Home > Discussion - Behavioral Counseling to...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Lin J, Whitlock E, O'Connor E, et al. Behavioral Counseling to Prevent Sexually Transmitted Infections [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2008 Oct. (Evidence Syntheses, No. 64.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.


Summary of Review Findings

Variation in population risk and intervention intensity

We did not find any trials evaluating primary care feasible behavioral counseling interventions to prevent STIs in a general-risk (average) population. All the populations studied in the trials we identified address “at risk” population groups who have a higher than average STI incidence (e.g., adolescents and young adults, African Americans, Hispanics, mentally ill persons, mentally disabled persons, persons living in low-income urban areas, or persons with a previous or present history of an STI). For this discussion, therefore, we use the term low- risk to describe the general primary care population and high-risk to describe “at-risk” groups. We also stratify our discussion of populations based on trial setting (i.e., primary care setting versus STI clinic), as setting is another measure of population risk.

Given that there is a very large variation in the intensity of behavioral counseling interventions studied, we use the term low intensity to describe single-visit counseling interventions lasting less than 30 minutes, or any intervention that could be added to usual primary care without significant additional visit time. We use moderate intensity to describe interventions lasting longer than 30 minutes, but less than 2 hours in total. We use high intensity to describe multiple-visit interventions requiring longer than 2 hours.

Effects of behavioral counseling interventions on STI incidence and self-reported sexual behavior

STI incidence. Only two trials targeting high-risk populations in primary care settings used self-reported STI incidence as an outcome measure (one in adults, 15 and one in adolescents). 18 Neither of these trials reported an intervention effect. Both of these trials used low-intensity counseling interventions and had low proportions of persons self-reporting STI incidence at followup (less than four percent). One trial targeting a high-risk population in a psychiatric clinic used self-reported STI incidence as an outcome measure. 13 This trial by Carey et al. showed an eight percent reduction in self-reported STI at 6 months in the intervention arm, as compared to a two percent reduction in the matched intensity control arm (p<0.013).

Six trials targeting high-risk populations- three in adults, 12, 13, 16, 17 two in adolescents, 19, 20 and one in both 14 - used STI incidence as an outcome measure. All but one of these trials (five RCTs, n= 8384) showed a reduction in STI incidence. This one trial (n=393) targeted heterosexual adults attending an STI clinic. There was no reduction, however, in STI incidence in the high-intensity counseling intervention arm. 12

Two trials in predominantly adult women attending STI clinics, Project SAFE and Project SAFE 2, showed a similar reduction in laboratory diagnosed gonorrhea and chlamydia at 12 months, OR 0.52, 95 percent CI (0.34, 0.81) and OR 0.51, 95 percent CI (0.31, 0.83), respectively. Project SAFE 2 showed a persistent effect in risk reduction in bacterial STI incidence (OR 0.57, 95 percent CI (0.34, 0.96)) at 24 months in Project SAFE 2. 16, 17 Due to the heterogeneity amongst these trials, we caution the extrapolation of the effect estimated by Project SAFE 2, which showed a similar magnitude of risk reduction at 1 and 2 years. In fact, other studies suggest some decay of effect over time, although a benefit remains for cumulative risk reduction. 32 35

For adolescents, both trials using high-intensity group interventions in sexually active adolescent girls— one was conducted in an adolescent medicine clinic 20 and the other was conducted in a community health center 19 — showed a reduction in bacterial STIs at 12 months. In the trial by Jemmott et al., 20 10.5 percent of adolescents tested positive for chlamydia, gonorrhea, or trichomonas in the skills-based counseling intervention arm, compared to 18.2 percent in the control arm (adjusted p=0.05). In the trial by DiClemente et al., 19 there was a decrease in the incidence of chlamydia, gonorrhea, and trichomonas. This reduction, however, was only statistically significant for chlamydia, OR 0.17, 95 percent CI (0.03, 0.92), probably due to the low incidence of gonorrhea and trichomonas in this study.

One trial, Project RESPECT, which included both adults and adolescents attending STI clinics, used two individual-based interventions combined with HIV testing— a moderate-intensity intervention of two 20-minute sessions and a high-intensity intervention of four 50-minute sessions. Both the moderate- and high-intensity counseling interventions showed a similar decrease at 12 months in overall laboratory diagnosed STIs (gonorrhea, chlamydia, syphilis, and HIV), RR 0.78, 95 percent CI (0.64, 0.94) and RR 0.81, 95 percent CI (0.67, 0.98) respectively. 14 In an a priori subgroup analysis, this risk reduction appeared to be more exaggerated in adolescents (age 14 to 20 years), RR 0.57, 95 percent CI (0.37, 0.90). 21 In another a priori subgroup analysis, this risk reduction was not significant for HSV incidence(Appendix B Table 1). 36

There is evidence for the effectiveness of moderate-to high-intensity behavioral counseling to reduce the incidence of overall STIs (not including HSV) in high-risk populations, with more robust evidence for common bacterial STIs( e.g., gonorrhea and chlamydia). Two trials in high- risk populations conducted in primary care (n=1429) did not show a reduction in incidence of self-reported STIs using low-intensity interventions. We found no trials in low-or general-risk populations (Figure 2).

Figure 2. Intervention intensity vs. population risk.


Figure 2. Intervention intensity vs. population risk. a Adolescent only population (+) positive findings

Self-reported behavioral outcomes. Four of the six trials targeting high-risk populations in primary care settings only used self-reported sexual behaviors as outcome measures—three in adults 22 24, 37 and one in adolescents. 25

For adults, we found only one trial, Project FIO, that showed a beneficial effect of a high-intensity counseling intervention on self-reported sexual behaviors. 22 This trial was conducted in women attending a family planning clinic who had a relatively high baseline risk for STI (self-reported 58 percent with a history of an STI). This trial showed a beneficial effect for the nine-session (total 18 hours), group-counseling intervention, but not for the five-session (total 10 hours) intervention, compared to women receiving no intervention. At 12 months, women in the nine-session intervention group reported four fewer unprotected sexual acts during the past three months, p<0.001, and an 18 percent increase in self reported male or female condom of borderline statistical significance, p=0.06. 22

The other two trials in high-risk adult populations in primary care settings did not show an intervention effect, including: one trial using a low-intensity individual intervention in young adults attending family practices in Australia; 23 and one trial using a moderate-intensity individual intervention with and without HIV testing in young adults attending a university health center. 24 The one trial in an adolescent population in a primary care setting using a moderate-intensity individual intervention did not increase condom use or abstinence. 25

In general, there is no additional evidence for the effectiveness of low-, moderate-, or high-intensity behavioral counseling interventions in high-risk populations based on self-reported sexual behavior outcomes. One trial did show a benefit using a very-high-intensity intervention, but in retrospect this population's risk was similar to those studied STI clinics, although based in a family planning clinic. We identified no trials that met our inclusion criteria in low- or general-risk populations (Figure 2).

Effect of behavioral counseling interventions on other positive outcomes and adverse outcomes

In general, few trials report on other positive outcomes in addition to psychosocial mediators of behavior (e.g., knowledge, attitudes, self-esteem, and ability changes). For adults, high-intensity counseling interventions for women in STI clinics can also increase treatment compliance (i.e. decrease unprotected sex with untreated or incompletely treated sexual partners). For sexually active adolescents, high-intensity behavioral counseling may transiently increase general contraceptive use and decrease pregnancy.

Based on nine fair-to-good-quality trials, there appears to be no significant harm in primary care feasible counseling interventions for adults (i.e., there is no increase in risky behaviors or increase in STI incidence). For adolescents, risk reduction behavioral counseling may transiently increase vaginal sexual intercourse in young adolescents. However, the significance of this transient finding is unclear, given that no change in overall sexual activity or vaginal sexual activity by the end of the trial at 9 months. 18 Only one study reported on sexual debut. This study found that risk-reduction counseling did not increase sexual activity in those previously not sexually active. 25 We found no trials for risk-avoidance or abstinence-only counseling that met our inclusion criteria. Our findings are consistent with a recent meta-analysis that included all studies examining a deliberate HIV risk-reduction counseling intervention in a nonperinatal context, which found no inadvertent increase the number of sexual occasions or sexual partners. 38

Effect of behavioral counseling intervention elements

Given the heterogeneity amongst these trials and limited evidence for effective interventions, it is not possible to draw definitive conclusions about the differential effect of interventions on specific populations or the differential effect of specific intervention elements (e.g., theory and content, format, and intensity). Based on the existing body of evidence, however, it appears that population risk and intervention intensity are the biggest predictors of a counseling intervention's effect on STI incidence and self-reported behavior change. In adults, there is strong evidence for high-risk individuals, specifically for ethnically diverse populations, predominantly African American and Hispanic; low-income urban populations; populations with a high baseline prevalence of STIs or history of STIs (30–100 percent); and some evidence for persons with major psychiatric disease with comorbid recent history of substance abuse. There is also strong evidence for sexually active adolescents, and specifically for ethnically diverse and low-income urban populations.

Intervention intensity, more than format or a particular behavioral model, may also be an important factor in the effectiveness of counseling interventions. However, there were no low-intensity or single-visit counseling interventions used in the highest-risk populations. The range of intensity for effective interventions was 40 minutes delivered in two sessions with HIV testing, 14 to 10 sessions of unknown duration, 13 to 18 hours over nine sessions. 22 All effective interventions were based on individual risk-based counseling and included a tailored risk-reduction plan. Most of these interventions were developed with some amount of formative research within the targeted population. For adolescents, two of the effective interventions also included instruction on condom skills. In one study by Jemmott et al., only the condom skills intervention group showed an effect in STI reduction. 20 All effective interventions were based on common behavioral models, including the AIDS risk reduction model (ARRM), cognitive behavioral theories, harm reduction, stages of change theory and motivational techniques, theory of reasoned action, and social cognitive theory. However, individually tailored counseling and these behavioral models and other social theories were also the basis for interventions that did not show a risk reduction in STIs or behavioral change in moderate-risk populations. 15, 18, 25

Of the thirteen trials we reviewed, five of them used culturally tailored interventions. 16, 17, 19, 20, 22 Of the seven positive trials, six were conducted in predominantly nonwhite populations, and five of these six interventions were tailored to be culturally sensitive. Project RESPECT had about 84 percent nonwhite participants, but was not explicitly culturally tailored. Without direct comparisons of culturally tailored interventions versus nontailored interventions, the significance of this observation is uncertain. These trials also matched counselor gender, ethnicity, or both to participants, and one trial used adolescent peer co-facilitators. Based on Project RESPECT, however, Pealer et al. report that there is no significant association between intervention completion or new STIs and counselor characteristics (e.g., gender, ethnicity, education, and counseling experience), or counselor-client dyad characteristics (e.g., concordance or discordance of gender and ethnicity). 39 This finding is supported by a nonprimary care feasible counseling intervention trial by Jemmott et al., which found that the effects of an HIV risk-reduction intervention in adolescents did not vary as a function of the facilitator's race or gender, participant's gender, or the gender composition of the intervention group. 40

Effectiveness of female condoms

In this report, we examined the effectiveness of female condoms in reducing the incidence and transmission of STIs. We did not reexamine the effectiveness of the consistent and correct use of male condoms (latex or polyurethane), which has previously been shown to be highly effective in preventing the sexual transmission of most STIs and HIV, 41, 42 the development of PID, 43 and possibly in reducing the risk of HSV and HPV. 44, 45 The FDA approved female condoms based on laboratory studies that demonstrated that polyurethane female condoms offer protection against STI pathogens. 46 Since their approval, we identified only three fair-to-good-quality studies directly examining the effectiveness of female condoms in preventing STIs. 29 31 These studies, only one of which was conducted in the US with poor followup, suggest that hierarchical counseling to use female condoms (i.e., use female condoms when male condoms are not used) and female condom availability offer a similar risk reduction in bacterial STI incidence as counseling to use male condoms only. Overall, however, the use of female condoms was low, 7 to 11 percent reporting consistent female condom use, 29 or 12 percent of all acts using a female condom. 30

Contextual Issues

The delivery, receipt, and impact of behavioral counseling interventions are influenced by various socio-cultural and health care system factors. Unfortunately, because of the heterogeneity of populations targeted, intervention settings, and intervention designs, it is difficult to draw conclusions about potential mediating and moderating factors by comparisons between studies. Some inferences, however, can be made from the body of evidence as a whole, and some conclusions can be drawn from larger individual trials that might inform the feasibility and generalizability of behavioral counseling interventions to prevent STIs in a primary care setting.

Socio-cultural factors. All seven trials that showed a beneficial impact on STI incidence or sexual behavior were conducted in low-income urban environments with an ethnically diverse patient population. Only the trial by Carey et al. among psychiatric outpatients was about 20 percent African American. The other six trials were predominantly or completely conducted with African Americans and Hispanics. Project SAFE, which had a subgroup analysis by ethnicity, reported similar point estimates of reduction in reinfection rate for bacterial STIs in Hispanics and African Americans. This risk reduction, however, was not statistically significant for African Americans, likely because there were about half the number of African Americans in the trial. Korte et al., therefore, concluded that the impact of the behavioral counseling intervention was comparable in both ethnic groups.(Appendix B Table 2)

Two trials in predominantly African American female adults and adolescents 19, 22 had subsequent subgroup analyses among participants who self-reported a history of intimate partner violence. 47, 48 Both subgroup analyses found that high-intensity behavioral counseling interventions had a similar beneficial effect (i.e., decrease in bacterial STI incidence in female adolescents and decrease in self-reported risky sexual behavior in female adults) as in the original trial analyses.(Appendix B Table 1)

Health care system factors. All studies meeting our inclusion criteria were conducted primarily in a clinic setting. For adolescents, two trials recruited patients and conducted the intervention in an HMO; 18, 25 one trial recruited from and was conducted in a hospital-based adolescent clinic; 20 one trial recruited from various community health agencies and was conducted in a family medicine clinic on weekends; 19 and one multi-center trial recruited from and was conducted in STI clinics. 14 For adults, one trial recruited from a managed care clinic and the intervention was delivered through the mail; 15 one trial was recruited from and conducted in a family practices in Australia; 23 one trial was recruited from and conducted in a family planning clinic; 22 one trial was recruited from and conducted in a university based health clinic; 24 one trial recruited from and was conducted in a psychiatric clinic; 13 and four trials were recruited from and conducted in STI clinics. 12, 14, 16, 17

All trials had dedicated research staff for the recruitment (screening), intervention, and assessments. For adolescents, four of the five trials had trained facilitators delivering the intervention, 14, 19, 20, 25 one of which also used peer co-faciliators. 19 For adults, six of the nine trials had trained facilitators delivering the intervention. 12 14, 16, 17, 22 One of these nine did not report on the intervention staffing. 24

For both adolescents and adults, only two trials used the existing primary care provider to deliver the intervention 18, 23 and one trial used a mailed intervention. 15 These three trials evaluated brief interventions conducted in lower-risk settings, and none of these trials showed a consistent change in self-reported STI incidence or sexual behavior.

In most of the trials, participants were given a financial incentive as reimbursement for time and travel, which ranged from $10 to $40 per session. In most of these trials, participants were given additional financial incentives for baseline and followup assessments.

Cost-effectiveness. In our targeted search for evidence on cost-effectiveness, we identified multiple cost-effectiveness and cost-benefit analyses primarily addressing HIV prevention. Most of these analyses, however, were based on evidence not applicable to primary care feasible behavioral counseling (e.g., community- or school-based interventions). In total, we only found two relevant articles that considered primary care feasible interventions, 49, 50 one of which we excluded because it was an analyses based on a trial by O'Donnell et al. evaluating a video-based intervention in an STI clinic, which we excluded for poor quality. 49, 51

We found one comparative cost-effective analysis that included a broad range of HIV-prevention interventions, ranging from individual-level counseling or partner notification to structural interventions like condom availability. 52 For each intervention, Cohen et al. selected one particular study that demonstrated effectiveness in changing HIV incidence, STI incidence, or self-reported risk behavior (e.g., unprotected sexual intercourse). Among the 26 types of interventions they addressed, they calculate the cost-effectiveness of the interventions from Project RESPECT and Project SAFE based on their model as prototypes for client-centered HIV Counseling and Testing and multiple sessions of group counseling, respectively. In their model, cost-effectiveness is assessed from the perspective of the public health system. HIV prevalence estimates were taken directly from the study or, when not available, from published estimates based on the location in which the study was conducted. Because of the uncertainty of some cost-estimates (e.g., cost of person hours, supplies, and overhead), Cohen et al. conducted a sensitivity analysis of the various cost parameters in their model.

For Project RESPECT, based on a 20 percent STI reduction at 12 months, 1447 persons were reached at an estimated cost of $74 per person, and 1.01 HIV cases averted in 12 months, leaving a cost-effectiveness ratio of $110,000 per case averted. For Project SAFE, based on an 38 percent STI reduction at 12 months, 313 persons were reached at an estimated cost of $160 per person, and 0.29 HIV cases averted at 12 months, leaving a cost-effectiveness ratio of $170,000 per case averted. Cohen et al. estimate that interventions for which the cost-effectiveness ratio remains below $200,000 may be considered “cost-effective” compared with the lifetime costs of treatment of HIV infection. In their sensitivity analyses, prevalence of HIV infection in the target population and cost per person reached by the intervention had the greatest impact on cost-effectiveness. Therefore, Cohen et al. conclude that the interventions delivered in Project RESPECT and Project SAFE may not be cost-effective for low-incidence populations. Their analyses of cost-effectiveness is based only on incident HIV cases averted and not other averted STIs. We did not identify any cost-effectiveness or cost-benefit analyses addressing STI prevention in general.


This body of evidence has several limitations that should be considered when interpreting this report's findings.

Interpreting results within studies. Trials reporting STI incidence with non-significant intervention effects do not imply the interventions are not effective. 12, 15, 18 The overall incidence of even common bacterial STIs is low. These studies, therefore, are subject to type-2 measurement error (e.g., inadequate power). In general, these trials do not report their calculation of sample size. Trials either had relatively small sample sizes, low percentages of STIs, or both. In addition, the study by Scholes et al. in adults and Boekeloo et al. in adolescents both used self-reported measures of STI incidence, which are particularly subject to assessment and reporting bias. 53, 54

Studies for KQ2 using self-reported behavioral outcomes should be interpreted cautiously, especially if there is no consistency in direction or magnitude of effect among different behavioral outcomes. Similar to self-reported STI incidence, self-reported behavioral outcomes are also particularly subject to assessment and reporting bias. 54 Measurement of these outcomes are not standardized, and improved validity and reliability of these measures require rigorous study design (e.g., extensive training of interviewer and continuous quality control monitoring) or alternative strategies, (e.g., audio computer assisted self-interview (A-CASI)) to decrease social desirability bias. In addition, there exist more permutations for selective reporting with behavioral outcomes (e.g., on different types of behavioral outcomes).

Although in the studies that reported both biological and behavioral outcomes, we generally found concordance of direction of effect of STI incidence and self-reported behavior for both positive 13, 14, 16, 17, 19, 20 and negative studies, 12, 15, 18 there were a few self-reported behavioral measures that were not concordant with the STI incidence. In a subsequent analysis of the Project RESPECT data, many behaviors had paradoxical associations with STI incidence. 55 The relationship between STI incidence and behavior is exceedingly complex, therefore, it is difficult to extrapolate disease prevention effectiveness from self-reported measures of behavioral outcomes without rigorous and comprehensive behavioral measurement. 56

Interpreting results across studies. Unfortunately there is minimal conclusive information that can be gathered from comparisons between studies. In general, there are a relatively small number of trials, 13 in total, 8 in adults, 4 in adolescents, and 1 in both. This small number is further diluted by the heterogeneity of the populations studied; intervention setting, format and intensity; and comparators (e.g., wait-list or usual care control; matched health promotion intervention; usual care involving minimal STI prevention education or counseling). Therefore, we are unable to pool effect sizes or make direct comparisons of effect sizes by population or intervention characteristics. In addition, we did not include comparative effectiveness studies of counseling interventions, which were beyond the scope of this review. Therefore, we are unable to comment on the differential effect of different types of interventions or intervention elements, other than findings from the studies we included with multiple intervention arms. 14, 17, 20, 22, 24

External validity. Our stringency around internal validity and scope of interventions, which focused primarily on whether primary care interventions are effective, affects the generalizability of our results. To minimize bias, we considered only randomized and non-randomized controlled trials and excluded all poor-quality studies and studies with limited reporting of trial details. However, we contacted first authors of articles to supplement or clarify issues around incomplete or unclear reporting of information, when appropriate. Second, we focused only on those interventions that we considered primary care feasible, and therefore excluded a large number of primarily community-based trials evaluating counseling interventions to prevent STIs, including HIV. Third, comparative effectiveness studies were beyond the scope of this review and, therefore, many trials in high risk populations (e.g., MSM, IVDU) in which HIV Counseling and Testing was considered “standard of care,” were excluded from this review. Therefore, our review does not address trials on many high risk populations (e.g., pregnant women, MSM, sex workers, military recruits, inmates or former inmates, mentally disabled, or IVDUs).

For some of these populations, sexual risk reduction is addressed elsewhere. In MSM and IVDU, for example, there is good evidence that community-based and community-level interventions can reduce risky behaviors. 57 59 In general, we found limited rigorous trial evidence for other “at risk” groups (e.g., pregnant women, mentally ill and mentally disabled, military recruits, sex workers, and inmates or former inmates). One study by Shain et al. included women who were pregnant, although results were not reported separately for this subgroup. One study by Carey et al. included mentally ill persons with co-morbid substance abuse. We identified trials for military recruits and sex workers in settings not considered to be primary care feasible. We identified one trial underway in newly released inmates (Appendix D).

In addition, as a result of our stringency around internal validity and scope, only two HIV Counseling and Testing intervention trials, 14, 24 and no risk-avoidance (abstinence-only) intervention trials, met criteria to be included in our review. However, the evidence for these interventions' effectiveness are addressed elsewhere. The effectiveness of HIV Counseling and Testing is covered in the USPSTF's report on HIV Screening. 60 The effectiveness of community-based risk avoidance behavioral counseling is currently under review by the CDC's Guide to Community Preventive Services in their review of STI prevention in adolescents.

Even more important than the limitations of applicability to different populations or interventions types, however, are the translational issues of delivering behavioral counseling interventions in practice (e.g., cost and opportunity cost of implementation and maintenance of counseling and other operational barriers). These issues are particularly pertinent for this body of evidence, since all identified effective counseling interventions were moderate- to high-intensity, and at minimum involved multiple sessions and trained counselors.

Evidence gaps and future research

In general, there still exists a paucity of methodologically rigorous trial evidence for the effectiveness of primary care feasible behavioral counseling to prevent STIs - in particular, for lower-risk populations and lower-intensity counseling interventions. Trials with biological outcomes (e.g., STI incidence) addressing readily identifiable “at-risk” populations (e.g., pregnant women, adolescents and young adults, ethnic minorities, mentally ill persons, and urban poor) in nonSTI clinics are needed to demonstrate the effectiveness of counseling in lower risk settings. Currently, there are several trials in press or in progress that may address some of these gaps (see Appendix D). If possible, trials addressing general-risk populations in primary care settings should also be conducted. We do not know of any trials underway that address a general-risk population in primary care.

There is also a need for effectiveness trials of lower-intensity counseling interventions that can more feasibly be incorporated into health care delivery. We identified several ongoing or recently completed trials that may address this evidence gap using brief single-session interventions, waiting-room interventions, or computer-delivered interventions (Appendix D).

Finally, there are few trials that replicate the effectiveness of proven counseling interventions in other populations. For example, there is a paucity of rigorous trial evidence for men, male adolescents, pregnant women, and other high-risk population groups. Similarly, there are few trials of different interventions in similar populations. Reproducibility of counseling interventions' effectiveness among different populations is important to establish both feasibility and generalizability. Evaluating different interventions in similar populations, as well as comparative effectiveness trials of counseling interventions, will help define essential elements for an effective counseling intervention for STI prevention.


From rigorous trials evaluating primary care feasible behavioral counseling interventions, we conclude that population risk and intervention intensity appear to correlate most strongly with intervention effect. There is fair evidence to suggest that moderate-to high-intensity behavioral counseling is effective in reducing STI incidence in high-risk populations (e.g., “at-risk” populations in STI clinics). There is no evidence for or against the extrapolation of the effectiveness of these interventions to similarly “at-risk” populations in non-STI clinics. In addition, there is fair evidence to suggest that among sexually active adolescents, high-intensity behavioral counseling is effective in reducing STI incidence in a primary care setting. All effective counseling interventions were based on individual-risk based counseling in individual or small group sessions. Within this body of evidence, we found fair evidence to suggest that risk reduction counseling interventions do not have significant behavioral or biological adverse effects.

We found surprisingly few studies evaluating the effectiveness of female condom use in preventing transmission of STIs. While female condom use in these trials was extremely low, female condom use combined with male condom use was not significantly different from male condom use alone in preventing incident STIs.

There is a lack of evidence for the effectiveness of low-intensity behavioral counseling interventions, especially in lower-risk populations. The few trials that evaluated low-intensity interventions had study design factors which may have contributed to their non-significant intervention effect findings. 15, 18, 23 We identified no trials conducted in a general-risk, primary care population.

Cover of Behavioral Counseling to Prevent Sexually Transmitted Infections
Behavioral Counseling to Prevent Sexually Transmitted Infections [Internet].
Evidence Syntheses, No. 64.
Lin J, Whitlock E, O'Connor E, et al.

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...