenever they had genital symptoms or concerns about STD exposure, although HSV serologic or virologic testing was not routinely performed at these unscheduled visits.
Laboratory methods. Stored serum samples were tested for antibodies to HSV-1 and HSV-2 by use of a type-specific strip immunoblot assay (SIA; Chiron) based on recombinant antigen bands for glycoproteins gB1, gD2, gG1, and gG2 . The sensitivity of the SIA has been found to be 98% for HSV- 2 and 95% for HSV-1, with a specificity of 99% for both, compared with Western blot analysis . All tests were performed in the same laboratory, by use of the manufacturer's automated processor. For participants with signs or symptoms of genital herpes, a diagnosis was made in each STD clinic on the basis of the examiner's clinical impression of the participant's history and physical findings, without the standardized use of confirmatory HSV culture or other virologic testing. Diagnostic criteria for Neisseria gonorrhoeae and Chlamydia trachomatis have been described elsewhere . Infection with Trichomonas vaginalis in females was defined by trichomonads seen on a wet mount from a vaginal swab (all sites) or a positive culture (Denver and Baltimore). Bacterial vaginosis was diagnosed by use of standard clinical criteria .
Definition of incident infections. To be included in the present analysis, participants had to be HSV-2 seronegative and without a clinical diagnosis of genital herpes at enrollment. The acquisition of HSV-2 infection was defined by seroconversion (i.e., new detection of HSV-2 antibody) at either the 6- or the 12-month visit. For purposes of analysis, the midpoint between the last negative test result and the first positive test result was used as the date of acquisition. Symptomatic genital herpes was not included in the definition of acquisition of HSV-2 infection, since participants were not educated specifically about the signs or symptoms of genital herpes, and clinicians were not provided uniform protocols for viral culture testing of lesions suggestive of genital herpes. Among participants who experienced HSV- 2 seroconversion, we did evaluate the proportion who received a clinical diagnosis of genital herpes during a study visit or reported a diagnosis from an outside clinic.
Participants with gonorrhea or chlamydia at enrollment were treated with standard therapy  and advised about the importance of treatment of their sex partners . Incident infection with gonorrhea or chlamydia was defined as the first laboratory diagnosis of each infection during 1 of the scheduled follow-up visits or during a voluntary (unscheduled) visit. The date of the first positive follow-up test result was considered to be the date of acquisition of symptomatic infection. For asymptomatic infection with 1 of these bacterial STDs, the midpoint between the first positive follow-up test result and the previous test was used as the date of acquisition.
Statistical analysis. Incidence rates were calculated as the number of new infections per person-time of follow-up, with 95% confidence intervals (CIs) obtained by a normal approximation to the Poisson distribution. Incidence rates were compared by use of a Cox proportional hazards model. Crude and adjusted ratios of incidence rates (hazard ratios [HRs]) were calculated for both baseline data and time-varying covariates. Because serum samples were drawn at baseline and at the 6- and 12-month follow-up visits, there were two 6-month intervals that began and ended with an HSV-2 assay. Quarterly interviews asked about behavior during the previous 3 months, so data for each 6-month interval were the sum of the 2 interviews during that interval. Therefore, in proportional hazards models including time-varying covariates, only participants who attended visits at all 3 time points during either the first interval (baseline, 3, and 6 months) or the second interval (6, 9, and 12 months) could be included, and data were analyzed using these 6-month blocks. Interval censoring was accounted for by use of an exact distribution for ties . Independent predictors of acquisition of HSV-2 infection were determined by use of a multivariate Cox proportional hazards model. Separate models for males and females were used to assess variables obtained from only one sex. An additional multiple logistic regression model was created among only participants who experienced HSV-2 seroconversion, to determine independent predictors of remaining undiagnosed with genital herpes despite acquisition of new HSV-2 infection. Comparisons of incidence of HSV-2 infection among intervention arms were performed in an intention-to-treat fashion. Statistical analyses were performed using SAS software (version 8.1; SAS Institute). All reported P values are 2 sided.
Of 4128 Project RESPECT participants from whom baseline serum samples were available, 1686 (41%) were already HSV-2 seropositive , and another 77 were excluded because of a clinical diagnosis of genital herpes at enrollment. Of the 2365 remaining participants, 599 (25%) did not return for any follow- up serologic testing, leaving a final effective study population for this analysis of 1766 participants. A total of 37% of the study population were female, and 52% of the participants were black. The median age of the participants was 24 years, and the median number of lifetime sex partners was 14. There were no significant differences by arm for any baseline characteristics.
Incidence of HSV-2 infection. Overall, there were 185 new cases of HSV-2 infection during 1587 person-years (py) of follow-up, yielding an incidence rate of 11.7 cases/100 py (95% CI, 10.0–13.3 cases/100 py) (table 1). During the first 6-month study interval, 119 (7.5%) of 1587 participants experienced HSV-2 seroconversion. During the second 6-month interval, 42 (3.5%) of 1210 participants experienced seroconversion. There were also 179 participants who attended only the baseline and 12-month visits, 24 (13.4%) of whom experienced HSV- 2 seroconversion; it is unclear during which 6-month study interval these infections were acquired. Females had a higher incidence rate than did males (14.8 vs. 9.9 cases/100 py; P=.008 ), and black participants had a higher rate than did non- black participants (15.1 vs. 8.1 cases/100 py; P < .001). There was a bimodal distribution of incidence of HSV-2 infection according to age; participants <25 years old and those>40 years old had the highest HSV-2 infection incidence rates, whereas those 25–29 years old had the lowest incidence rate. Incidence of HSV-2 infection also varied by geographical location, educational level, and previous HSV-1 antibody status (table 1).
Independent predictors of incidence of HSV-2 infection. We also performed a multivariate analysis of predictors of incidence of HSV-2 infection on the 1495 participants who had at least one 6-month interval of complete behavioral data that began and ended with an HSV-2 serologic test, creating a total of 2510 person-6-month intervals of observation (table 2). After controlling for other risk factors, the risk of acquiring HSV- 2 infection was almost twice as high in females as in males and in black participants as in nonblack participants. Participants at the Newark clinic were also more likely than participants at other sites to acquire HSV-2 infection. Previous infection with HSV-1 did not protect against incident HSV-2 infection; in fact, HSV-1 infection at baseline was associated with a slightly increased risk of acquisition of HSV-2 infection, although this was not statistically significant.
The only sexual behavior characteristic associated with new HSV-2 infection was unprotected sex with an occasional partner. Participants who used condoms 100% of the time with an occasional partner had an incidence of HSV-2 infection similar to those with no occasional partners, and, thus, these groups were combined for further comparisons. Compared with this referent group, participants who used condoms 50%–99% of the time with an occasional partner had a slightly higher risk of acquiring HSV-2 infection, although this was not statistically significant. Participants who used condoms <50% of the time with occasional partners were twice as likely to acquire HSV- 2 infection. The protective effect of condoms in preventing acquisition of HSV-2 infection from occasional partners was not different between males and females (P=.98). There was Pp.98 no association between incidence of HSV-2 infection and the number of new or ongoing sex partners, number of vaginal sex episodes, level of condom use with main partners, or incident bacterial STDs.
Some variables were available only for females or for males, and these additional factors were evaluated by sex in separate multivariate models. Among females, after controlling for other risk factors, incident trichomoniasis and bacterial vaginosis also predicted incidence of HSV-2 infection (table 2). Females with newly diagnosed trichomoniasis during the follow-up period were almost 4 times as likely to acquire HSV-2 infection, and those with bacterial vaginosis were twice as likely. Among males, no additional factors predicted incidence of HSV-2 infection, including circumcision status (table 2).
Proportion of new HSV-2 infections diagnosed with genital herpes. Of the 185 participants who developed incident HSV- 2 infection during the study period, all had genital examinations by experienced clinicians, but only 20 (10.8%) received a clinical diagnosis of genital herpes, and 165 (89.2%) remained undiagnosed. Among participants with new HSV-2 infections, genital herpes was more likely to remain undiagnosed among black participants than among nonblack participants (93.4% vs. 81.0%; P=.01) and among HSV-1-seropositive participants than among HSV-1-seronegative participants (92.3% vs. 78.6%; P=.01). Males were slightly more likely to remain undiagnosed than were females, although not significantly so (93.1% vs. 84.5%; P=.01). Among males, circumcision status was not associated with a clinical diagnosis of genital herpes, with 96.3% of uncircumcised men remaining undiagnosed, compared with 91.9% of circumcised men (P=AA). On multivariate analysis, only black race (odds ratio [OR], 4.2 [95% CI, 1.5–11.4]) and HSV- 1 seropositivity (OR, 4.2 [95% CI, 1.5–11.6]) predicted remaining undiagnosed with genital herpes among those acquiring new infections.
Incidence of HSV-2 infection by intervention arm. Figure 1 shows HSV-2 infection incidence rates by counseling intervention arm, compared with the incidence rates of first new infections with chlamydia and gonorrhea by arm. The overall incidence rate was highest for HSV-2 infection (11.7 cases/100 py), followed by chlamydia (8.9 cases/100 py), and then gonorrhea (6.2 cases/100 py). Assessing incidence of HSV-2 infection by intervention arm, the incidence rate was 12.9 cases/100 py for participants in the control arm, 11.8 cases/100 py for participants in arm 2 (HR, 0.9 [95% CI, 0.7–1.3], vs. controls), and 10.3 cases/100 py for participants in arm 1 (HR, 0.8 [95% CI, 0.6–1.1], vs. controls). The same pattern was seen for all 3 STDs with respect to counseling intervention arm, with a progressive decrease in incidence from control to brief counseling to enhanced counseling (figure 1). However, the decrease in incidence by arm was greater for chlamydia and gonorrhea than it was for HSV-2 infection. The effect of RRC on acquisition of HSV-2 infection did not significantly differ by sex. After controlling for baseline characteristics—including sex, race, age, study site, and HSV-1 status—the HRs for incidence of HSV- 2 infection by arm did not substantially change (arm 1 vs. controls: HR, 0.7 [95% CI, 0.5–1.1]).
The data from our analysis represent one of the largest evaluations of incidence of HSV-2 infection in patients attending STD clinics and the first published evaluation of the benefit of an HIV/STD RRC intervention on the prevention of genital herpes. The incidence of HSV-2 infection was high in this population, with 11.7 new cases/100 py of follow-up, which was higher than the incidence of either chlamydia or gonorrhea. The HSV-2 infection incidence rate in the present study was substantially higher than the rate of 5.1 cases/100 py seen in a trial of an ineffective HSV-2 vaccine (Chiron), which studied 1862 patients attending STD clinics and 531 couples discordant for HSV-2 infection [23, 24]. A portion of this discrepancy could be explained by an inclusion criterion in the Chiron trial, which required that discordant couples had been sexually active for at least 6 months . Thus, couples in the stage of their relationship with the highest risk of HSV-2 transmission would not have been included in the trial [25, 26]. A larger portion of the difference in observed incidence rates may relate to the fact that, unlike our study, the vaccine trial provided consistent counseling about genital herpes (e.g., counseling about the signs and symptoms of genital herpes and the need to avoid intercourse when lesions are present) . Greater awareness of potential signs and symptoms of genital herpes in partners could encourage avoidance of sexual activity during symptomatic outbreaks and thus decrease rates of acquisition of HSV-2 infection.
Similar to the investigators of other studies [23, 27], we have found that females were almost twice as likely to acquire HSV- 2 infection as were males. In addition, the females in the present study were more likely to acquire HSV-2 infection if they had incident trichomoniasis or bacterial vaginosis. The precise timing of these conditions with respect to acquisition of HSV-2 infection is unknown; nonetheless, these associations may shed light on the greater risk of incident HSV-2 infections in females. Both trichomoniasis and bacterial vaginosis have been found to be risk factors for acquisition of HIV infection in prospective studies [28, 29], and at least 1 prospective study has linked bacterial vaginosis with incident HSV-2 infection . Alterations in the vaginal microenvironment seen with these conditions may contribute to an increased susceptibility to HSV- 2, as they are thought to do for HIV . Since trichomoniasis and bacterial vaginosis represent alterable conditions that may account for a portion of the increased risk of acquisition of HSV-2 infection in females, their association with HSV-2 deserves further study.
We have also found a higher incidence rate of HSV-2 infection in black participants than in nonblack participants. Numerous cross-sectional studies have reported a higher prevalence of HSV-2 infection in black participants [3, 4, 6, 32], and the Chiron vaccine trial found a higher incidence of HSV-2 infection in nonwhite participants than in white participants [11, 23]. To the extent that sex partnerships tend to develop among members of the same race or ethnic group , with each new partnership, black persons may have a higher risk of coming into contact with an infected partner, because of the higher prevalence among black persons. In addition, as noted in both our baseline study  and the present study, HSV-2 infection is more likely to remain undiagnosed among black persons than among nonblack persons. Therefore, black persons may not only be more likely to encounter partners with HSV-2 infection, but those partners may also be less likely to know they are infected and, thus, to take precautions to prevent transmission.
In the present study, previous HSV-1 infection did not protect against subsequent acquisition of HSV-2 infection. These results are consistent with findings from the Chiron vaccine trial , 2 prospective trials among females [30, 34], and an evaluation of nationally representative seroprevalence data from 1988 to 1994 . However, conflicting results have been found in other studies. In a large vaccine trial  and in 2 smaller trials [27, 37], previous HSV-1 infection was associated with reduced acquisition of HSV-2 infection. Regardless of its effect on acquisition of HSV-2 infection, several studies, including ours, have found that previous HSV-1 infection decreases the likelihood that new HSV-2 infections will be clinically diagnosed [23, 35, 36].
The only behavioral risk factor for acquisition of HSV-2 infection identified in our analysis was unprotected sex (<50% condom use) with an occasional partner. Unprotected sex with an occasional partner was associated with incident bacterial STDs in another analysis of data from Project RESPECT , which was similarly unsuccessful in finding strong associations between other sexual behaviors and new bacterial STDs. We had no information on whether a participant's partner was infected with HSV-2. Regardless of behavior, there is no risk of acquiring HSV-2 infection from uninfected partners; thus, including participants whose partners are uninfected would tend to obscure any relationship between behavior and acquisition of HSV-2 infection, since it would reduce precision in measuring these associations. This concept is highlighted by a recent subanalysis of Project RESPECT: when the infection status of sex partners was unknown, no relationship was found between bacterial STDs and condom use; yet, when only participants with known infected partners were analyzed, a significant reduction in bacterial STDs was found with consistent condom use .
The vast majority of new HSV-2 infections in the present study were not diagnosed clinically. The proportion of new infections that remained undiagnosed under the routine clinical conditions in the present study (89%) was much higher than that in the Chiron trial (63%) , in the control group of a recent vaccine trial (64%) , and in a study of pregnant females (64%) . Project RESPECT participants were not explicitly asked about signs and symptoms of genital herpes between study visits. Thus, there may have been participants who recognized symptoms of genital herpes but never received a diagnosis. However, our estimates are likely to be reflective of the actual proportion of HSV-2 infections that would remain undiagnosed in clinical settings outside of genital herpes research protocols. In all of the above studies [23, 34, 36], subjects were counseled about herpes signs and symptoms, whereas our subjects were counseled about safer sexual behaviors but not specifically about herpes. Thus, the different proportions of undiagnosed infections among these studies suggest that persistent herpes-specific counseling can have a positive effect on recognition of new HSV-2 infections.
Finally, the trend toward lower incidence of HSV-2 infection among those participants receiving enhanced RRC is a promising observation. Project RESPECT showed that client-centered RRC not only reduced risky behavior but also decreased incidence of bacterial STDs by 20% over the course of 12 months . The present analysis has found that RRC interventions that decreased incidence of bacterial STDs showed a similar pattern with respect to HSV-2 infection. Of importance, 43% of the initial Project RESPECT population had evidence of HSV-2 infection at enrollment and, therefore, could not be included in the present analysis. With the remaining sample size, we had only 20% power to detect a 20% difference between study arms. Additionally, it is possible that some participants were infected with HSV-2 before enrollment in the study and were already in the process of seroconverting at the time of the counseling interventions, since it may take up to several weeks after infection to seroconvert . Twice as many HSV-2 seroconversions occurred during the first 6 months than during the second 6 months, raising concern that some early infections were acquired before enrollment in the study. This would bias the comparison of the counseling interventions toward the null hypothesis that there is no difference in acquisition of HSV-2 infection between the study arms. Thus, with respect to prevention of acquisition of HSV-2 infection, we may have underestimated the benefits of RRC in the present analysis. Although not definitive, our findings suggest that RRC interventions may have additional benefits beyond their effect on bacterial STDs.
The present study had some limitations. Since the trial was conducted in 5 inner-city public STD clinics, these data are not applicable to other, perhaps lower-risk, populations. In addition, the study population consisted of selected patients, representing only 43% of those eligible, who enrolled in a randomized, controlled trial; thus, these findings may not be generalizable to all patients attending STD clinics. Finally, because this was a secondary analysis, we were unable to assess certain herpes-specific risk factors, such as whether a participant's partner had signs or symptoms of genital herpes.
Nonetheless, our findings offer observations important to the development of HSV-2 infection prevention strategies. First, the high risk of acquisition of HSV-2 infection in this population, especially among females and black participants, emphasizes the need for new prevention efforts to stem the transmission of genital herpes . Second, the association of trichomoniasis and bacterial vaginosis with acquisition of HSV-2 infection may contribute to higher incidence of HSV-2 infection among females and deserves future study. Third, the finding that condom use with occasional partners reduced the risk of acquisition of HSV-2 infection supplements the limited evidence available regarding the effectiveness of condoms in preventing genital herpes [12, 26, 41]. Finally, the high rate of undiagnosed HSV-2 infection in this population is cause for concern, since it likely contributes to ongoing transmission. Studies have shown that most patients with previously asymptomatic HSV-2 infection can be taught to recognize subsequent symptomatic outbreaks [42, 43], and preliminary data indicate that knowledge of genital herpes infection may be associated with delayed transmission to a partner . Thus, a logical approach to HSV-2 infection prevention might involve HSV-2 screening of selected populations, counseling for recognition of genital herpes symptoms in those persons with unrecognized infection, and subsequent HSV-targeted counseling in those persons who recognize their symptoms . The present study has evaluated non-herpes-specific RRC that focused on preventing acquisition of STDs, whereas the above herpes-specific approach would focus on preventing transmission. Nonetheless, the possible benefit of RRC for the prevention of acquisition of HSV- 2 infection seen in the present study offers encouragement that interventions more specifically tailored to genital herpes may be useful and should be an important focus of future studies.
PROJECT RESPECT STUDY GROUP
The members of the Project RESPECT Study Group are as follows. Baltimore: Carolyn Erwin-Johnson, Andrew L. Lentz, Mary A. Staat, Dawn Sweet, and Jonathan M. Zenilman (principal investigator [PI]). Denver: John M. Douglas, Jr., (PI), Tamara Hoxworth, Ken Miller, and William McGill. Long Beach: Ruth Bundy (co-PI), Laura A. Hoyt, C. Kevin Malotte, and Fen Rhodes (PI). Newark: Michael Iatesta, Eileen Napolitano (co- PI), Judy Rogers, and Ken Spitalny (PI). San Francisco: Gail A. Bolan (PI), Coleen LeDrew, Kimberly A. J. Coleman, Luna Hananel, and Charlotte K. Kent. NOVA, Bethesda: Robert Francis (PI), Christopher Gordon, Nancy Rosenshine, and Carmita Signes. Centers for Disease Control and Prevention: Sevgi Aral, Robert H. Byers, Beth Dillon, Martin Fishbein, Sandra Graziano, Mary L. Kamb, William Killean, James Newhall, Daniel Newman, Thomas A. Peterman, and Karen L. Willis.
We are grateful to the study participants who volunteered for Project RESPECT and to our counselors, clinical staff, and research associates in Baltimore, Denver, Long Beach, Newark, and San Francisco. We thank Rae Lynn Burke (Chiron) for provision of the strip immunoblot reagents and equipment and Denise Brown (Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention [CDC]) for performance of the laboratory assays. We are also indebted to Jim Dobbins (Division of Viral and Rickettsial Diseases, CDC) for his assistance in study concept and design and to Mike Hoekstra (Division of Bacterial and Mycotic Diseases, CDC) for his statistical expertise.