糖心vlog

Importance听 Increasing rates of preventable sexually transmitted infections (STIs) in the US pose substantial burdens to health and well-being.

Objective听 To update evidence for the US Preventive Services Task Force (USPSTF) on effectiveness of behavioral counseling interventions for preventing STIs.

Data Sources听 Studies from the previous USPSTF review (2014); literature published January 2013 through May 31, 2019, in MEDLINE, PubMed (for publisher-supplied records only), PsycINFO, and Cochrane Central Register of Controlled Trials. Ongoing surveillance through May 22, 2020.

Study Selection听 Good- and fair-quality randomized and nonrandomized controlled intervention studies of behavioral counseling interventions for adolescents and adults conducted in primary care settings were included. Studies with active comparators only or limited to individuals requiring specialist care for STI risk-related comorbidities were excluded.

Data Extraction and Synthesis听 Dual risk of bias assessment, with inconsistent ratings adjudicated by a third team member. Study data were abstracted into prespecified forms. Pooled odds ratios (ORs) were estimated using the DerSimonian and Laird method or the restricted maximum likelihood method with Knapp-Hartung adjustment.

Main Outcomes and Measures听 Differences in STI diagnoses, self-reported condom use, and self-reported unprotected sex at 3 months or more after baseline.

Results听 The review included 37 randomized trials and 2 nonrandomized controlled intervention studies (N鈥�=鈥�65鈥�888; 13 good-quality, 26 fair-quality) recruited from primary care settings in the US. Study populations were composed predominantly of heterosexual adolescents and young adults (12 to 25 years), females, and racial and ethnic minorities at increased risk for STIs. Nineteen trials (n鈥�=鈥�52鈥�072) reported STI diagnoses as outcomes (3 to 17 months鈥� follow-up); intervention was associated with reduced STI incidence (OR, 0.66 [95% CI, 0.54-0.81; I²鈥�=鈥�74%]). Absolute differences in STI acquisition between groups varied widely depending on baseline population STI risk and intervention effectiveness, ranging from 19% fewer to 4% more people acquiring STI. Thirty-four trials (n鈥�=鈥�21鈥�417) reported behavioral change outcomes. Interventions were associated with self-reported behavioral change (eg, increased condom use) that reduce STI risk (OR, 1.31 [95% CI, 1.10-1.56; I²鈥�=鈥�40%, n鈥�=鈥�5253). There was limited evidence on persistence of intervention effects beyond 1 year. No harms were identified in 7 studies (n鈥�=鈥�3458) reporting adverse outcomes.

Conclusions and Relevance听 Behavioral counseling interventions for individuals seeking primary health care were associated with reduced incidence of STIs. Group or individual counseling sessions lasting more than 2 hours were associated with larger reductions in STI incidence, and interventions of shorter duration also were associated with STI prevention, although evidence was limited on whether the STI reductions associated with these interventions persisted beyond 1 year.

In 2008 the Centers for Disease Control and Prevention estimated that more than 20 million sexually transmitted infections (STIs) occur each year in the US.¹ Surveillance data since then, and especially from 2014 to 2018, indicate substantial increases in the incidence of most STIs,² including syphilis and gonorrhea, that were previously rare or declining.³ STI incidence is highest among adolescents and young adults.¹ For example, 61% of the 1.1 million cases of chlamydia reported in 2018 in the US were among individuals aged 15 to 24 years.³ The high rates of STI in the US contribute to chronic disease, cancer, infertility, adverse birth outcomes, and mortality, posing a substantial burden to population health.³ Reported disparities in STI rates by region of the country, age, race, ethnicity, and sexual orientation are attributed to structural and social conditions, sexual behavior, and unequal access to health care, education, and public health programs.^4,5

Primary care clinicians have an important role in primary and secondary prevention of STIs. In addition to providing recommended STI screening and treatments, primary care clinicians can identify patients at increased risk for STIs by obtaining comprehensive sexual histories.^2,6 In 2014 the US Preventive Services Task Force (USPSTF) recommended intensive behavioral counseling for all sexually active adolescents and for adults at an increased risk for sexually transmitted infections (B recommendation).^7,8 The current review of the evidence on behavioral counseling interventions for STI prevention was conducted to inform a USPSTF update to its current recommendation.

An analytic framework was developed with 3 key questions (KQs) (Figure 1) that examined the effectiveness of behavioral counseling interventions in reducing STIs, related morbidity and mortality, or other health outcomes (KQ1); decreasing risky sexual behaviors or increasing protective behaviors (KQ2); and potential harms of behavioral counseling interventions (KQ3). Additional methodological details are publicly available in the full evidence report.¹⁰

To identify studies published since the previous review,⁸ literature searches were conducted from January 2013 through May 31, 2019, in MEDLINE, PubMed (for publisher-supplied records only), PsycINFO, and the Cochrane Central Register of Controlled Trials (eMethods in the Supplement). Additional studies were located by reviewing reference lists of other systematic reviews and through suggestions from experts. Ongoing surveillance was conducted after May 2019 through May 22, 2020, to identify newly published studies that might affect the findings of the review. This was accomplished through article alerts and targeted searches of journals with a high impact factor and journals relevant to the topic. The last surveillance on May 22, 2020, identified no new studies.

Two reviewers independently evaluated articles from the previous review in addition to citations and full-text articles from the literature searches against prespecified inclusion criteria (Figure 2 and eTable 1 in the Supplement). Given changes to the medical and social context of STI risk and prevention, studies from the previous review published more than 20 years ago (years 1999 or earlier) were not included. For all KQs, included studies were required to target sexual behavior change to prevent STIs. Studies among adults and adolescents, including those who were pregnant, were included. Studies that included people living with HIV were not excluded unless the study population and intervention was focused solely on HIV-positive populations. Studies that focused solely on behavior change to prevent unintended pregnancy or change behaviors such as drug and alcohol use associated with risky sexual behavior were not included. Additionally, studies limited to populations requiring complex, specialized interventions outside the scope of primary care to address their specific STI risk (eg, HIV-serodiscordant couples, commercial sex workers) were not included in the review. Randomized clinical trials (RCTs) and nonrandomized controlled intervention studies involving behavioral counseling to prevent or reduce STIs were included. Trials that assessed effectiveness of circumcision for HIV prevention, STI testing only, biomedical prevention interventions, or those conducted within closed or preexisting social networks were not included.

Studies were required to have outcomes assessed at 3 months鈥� postbaseline or longer and be conducted in or recruited from general primary care settings and other preventive care outpatient settings, including family planning clinics, STI clinics, school-based health clinics, and behavioral/mental health clinics. Studies were limited to those conducted in countries categorized as 鈥渧ery high鈥� on the Human Development Index (2016) published by the United Nations Development Programme.¹¹ Included studies were limited to those published in English and deemed good or fair quality based on USPSTF quality rating standards.⁹

Two reviewers applied USPSTF design-specific criteria⁹ to assess the methodological quality of all eligible studies, and each study was assigned a quality rating of 鈥済ood,鈥� 鈥渇air,鈥� or 鈥減oor鈥� (eTable 2 in the Supplement). Discordant quality ratings were resolved by discussion or by a third reviewer and adjudicated as needed. Studies rated as poor quality were excluded from the review. Good-quality RCTs were those that met all or nearly all the prespecified quality criteria. Fair-quality studies did not meet all criteria but did not have serious threats to their internal validity related to design, execution, or reporting. Intervention studies rated as poor quality generally had several important limitations, including at least 1 of the following risks of bias: very high attrition (defined as >40%); differential attrition between intervention groups (defined as >20%); lack of baseline comparability between groups without adjustment; or problematic issues in trial conduct, analysis, or reporting of results. One reviewer extracted data from all included studies rated as fair- or good quality directly into evidence tables, and a second reviewer checked the data for accuracy.

Data were synthesized separately for each KQ, and tables were created to describe study results for all included outcomes, with stratification by intervention and study population characteristics. Narrative summary and summary tables were used to describe important features of the included evidence, including study design and setting, interval validity, and important characteristics about patients and interventions. Based on a priori plans to evaluate differences in intervention effects by age, 2 sets of data tables were prepared鈥�1 set for studies focused on adolescent and young adult populations and the second for those on adult populations. Age categories represented broad age spans of included populations. To examine age effects, within-study age-specific subgroup analyses were extracted when available, regardless of the subgroup comparison methods.

Meta-analyses of STI incidence, condom use, and unprotected intercourse were conducted. The DerSimonian and Laird model was used for pooling odd ratios (ORs) of STI incidence. Adjusted study-reported ORs were used if reported, and ORs were calculated based on the reported proportion of participants with an STI in each group if adjusted differences were not reported or results were presented in a format that could not be combined for meta-analysis of ORs (eg, hazard ratios). A funnel plot was created and an Egger test was conducted to explore small-study effects, which can be related to publication bias.¹² Additionally, meta-regression and subgroup analyses were conducted to explore factors associated with effect size. Because fewer than 10 trials could be included in these analyses and statistical heterogeneity was fairly high (I²鈥�=鈥�50% and higher), restricted maximum likelihood models were run with the Knapp-Hartung correction for small samples, because the DerSimonian and Laird method tends to underestimate the width of the 95% CI when the number of studies is small or statistical heterogeneity is high.¹³ Several factors were examined as potential effect modifiers, including but not limited to recruitment setting (eg, STI clinic vs other settings), participant characteristics (eg, whether the study was limited to adolescents or young adults, whether the study was limited to male or female participants only), amount of intervention contact time (eg, length and number of sessions), modality (eg, in-person individual or group counseling), administrator (eg, licensed health professional), and study design features (eg, type of control). Analyses to assess the sensitivity of pooled effects to individual studies or types of studies鈥攆or example, those unique in terms of reported STI outcomes or study design鈥攚ere conducted when potential outliers were identified.

For KQ1, STI outcomes based on laboratory-confirmed clinical tests, conducted as part of the study, recorded in the medical records of study participants, and/or found in public health surveillance registries were analyzed. A few studies reported incidence of a specific STI, while others considered intervention effects on the diagnosis of any STI. For meta-analysis, these outcomes were combined. The follow-up periods ranged from 6 to 24 months; when several time points were reported, those closest to 12 months were included in the meta-analysis. For KQ2, condom use outcomes were reported using a variety of measures. A pooled analysis of dichotomous condom use measures was conducted, with selection based on any of the following in order of preference: consistent condom use (variably defined), condom use at last intercourse, any condom use, and condom used at first intercourse with new partners. Female condom use was included for this outcome and was reported in 1 trial.¹⁴

Stata version 15.1 (StataCorp LP) was used for all analyses. All significance testing was 2-sided, and results were considered statistically significant at P鈥�&濒迟;鈥�.05.

The strength of evidence was rated for each KQ based on consistency (similarity of effect direction and size), precision (degree of certainty around an estimate), reporting bias (potential for bias related to publication, selective outcome reporting, or selective analysis reporting), and study quality (ie, study limitations).

Two reviewers evaluated 4649 citations and 273 full-text articles against inclusion criteria, and 39 studies (62 articles)^14-75 met inclusion criteria for this systematic review (Figure 2). Twenty-two studies (20 RCTs^{14,17-20,23-25,28,30-32,36,38,45-47,49,50,53}; 2 nonrandomized controlled intervention studies^42,51) were carried forward from the previous USPSTF report, and 17 new RCTs^{15,16,55,58-61,66-75} were identified for inclusion (Table 1, Table 2, Table 3, Table 4). The newly available studies included 4 conducted among adolescents,^58,67,72,74 5 conducted among a mix of adolescents and young adults,^{59,61,68,71,75} and 1 focused solely on young adults.⁵⁵ The remaining 7 enrolled mostly adults and spanned a wider age range.^{15,16,60,66,69,70,73} Among the studies of adults and mixed-age populations, the mean age of participants was younger than 30 years in most studies, and none reported a mean age of 40 years or older. Thus, among studies that included adults, young adults (ie, aged 18-25 years) were most often recruited. Of the 39 included studies, 35 (90%) tested intervention effectiveness among individuals at an increased risk for an STI, including some participants with a recent STI diagnosis. Characteristics used to identify trial participants at increased risk for acquiring an STI included demographic characteristics associated with higher prevalence of STI in the US (eg, aged 15 to 24 years, African American race, Hispanic ethnicity), self-reported sexual and behavioral factors associated with STI risk (eg, men who have sex with men [MSM], multiple partners, unprotected intercourse), and personal health history (eg, recent STI or visit to STI clinic).

Most interventions were conducted in the following health care settings: general primary care, obstetrics and gynecology, STI clinics, women鈥檚 health clinics, adolescent medicine, and family planning clinics (eTables 3 and 4 in the Supplement). Nine studies^{20,31,47,49,51,55,60,71,75} included a low鈥揷ontact time intervention group (<30 minutes), 13^{15-17,24,28,36,38,42,45,46,61,68,69} included a moderate鈥揷ontact time intervention group (30-120 minutes), and more than half (21/39 [54%])^{14,17-20,23,25,30-32,50,53,58,59,66-68,70,72-74} tested a high鈥揷ontact time intervention group (>2 hours) (eTable 5 in the Supplement). The most common therapeutic approach in the included interventions was motivational interviewing,^76,77 which was used in 11 studies^{16,19,20,23,36,46,69,72,74,75}; another 7 studies used cognitive behavioral therapy approaches.^{17,30,50,53,68,72,73,78} Group counseling was the most common intervention component and was used in 25 intervention groups. Group counseling was frequently paired with other treatment components, such as individual counseling, videos, video games, or telephone contact. In most studies of group counseling, participants were of the same gender. Fourteen of the intervention groups primarily involved mobile-phone text or computer interventions, and all but 2 of these involved low or moderate contact times ranging from 5 to 90 minutes.^{15,16,19,45,55,60,61,67,69,71,74,75}

Key Question 1. Do behavioral counseling interventions that aim to decrease risky sexual behaviors or increase protective behaviors, or both, reduce sexually transmitted infections (STIs) or related morbidity and mortality?

Key Question 1a. Does the effectiveness of behavioral counseling interventions differ for subpopulations (eg, defined by age, STI history, sexual orientation, gender, pregnancy status)?

Key Question 1b. Does the effectiveness of behavioral counseling interventions differ by intervention characteristics (eg, intensity or mode)?

No studies reported intervention effects on STI morbidity or mortality, but 21 included studies^{16,17,20,23-25,30-32,36,38,42,45,46,50,51,53,60,61,69,75} (n鈥�=鈥�59鈥�328) reported on the effectiveness of behavioral interventions for preventing STI, nearly all were conducted among populations at increased risk (20/21 [95%]), and 19 reported measures that could be combined for meta-analysis (eTables 6 and 7 in the Supplement). These interventions were significantly associated with a lower incidence of STI at follow-up, most often reported at 6 to 12 months, with an overall pooled OR of 0.66 (95% CI, 0.54-0.81; I²鈥�=鈥�74%; n鈥�=鈥�52鈥�072) (Figure 3). Only 4 studies^16,36,38,69 reported higher absolute STI rates in the control condition, but the rates were not significantly different from rates in the intervention groups. STI incidence rates were highly variable across studies; control group rates ranged from 0% to 50%, while intervention group rates ranged from 0% to 37%. Statistical heterogeneity was substantial, and there was also considerable clinical heterogeneity, particularly in terms of populations and intervention characteristics. Planned subgroup comparisons were conducted to examine sources of heterogeneity (described below). A sensitivity analysis excluding a study of bacterial vaginosis prevention among lesbian and bisexual women did not change the overall result (pooled OR, 0.65 [95% CI, 0.52-0.80]; I²鈥�=鈥�75.3%; 18 studies).

Rigorous tests of intervention effectiveness for different subpopulations within trials were not reported. One trial⁵¹ reported several prespecified exploratory subgroup comparisons but did not test for interactions or account for multiple comparisons, while other trials presented post hoc exploratory subgroup comparisons⁶⁷ or secondary analyses on a subset of trial participants.²⁵ Instead, most of the included evidence focused exclusively on a specific subpopulation defined by sex, age, sexual history, pregnancy status, and racial/ethnic identity. Some subpopulations were not well-represented in the body of evidence; for example, only 2 included studies^32,69 recruited participants for interventions occurring during pregnancy. Many of the interventions designed for specific subpopulations were found to be effective, such as those for African American and Latina adolescents, adult minority women, and mixed-gender populations making visits to STI clinics. Trials among men, adolescent boys, MSM, and average risk populations were not well-represented in the body of evidence. Based on meta-regression tests for subgroup effects for STI prevention, larger effect sizes were associated with studies conducted among adolescents (P鈥�=鈥�.002), high-contact interventions (more than 2 hours compared with 2 hours or less) (P鈥�=鈥�.02), or group counseling sessions (P鈥�=鈥�.02) (eTable 12 in the Supplement). Overall, subgroup comparisons tended to include few studies in at least 1 category, and confounding of study population and intervention characteristics limited conclusions about which specific factors were responsible for intervention effectiveness.

Key Question 2. Do behavioral counseling interventions decrease risky sexual behaviors or increase protective behaviors that can reduce the risk of STIs?

Key Question 2a.Does the effectiveness of behavioral counseling interventions differ for subpopulations (eg, defined by age, STI history, sexual orientation, gender, pregnancy status)?

Key Question 2b. Does the effectiveness of behavioral counseling interventions differ by intervention characteristics (eg, intensity or mode)?

In total, 34 studies (n鈥�=鈥�21鈥�471) contributed evidence on the effectiveness of behavioral counseling interventions in changing a variety of sexual risk and protective behaviors (eTables 8 and 9 in the Supplement). Like the studies reporting STI outcomes, most of included evidence (30/34 [88%]) was from studies of people at increased risk for STI. Follow-up time in the studies reporting behavioral outcomes ranged from 3 to 14 months (mode, 12 months), and for the few studies reporting extended follow-up beyond 1 year, effects tended to diminish. Statistical heterogeneity was modest, but clinical heterogeneity was high because of the diverse populations and intervention approaches. Behavioral outcomes were not consistently reported, and measures were variable.

Eighteen^{14,15,17,18,24,25,31,32,45-47,49,55,58,61,68,70,74} studies reported on the effectiveness of behavioral interventions for condom use. Of these, 13 studies^{14,17,24,25,31,45,47,49,55,58,61,70,74} (n鈥�=鈥�5253) reported dichotomous condom-use outcomes with measures that were pooled for meta-analysis (Table 5; eTables 8 and 9 in the Supplement). The pooled result across these studies suggested that the intervention was associated with a higher odds of condom use (OR, 1.31 [95% CI, 1.10-1.56]; I²鈥�=鈥�40%) (Table 5).The percentage reporting condom use varied depending on the measure reported (eg, condom use at last sex, consistent condom use). Absolute differences between groups ranged from 鈥�7% to 27%. Eight^{17,24,25,31,32,45,46,61} of the studies reporting condom use outcomes also reported STI incidence, and the direction and statistical significance of effects were internally consistent; studies with higher levels of reported behavioral changes found greater reductions in STI diagnoses.

Twenty-one studies (n鈥�=鈥�13鈥�665) reported a measure of unprotected intercourse (eTables 8 and 9 in the Supplement). Of these, 14 studies (n鈥�=鈥�9183)^{15,16,19,20,24,25,30-32,38,58,67,70,73} reported the number of times participants engaged in unprotected sexual intercourse or intercourse without a condom over various time frames, with different degrees of specificity regarding the type of partner or sexual intercourse act. When combined to estimate an overall association, there was a small but statistically significant difference between groups. Fewer unprotected intercourse occasions were reported among those assigned to the intervention conditions (pooled mean difference, 鈥�0.94 [95% CI, 鈥�1.40 to 鈥�0.48]; I²鈥�=鈥�16%) (Table 5).

Other behavioral outcomes, such as the number of sexual partners and sexual abstinence, were reported by fewer studies and tended to be consistent with effects seen for other reported outcomes (eTables 10 and 11 in the Supplement). Only 1 trial tested an intervention aimed at reducing future STI risk before the onset of sexual activity. The moderate鈥揷ontact-time trial (Families Talking Together) enrolled male and female Hispanic or African American adolescents and their mothers at the time of a pediatric care visit and supported parental communication and education about sexual health.²⁸ At 9 months of follow-up, a lower proportion of adolescents in the intervention group reported ever having had sexual intercourse compared with the control group (6.8% vs 22.2%, P鈥�<鈥�.05 as reported in primary study).

Evidence on potential subpopulation differences available from within-trial comparisons was limited for the same reasons described above for STI outcomes. Comparisons of effect differences at the study level were not conducted for self-reported behavioral outcomes due to inconsistency in the outcome measures and the relatively small numbers of studies available for subgroup meta-regression analysis.

Key Question 3. What potential harms are associated with behavioral counseling interventions to reduce STI infections?

Few studies reported potential adverse consequences related to the STI-prevention behavioral interventions that were evaluated, and no evidence of statistically significant intervention harms was identified. Seven trials (n鈥�=鈥�3458) reported potentially adverse consequences that might theoretically arise from the behavioral interventions evaluated.^{17,25,32,45,61,66,75} A study conducted among adolescents and young adults that involved a text-messaging intervention reported 3 driver-caused road traffic crashes related to texting: 2 in the intervention group and 1 in the control group. The study was too small to permit estimation of the difference with any precision, and it was unclear whether the texting associated with the accidents was study related. No other studies relying on text-messaging interventions reported adverse events related to driving and texting.

This evidence report reviewed studies on behavioral counseling interventions to prevent STIs; the evidence is summarized by KQ in Table 6. An individual鈥檚 risk of acquiring an STI is influenced by many factors, including their sexual practices; relationship dynamics (eg, power and communication); skills, ability, and willingness to engage in health protective behaviors; and the prevalence of untreated STI in their social networks and community. For those at increased risk, behavioral counseling interventions provided in or referred from primary care settings can be effective in changing sexual risk and protective behaviors, lowering the risk of STI diagnosis for up to 1 year.

The randomized trials included in this review recruited participants at increased risk based on their reported sexual risk behaviors, history of STIs, and sociodemographic risk factors, including age and race/ethnicity. Most trials were focused on the prevention of STIs from heterosexual intercourse, with a few exceptions.^36,67,73 Increasing the use of male condoms during sex and reducing unprotected sexual intercourse was a primary aim of most interventions. Group and individual counseling provided by trained facilitators or health professionals were commonly included components of interventions, tending to involve more than 2 hours of contact time over the course of several weeks or months. Counseling interventions involving more than 2 hours of contact time may be associated with larger reductions in diagnosed STI at follow-up, although confounding of intervention and population characteristics limited conclusions that could be drawn about differences in effectiveness. Other reviews that have focused on different settings, study designs, and populations in evaluating STI behavioral counseling interventions have reported beneficial effects in a comparable range.^79-81

Compared with the 2014 USPSTF review,⁸ this review included 17 new studies^{15,16,55,58-61,66-75} and broadly supports the previous conclusions.⁸² Of the 17 additional studies contributing evidence to this update, just 5 reported STI diagnosis outcomes.^{16,60,61,69,75} Three of these 5 studies were low鈥� or moderate鈥揷ontact time interventions using tailored computer-based⁷⁵ or text messages⁶¹ delivered to male and female adolescent and young adult populations or, in the case of a small trial in Great Britain (Men鈥檚 Safer Sex), a 10-minute website visit intervention for adult men.⁶⁰ The previous USPSTF review called for the development and testing of less time-intensive and media-based interventions that might be applicable to broader populations. It is encouraging that there were new studies evaluating low鈥� and moderate鈥揷ontact time interventions available for the current review. Effect sizes of the newly added low- and moderate-contact interventions were comparable in size to effects seen in high-contact trials and contributed to the overall estimate of the effectiveness of behavioral interventions for STI prevention. Overall, however, the evidence on STI prevention in this review was primarily from previously included trials,^{23-25,30,31,42,50,51,53} and most of these were high鈥揷ontact time interventions focused on narrowly defined populations at increased risk for STIs.

Many studies reported outcomes at 12 or more months of follow-up, including several that reported statistically significant differences at 6 months that were no longer observed at 12 months. Thus, for studies reporting shorter follow-up times, it is possible that observed effects would have diminished with longer observation. The sustainability of the intervention effects observed in this review is unclear. Establishing new sexual habits and skills may have lasting effects or may require reinforcement to have continued health benefits. A comparative effectiveness study that was not included in this review, but that tested an adaptation of an included intervention,²⁵ found that ongoing telephone reinforcement was effective for maintaining STI prevention benefits for African American adolescents receiving the HORIZONS intervention.^83,84

The review was designed to identify evidence for interventions that can be feasibly implemented in or referred from US-based primary care health care settings. A broader body of evidence on STI prevention programs in the US and globally provides evidence on prevention strategies beyond this context. For example, evidence on interventions evaluated among people identified through social groups or institutions was not reviewed (eg, schools, churches, worksites), so an effective intervention with Latino MSM that recruited individuals from social settings such as bars or dance clubs and involving recruitment of friends was not included.⁸⁵ Thus, the absence of studies for some important populations at increased risk for STI in this review does not necessarily mean there are no effective interventions available. Other resources, such as the Community Guide of the Community Preventive Services Task Force,⁸⁶ provide reviews on effective community-based STI prevention programs.

Clinician risk assessments are necessary when ascertaining the eligibility of patients at increased risk of STIs for behavioral counseling interventions, yet few of the included studies incorporated risk assessment by primary care clinicians into the study design. Tests of interventions among populations of patients identified using defined STI risk assessment procedures could strengthen the applicability of evidence to primary care populations. Given that there may be considerable variation in sexual history鈥搕aking practices in primary care, the implementation of behavioral counseling interventions for populations at increased risk will likely require greater attention to this aspect of clinician training and health care delivery. There also remains a need for research on lower-intensity behavioral interventions that could be applied in primary care settings serving patients across a range of levels of STI risk. Pragmatic studies in health systems could provide important evidence on intervening in the context of routine primary care, especially when patients screen positive for an STI or report sexual risk behaviors such as multiple concurrent partners and unprotected intercourse.

One included trial was implemented in a pediatric health care setting and found to be effective for delaying self-reported initiation of sexual intercourse. The Families Talking Together intervention enrolled mother-child dyads and was designed for African American and Hispanic youth before they became sexually active.²⁸ A trial replicating these results among a larger sample of Hispanic, African American, and mixed-race youth published in May 2020 was identified in ongoing surveillance of the literature and provides additional evidence that the intervention was effective for reducing self-reported behavioral risk factors for STI (ie, time to sexual debut, condom use).⁸⁷ The included literature primarily evaluated intervention effectiveness within narrowly defined population groups (such as African American women or adolescent girls), and it is unknown whether the reported effect estimates would also be observed for the interventions if adapted to other populations. Most studies primarily included heterosexual participants at increased STI risk. Several groups that can experience increased risk for STIs were underrepresented in the evidence, however, including older adults, gay boys and men, and transgender populations. There also were only 2 studies in this review conducted among pregnant women, despite the clinical importance of preventing vertical transmission of STIs and complications that can be associated with STIs during delivery. Recent reports of a steep increase in mortality from congenital syphilis heighten the urgency of identifying pregnant individuals at increased risk and providing effective preventive interventions, as well as STI screening and treatment during pregnancy.³ Notably, none of the effectiveness trials eligible for this review enrolled sexual partner dyads despite the recognized role of interpersonal communication and relationship factors in STI risk.^88,89

This review has several limitations. First, studies that assessed the comparative effectiveness of different types of behavioral counseling interventions were not included because the topic scope was informed by the USPSTF Procedure Manual.⁹ Accordingly, the review was designed to support a recommendation based on intervention efficacy, and several studies were excluded because they did not include a usual-care or no-intervention control condition. Some excluded studies focused on high-risk populations not well-represented in the included body of evidence, such as African American boys and men,^90,91 MSM,^92-96 bisexual Black men,^97,98 transgender women,⁹⁹ and homeless young adults.¹⁰⁰ In addition, relatively few of the included studies were replications or adaptations of effective interventions.⁴²

Second, this review focused on interventions evaluated in the past 20 years, more than one-third of which were published prior to 2010. Changes in sexual risk behaviors and risks during this period may have implications for the applicability of the body of evidence. The development of effective treatments and a prophylactic agent for HIV infection may have shifted population risk perceptions. In addition, the emergence of social networking and dating app platforms has altered the landscape that defines sexual networks and sexual practices.¹⁰¹ Since adolescents and young adults have the highest rates of STI, research is needed to develop interventions addressing the current sexual, social, and interpersonal contexts that emerging adolescents and young adults will navigate.

Third, a variety of measures of condom use were reported across the trials, limiting estimation of pooled effects for behavioral outcomes. STI prevention generally requires consistent condom use, but many of the studies evaluated reported condom use only at last sex or whether condoms were ever used. Furthermore, behavioral outcomes were self-reported and subject to recall bias, Hawthorne effects, and social desirability bias.^79,102,103 Nevertheless, in studies reporting the incidence of diagnosed STIs, results for behavioral measures were generally consistent in the degree and direction of effects.

Behavioral counseling interventions for individuals seeking primary health care were associated with reduced incidence of STIs. Group or individual counseling sessions lasting more than 2 hours were associated with larger reductions in STI incidence, and interventions of shorter duration also were associated with STI prevention, although evidence was limited on whether the STI reductions associated with these interventions persisted beyond 1 year.

Corresponding Author: Jillian T. Henderson, PhD, MPH, Kaiser Permanente Evidence-based Practice Center, Center for Health Research, Kaiser Permanente Northwest 3800 N Interstate Ave, Portland, OR 97227 (Jillian.T.Henderson@kpchr.org).

Accepted for Publication: June 19, 2020.

Author Contributions: Dr Henderson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Henderson, Bean, O鈥機onnor.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Henderson, Senger, Bean.

Critical revision of the manuscript for important intellectual content: Henderson, Henninger, Redmond, O鈥機onnor.

Statistical analysis: Redmond, O鈥機onnor.

Administrative, technical, or material support: Senger, Henninger, Bean, Redmond.

Supervision: Henderson.

Conflict of Interest Disclosures: None reported.

Funding/Support: This research was funded under contract HHSA2902015000017I, Task Order No. 5, from the Agency for Healthcare Research and Quality (AHRQ), US Department of Health and Human Services.

Role of the Funder/Sponsor: Investigators worked with USPSTF members and AHRQ staff to develop the scope, analytic framework, and key questions for this review. AHRQ had no role in study selection, quality assessment, or synthesis. AHRQ staff provided project oversight, reviewed the report to ensure that the analysis met methodological standards, and distributed the draft for peer review. Otherwise, AHRQ had no role in the conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript findings.

Disclaimer: The opinions expressed in this document are those of the authors and do not reflect the official position of AHRQ or the US Department of Health and Human Services.

Additional Contributions: We gratefully acknowledge the following individuals for their contributions to this project: Kathleen Irwin, MD, MPH, and Tina Fan, MD, MPH (AHRQ); current and former members of the US Preventive Services Task Force who contributed to topic deliberations; Jennifer Lin, MD (Kaiser Permanente Center for Health Research), for mentoring and project oversight, and Smyth Lai, MLS, and Katherine Essick, BS (Kaiser Permanente Center for Health Research), for technical and editorial assistance. The USPSTF members, peer reviewers, and federal partner reviewers did not receive financial compensation for their contributions.

Additional Information: A draft version of this evidence report underwent external peer review from 4 content experts (Errol Lamont Fields, MD, PhD, Johns Hopkins University School of Medicine; Marie Harvey, DrPH, MPH, Oregon State University; Ralph DiClemente, PhD, New York University; Michelle Ybarra, PhD, MPH, Johns Hopkins University School of Medicine); and 3 federal partners: the Centers for Disease Control and Prevention, US Food and Drug Association, and National Institutes of Health. Comments were presented to the USPSTF during its deliberation of the evidence and were considered in preparing the final evidence review.

Editorial Disclaimer: This evidence report is presented as a document in support of the accompanying USPSTF Recommendation Statement. It did not undergo additional peer review after submission to JAMA.