Evidence reviews for the US Preventive Services Task Force (USPSTF) use an analytic framework to visually display the key questions that the review will address to allow the USPSTF to evaluate the effectiveness and safety of a preventive service. The questions are depicted by linkages that relate interventions and outcomes. A dashed line indicates health outcomes that follow an intermediate outcome. HSV indicates herpes simplex virus. Further details are available from the USPSTF procedure manual.10
aStudies that screen using an HSV-2 serologic test alone or a paired (HSV-1 and HSV-2) serologic test will be included if they meet other eligibility criteria; however, only the accuracy of test characteristics related to HSV-2 serologic tests will be evaluated.
bKey question 7 will be addressed only if there is insufficient literature for key questions 1 and 5 but sufficient literature for key question 4.
KQ indicates key question.
eMethods. Systematic Review Literature Search Strategies
eTable1. Inclusion/Exclusion Criteria
eTable 2. Quality Assessment Criteria
eTable 3. Quality Rating Criteria
eTable 4. Quality Ratings of Studies Assessing the Accuracy of Serologic Screening Tests for HSV-2 (Key Question 2)
eTable 5. Quality Ratings of Studies Assessing the Harms of Serologic Screening for HSV-2 (Key Question 3)
eTable 6. Quality Ratings of Studies of Antiviral Medications on HSV-2 Viral Shedding, Symptomatic Episodes and Transmission (Key Question 4 and Key Question 5)
eTable 7. Quality Ratings of Studies of Assessing Harms of Preventive Interventions (Key Question 6)
eTable 8. Results of Included Studies Assessing the Accuracy of HerpeSelect for HSV-2 (Key Question 2)
eTable 9. Characteristics and Results of Included Studies Assessing the Accuracy of biokit Rapid HSV-2 (Key Question 2)
eTable 10. Characteristics and Results of Included Studies Assessing Harms of Serologic Screening for Genital Herpes (Key Question 3)
eTable 11. Characteristics and Results of Included Studies Assessing Harms of Preventive Interventions (Key Question 6)
eFigure 1. Hierarchical Summary Receiver Operating Characteristic (HSROC) Curve Comparing Sensitivity and Specificity for the 10 Studies Reporting on the Accuracy of HerpeSelect® at the Manufacturer’s Cutpoint for a Positive Test (1.1).
eFigure 2. Hierarchical Summary Receiver Operating Characteristic (HSROC) Curve Comparing Sensitivity and Specificity for the 7 Studies Reporting on The Accuracy of HerpeSelect® at Cutpoints Between 2.1 and 3.5
eFigure 3. Hierarchical Summary Receiver Operating Characteristic (HSROC) Curve Comparing Sensitivity and Specificity for the 4 Studies Reporting on the Accuracy of BiokitHSV-2
Feltner C, Grodensky C, Ebel C, Middleton JC, Harris RP, Ashok M, Jonas DE. Serologic Screening for Genital HerpesAn Updated Evidence Report and Systematic Review for the US Preventive Services Task Force. JAMA. 2016;316(23):2531-2543. doi:10.1001/jama.2016.17138
Genital herpes simplex virus (HSV) infection is a prevalent sexually transmitted infection. Vertical transmission of HSV can lead to fetal morbidity and mortality.
To assess the evidence on serologic screening and preventive interventions for genital HSV infection in asymptomatic adults and adolescents to support the US Preventive Services Task Force for an updated recommendation statement.
MEDLINE, Cochrane Library, EMBASE, and trial registries through March 31, 2016. Surveillance for new evidence in targeted publications was conducted through October 31, 2016.
English-language randomized clinical trials (RCTs) comparing screening with no screening in persons without past or current symptoms of genital herpes; studies evaluating accuracy and harms of serologic screening tests for HSV-2; RCTs assessing preventive interventions in asymptomatic persons seropositive for HSV-2.
Data Extraction and Synthesis
Dual review of abstracts, full-text articles, and study quality; pooled sensitivities and specificities of screening tests using a hierarchical summary receiver operating characteristic curve analysis when at least 3 similar studies were available.
Main Outcomes and Measures
Accuracy of screening tests, benefits of screening, harms of screening, reduction in genital herpes outbreaks.
A total of 17 studies (n = 9736 participants; range, 24-3290) in 19 publications were included. No RCTs compared screening with no screening. Most studies of the accuracy of screening tests were from populations with high HSV-2 prevalence (greater than 40% based on Western blot). Pooled estimates of sensitivity and specificity of the most commonly used test at the manufacturer’s cutpoint were 99% (95% CI, 97%-100%) and 81% (95% CI, 68%-90%), respectively (10 studies; n = 6537). At higher cutpoints, pooled estimates were 95% (95% CI, 91%-97%) and 89% (95% CI, 82%-93%), respectively (7 studies; n = 5516). Use of this test at the manufacturer’s cutpoint in a population of 100 000 with a prevalence of HSV-2 of 16% (the seroprevalence in US adults with unknown symptom status) would result in 15 840 true-positive results and 15 960 false-positive results (positive predictive value, 50%). Serologic screening for genital herpes was associated with psychosocial harms, including distress and anxiety related to positive test results. Four RCTs compared preventive medications with placebo, 2 in nonpregnant asymptomatic adults who were HSV-2 seropositive and 2 in HSV-2–serodiscordant couples. Results in both populations were heterogeneous and inconsistent.
Conclusions and Relevance
Serologic screening for genital herpes is associated with a high rate of false-positive test results and potential psychosocial harms. Evidence from RCTs does not establish whether preventive antiviral medication for asymptomatic HSV-2 infection has benefit.
Quiz Ref IDGenital herpes is a sexually transmitted infection (STI) caused by herpes simplex virus (HSV) type 1 or type 2. Many patients experience signs and symptoms during primary infection and recurrences, but some may have mild or no symptoms.1 Episodic subclinical viral shedding leads to the potential for transmission in the absence of symptoms.2,3 HSV-2 accounts for most prevalent genital herpes cases and is more likely to cause frequent symptomatic recurrences than HSV-1 (which is most commonly acquired in childhood and associated with orofacial infection).4- 6 Genital HSV infection during pregnancy poses a risk of neonatal transmission during delivery, particularly among women who acquire HSV near the time of delivery.1,7
The true prevalence of asymptomatic HSV-2 infection is unknown; prevalence estimates rely on serologic test results and are not confirmed with Western blot. The estimated seroprevalence of HSV-2 in the United States was 16% in 2005-2010; only 14% of seropositive persons reported having been diagnosed with genital herpes.6 It is unclear what proportion of HSV-2–seropositive participants with no prior diagnosis of genital herpes had true asymptomatic (or unrecognized) infection vs a false-positive test result.
In theory, screening to identify unrecognized HSV-2 infection followed by counseling, antiviral treatment (episodic or suppressive), or both could prevent transmission and reduce symptoms and shedding. Several US Food and Drug Administration (FDA)–approved type-specific HSV serologic tests are available.8 Since HSV-2 rarely causes infection outside the anogenital region, HSV-2 antibodies can be interpreted as an indicator of genital infection.
In 2005, the US Preventive Services Task Force (USPSTF) recommended against routine serologic screening for HSV in asymptomatic adolescents, adults, and pregnant women (grade D recommendation).9 To inform an updated recommendation, we reviewed the evidence on benefits and harms of serologic screening for HSV-2, screening test accuracy, and benefits and harms of antiviral treatment in populations relevant to US primary care.
Detailed methods are available in the full evidence report at https://www.uspreventiveservicestaskforce.org/Page/Document/final-evidence-review/genital-herpes-screening1. The full evidence report also provides additional information about the proposed methods for key question (KQ) 7 (focused on the association between subclinical genital HSV-2 viral shedding and health outcomes), additional background and contextual information about genital herpes in the United States, and a list of studies that were excluded during the full-text review phase of the literature search.
Figure 1 shows the analytic framework and KQs that guided the review.
PubMed/MEDLINE, the Cochrane Library, and EMBASE were searched for English-language articles published through March 31, 2016. Search strategies are listed in eMethods in the Supplement. We searched for unpublished literature in ClinicalTrials.gov and the World Health Organization’s International Clinical Trials Registry Platform. To supplement electronic searches, reference lists of pertinent articles and suggested citations from reviewers were reviewed. We conducted ongoing surveillance after March 2016 through article alerts and targeted searches of high-impact journals to identify major studies published in the interim that may affect the conclusions or understanding of the evidence and related USPSTF recommendation. The last surveillance was conducted on October 31, 2016.
Two investigators independently reviewed titles, abstracts, and full-text articles to determine eligibility using prespecified criteria (eTable 1 in the Supplement). Disagreements were resolved by discussion. English-language studies of immunocompetent adults or adolescents, including pregnant women, were included. Only studies rated as good or fair quality were included. For all KQs, studies of persons without symptoms or a clinical history of genital herpes were eligible, as were studies of asymptomatic partners of persons with known genital herpes (ie, discordant couples). For the overarching question on direct evidence that screening improves health outcomes (KQ1), only randomized clinical trials (RCTs) comparing groups that were screened with groups that were not screened were included.
Quiz Ref IDFor KQ2 (accuracy of serologic tests), we included studies of FDA-approved serologic tests for HSV-2 that reported accuracy compared with the Western blot, which has been used as a reference standard in studies assessing commercially available serologic tests in the United States. Eligible populations could be symptomatic, asymptomatic, or a combination of both.
For KQ3 (harms of screening), we included trials, systematic reviews, and observational studies assessing the harms of screening in asymptomatic populations with no prior diagnosis of genital herpes, with or without a comparison group.
For studies assessing benefits or harms of preventive medications in asymptomatic populations (KQ4 through KQ6), RCTs comparing FDA-approved oral antiviral medications for the suppression of recurrent genital herpes (acyclovir, famciclovir, or valacyclovir) with placebo were eligible. RCTs of behavioral counseling interventions (eg, education or counseling; partner notification; barrier protection; or combinations of these components) were also eligible. For studies assessing the harms of antiviral medications in pregnant women (KQ6b), multi-institution antiviral medication pregnancy exposure registries were eligible. Eligible outcomes included reduced rates of symptomatic episodes and transmission (including measures of HSV-2 seroconversion). For KQ5b (effectiveness of interventions in pregnant women), eligible outcomes also included rates of neonatal HSV infection and reduced rates of symptomatic genital herpes at delivery. For KQ4 (effects of antiviral medication on subclinical HSV-2 shedding), we included any outcome measure of subclinical HSV-2 shedding (eg, percentage of days with any shedding detected).
For each included study, one investigator extracted information about design, population, tests or treatments used, and outcomes, and a second investigator reviewed for completeness and accuracy. Two independent investigators assessed the quality of each study as good, fair, or poor, using predefined criteria developed by the USPSTF and adapted for this topic (eTables 2-3 in the Supplement). Individual study quality ratings are provided in the Supplement (eTables 4-7).
Findings for each question were summarized in tabular and narrative form. To determine whether meta-analyses were appropriate, the clinical and methodological heterogeneity of the studies was assessed following established guidance.11 To do this, we qualitatively assessed the similarities and differences in populations, tests, treatments, comparators, outcomes, and designs. For KQ2 (the only KQ with sufficient numbers of similar studies for quantitative syntheses), pooled sensitivities and specificities for each type of serologic test were calculated using a hierarchical summary receiver operating characteristic (HSROC) curve analysis when at least 3 similar studies were available. Separate models were developed for each type of serologic test, and separate analyses were conducted for HerpeSelect using the manufacturer-recommended cutpoint for test positivity and for higher cutpoints reported in the literature to determine whether accuracy is improved with using higher cutpoints. The metandi program in Stata version 1412 was used to conduct all quantitative analyses.
Seventeen included studies with a total of 9736 participants (range, 24-3290) and reported in 19 publications were identified (Figure 2).13- 29
Key Question 1a. Does serologic screening for HSV-2 or combined testing for HSV-1 and HSV-2 in asymptomatic nonpregnant adults and adolescents reduce future symptomatic episodes and transmission of genital herpes?
Key Question 1b. Does serologic screening for HSV-2 or combined testing for HSV-1 and HSV-2 in pregnant women reduce neonatal HSV infection and symptomatic episodes of genital herpes at delivery?
No eligible studies were identified.
Key Question 2. What is the accuracy of serologic screening for HSV-2 in asymptomatic adults, adolescents, and pregnant women?
We included 11 good- or fair-quality studies (n = 7129; range, 61-3290) assessing the accuracy of 1 or more type-specific HSV-2 serologic tests compared with Western blot.13- 23 All 11 studies enrolled adults and none enrolled pregnant women. Most studies (8) enrolled a population with HSV-2 prevalence greater than 40% based on Western blot (range, 41%-70%). Two included studies described whether participants had current or prior symptoms consistent with genital herpes17,30; 1 study enrolled US college students with no current or previous symptoms consistent with genital herpes,30 and the other enrolled men seeking care at US STI clinics (17% were later diagnosed with genital herpes).17 In the 9 other studies, the proportion of participants who had current or past symptoms of genital herpes was not described. Most studies enrolled participants from 1 or more African countries; 3 were set in the United States,13,17,18 and 1 enrolled participants from multiple countries (Argentina, Costa Rica, Korea, Mexico, Nigeria, Thailand, and Vietnam).14
All 11 studies compared the Focus HerpeSelect HSV-2 enzyme-linked immunosorbent assay with Western blot and used a test cutpoint value of 1.1 to define a positive test result (current manufacturer’s cutpoint). Seven studies also assessed higher test cutpoints to boost specificity (ranging from 2.1 to 3.5).14- 16,19,21- 23 Four studies also assessed accuracy of the biokit HSV-2 Rapid Test.18,20- 22 Further study characteristics are shown in Table 1.
Ten studies (n = 6537 participants analyzed; range, 89-3290) provided sufficient data to estimate sensitivity and specificity of HerpeSelect using a cutpoint value of 1.113- 16,18- 23; individual study results are shown in eTable 8 in the Supplement. Pooled estimates of sensitivity and specificity were 99% (95% CI, 97%-100%) and 81% (95% CI, 68%-90%), respectively (Table 2). eFigure 1 in the Supplement shows the HSROC with 95% confidence ellipse using pairs of sensitivity and specificity. Estimates for specificity were highly variable, ranging from 41% to 97.5%. Studies handled equivocal (or indeterminate) test results in various ways, which may contribute to heterogeneity in estimates of test accuracy (eTable 8 in the Supplement). Five studies (n = 1840; range, 61-776) reported a positive predictive value using the manufacturer’s cutpoint; estimates ranged from 37.5 to 86.0.13,14,17,18,22 Four studies (n = 1779; range, 89-776) reported negative predictive values; estimates ranged from 96.5 to 100.13,14,18,22
Seven studies (n = 5516 participants analyzed; range, 99-3290) assessed higher cutpoints for a positive test result than those recommended by the manufacturer (ranging from 2.2 to 3.5) (eTable 8 in the Supplement).14- 16,19,21- 23 Estimates of specificity were generally higher at cutpoints greater than 2.2. The pooled estimates of sensitivity and specificity were 95% (95% CI, 91%-97%) and 89% (95% CI, 82%-93%), respectively (Table 2). eFigure 2 in the Supplement shows the HSROC with 95% confidence ellipse using pairs of sensitivity and specificity.
Four studies (n = 1512 participants analyzed; range, 98-782) reported on the sensitivity and specificity of the biokit HSV-2 Rapid Test (eTable 9 in the Supplement).18,20- 22 The pooled estimates of sensitivity and specificity were 84% (95% CI, 73%-91%) and 95% (95% CI, 93%-97%), respectively (Table 2). eFigure 3 in the Supplement shows the HSROC with 95% confidence ellipse using pairs of sensitivity and specificity.
Key Question 3a. What are the harms of serologic screening for HSV-2 or combined testing for HSV-1 and HSV-2 in asymptomatic nonpregnant adolescents and adults?
Key Question 3b. What are the harms of serologic screening for HSV-2 or combined testing for HSV-1 and HSV-2 in asymptomatic pregnant women?
Characteristics and outcomes of the 2 included fair-quality studies are shown in eTable 10 in the Supplement.24,25 Both studies were set in the United States and assessed the effect of a positive HSV-2 serologic test result on psychosocial outcomes among people who reported no history of genital herpes.
The first study (n = 24) was a qualitative assessment of the psychosocial effects of receiving an HSV-2 diagnosis based on serologic testing.24 Investigators recruited 24 adult participants reporting no history of genital herpes who tested HSV-2 seropositive by Western blot. Participants were recruited from clinical settings (STI, maternal and infant care, family medicine, and research clinics) and completed in-depth interviews on their experience of HSV-2 diagnosis. The qualitative analysis identified 3 categories of themes: (1) short-term, emotional responses that included surprise, denial, confusion, distress, sadness, disappointment, and relief to know; (2) short-term, psychological responses that included fear of telling sex partners, anger at the source partner, guilt about acquiring or transmitting, and concern about transmitting to a child; and (3) perceived ongoing responses that included fear of telling future partners, concern about transmitting to a sex partner, feeling sexually undesirable, feeling socially stigmatized, feeling like “damaged goods,” sex avoidance because of social responsibility, fear of transmitting to a newborn child, and relationship concerns relating to the diagnosis. The authors concluded that participants exhibited strong emotional and psychological responses to their serologic diagnoses of HSV-2, while observing that some of these responses were time limited.24
The second study (n = 33) enrolled individuals aged 14 to 30 years from an urban university setting and various clinical settings, including STI, primary care, and adolescent clinics.25 Of the 1190 enrolled, 820 (68%) had serologic testing (type of test not described) and 149 (18%) were HSV-2–positive. Of those participants who screened positive for HSV-2, 93 (62%) returned for their initial test results and 33 returned for the 3-month follow-up. At 3 months, participants completed the herpes Health-Related Quality of Life Questionnaire (HRQOLQ).31 Participants responded to each item using a 4-point scale that ranged from “very” to “not at all.” For individual-item analysis, answers of “very” or “quite” were considered indicative of endorsing the experience.25 A number of individual HRQOLQ items were endorsed frequently as “very” or “quite,” including the following: “It is difficult to forget that I have herpes” (63%); “I worry about giving herpes to someone” (56%); and “I worry about people finding out I have herpes” (48%).25 Other items from the HRQOLQ endorsed less frequently are shown in eTable 10 in the Supplement.
Key Question 4. How effective are oral antiviral medications in reducing genital HSV-2 viral shedding in asymptomatic adolescents, adults, and pregnant women?
We included 2 fair-quality RCTs (n = 129; range, 63-66) that compared daily preventive antiviral medication with placebo over 6 to 8 weeks (Table 3).26,27 One study did not describe how symptoms of genital herpes were ascertained27; the other reported that participants had undergone HSV serologic testing as part of their clinical care but reported no current or prior symptoms consistent with genital herpes.26 One RCT (n = 63) assessed valacyclovir (1 g daily)26 and the other (n = 66) assessed famciclovir (250 mg twice daily).27 In both studies, participants were educated on performing self-administered swabs of the anogenital area, which they completed once daily during the treatment period. Study characteristics are shown in Table 3.
One RCT reported a statistically significant reduction in viral shedding outcomes, and the other did not (Table 3). In the RCT assessing valacyclovir (n = 63), valacyclovir treatment significantly reduced subclinical days with HSV-2 shedding compared with placebo (1.5% vs 5.1%, P < .001) and also resulted in a greater proportion of participants experiencing no days with shedding (84% vs 54%, P = .001; between-group difference, 30% [95% CI, 15%-46%]).26 In the RCT comparing famciclovir (250 mg twice daily) with placebo (n = 66), participants in both groups did not have a significantly different risk of subclinical viral shedding; polymerase chain reaction samples were positive on 5.0% and 5.7% of subclinical days, respectively (relative risk [RR], 0.8 [95% CI, 0.41-1.56]; P = .52).27
Quality limitations across both trials included attrition (23%-29% of participants) and unclear handling of missing data. In addition, the validity and reliability of daily self-swab to ascertain viral shedding is unclear, potentially contributing to measurement bias.
Key Question 5a. How effective are preventive medications and behavioral counseling interventions in reducing future symptomatic episodes and transmission of genital herpes in asymptomatic nonpregnant adults and adolescents?
Key Question 5b. How effective are preventive medications and behavioral counseling interventions in reducing neonatal HSV infection and symptomatic episodes of genital herpes at delivery in pregnant women?
We included 4 fair-quality RCTs (n = 2550; range, 63-1484) evaluating antiviral medications (Table 3).26- 29 Two focused on preventing transmission and enrolled adult heterosexual couples who were serologically discordant for HSV-2 infection (ie, one partner had known genital herpes and the other partner had no prior diagnosis and was also HSV-2 seronegative),28,29 and 2 enrolled asymptomatic adults with no history of genital herpes who were seropositive for HSV-2 infection.26,27
Two RCTs comparing daily suppressive antiviral medication with placebo also reported viral shedding outcomes and were described in KQ4. Both reported on incidence of genital herpes symptoms (Table 3). In the RCT assessing valacyclovir (n = 63), participants were educated on signs and symptoms of genital herpes and instructed to return to the clinic any time they suspected an outbreak. At 2 months, fewer participants in the valacyclovir group reported symptoms of genital herpes than in the placebo group (12% vs 23%, respectively). The authors report that the treatment effect was significant (P = .033, controlling for the crossover effect); however, the arithmetic mean of the difference between groups was not significant (11% [95% CI, −0.6% to 22%]).26 In the RCT assessing famciclovir (n = 66), the incidence of genital herpes symptoms was similar in the famciclovir and placebo groups at 6 weeks (17.5% and 17.2%, respectively; P value not reported).27
Quiz Ref IDQuality limitations across both trials included attrition (23%-29% of participants), unclear handling of missing data, and risk of measurement bias. Symptoms were ascertained by self-report (not using a validated questionnaire) over a relatively short duration (6-8 weeks). One study enrolled participants who had HSV serologic testing as part of their clinical care26; results may not be applicable to those who screen positive but are not seeking testing for HSV infection.
Two RCTs compared the benefit of daily suppressive antiviral medication with placebo for preventing genital herpes transmission between HSV-2–serodiscordant heterosexual couples; one measured outcomes at 8 months28,29 and the other at 12 to 24 months (Table 3).32 One RCT enrolled immunocompetent couples,28,29 while the other was a substudy of HIV-1–serodiscordant couples in which the HIV-1–negative partner was also susceptible to HSV-2.32 One RCT (n = 1484 couples) assessed valacyclovir (500 mg daily),28,29 and the other (n = 937 couples) assessed acyclovir (400 mg daily); in both studies, antiviral medication was provided to the infected partner.32 Both studies were conducted in multiple countries; one enrolled from 96 study sites in the United States, Canada, Europe, Latin America, and Australia,28,29 and one was conducted in 7 sub-Saharan African countries.32
The 2 trials found conflicting results (Table 3). In the RCT assessing valacyclovir, fewer HSV-2–susceptible partners in the valacyclovir group had symptomatic HSV-2 infection than partners randomized to placebo over 8 months (0.5% vs 2.2%, respectively; hazard ratio, 0.25 [95% CI, 0.08-0.75]); similarly, fewer HSV-2–suceptible partners in the valacyclovir group seroconverted to HSV-2 than those in the placebo group (1.9% vs 3.6%, respectively; P = .04; hazard ratio, 0.52 [95% CI, 0.27-0.99]).28,29 In contrast, the RCT assessing acyclovir use among HIV-1–serodiscordant couples did not find a reduction in transmission. At follow-up (median, 18 months), the number of susceptible partners with seroconversions was not statistically different between the acyclovir group (40) and placebo group (28), which indicated seroincidence of 5.9 and 4.3 cases per 100 person-years, respectively (hazard ratio, 1.35 [95% CI, 0.83-2.20]; P = .22).32
Key limitations across both trials included high attrition; 22% of couples withdrew from one trial,28,29 and the overall attrition was 66% in the other trial.32
Key Question 6a. What are the harms of preventive medications and behavioral counseling interventions for reducing future symptomatic episodes and transmission of genital herpes in asymptomatic nonpregnant adults and adolescents?
Key Question 6b. What are the harms of preventive medications and behavioral counseling interventions for reducing neonatal HSV infection and symptomatic episodes of genital herpes at delivery in asymptomatic pregnant women?
One RCT (n = 63) included in KQ4 and KQ5 reported on harms.26 Rates of reported adverse events were similar among groups randomized to valacyclovir and placebo, including dizziness, headache, and nausea (eTable 11 in the Supplement).26
This review did not identify any eligible studies directly assessing the benefits or harms of serologic screening for HSV-2 compared with no screening. Therefore, the literature that might establish an indirect chain of evidence from multiple questions that link screening to health outcomes (KQ2 through KQ7) was reviewed.
Table 4 provides a summary of findings in this evidence review. Estimates for specificity varied and were imprecise, without a clear explanation for the observed heterogeneity. Potential explanations for false-positive serologic test results include cross-reactivity with HSV-1 (or other viruses), recent seroconversion, geographic variability in HSV-2 strain variants, and laboratory error. At higher cutpoints, estimates of sensitivity and specificity from 8 studies in Africa were still imprecise. Quiz Ref IDThere was evidence from 2 uncontrolled observational studies that detection of unexpected HSV-2 by screening is associated with potential psychosocial harms, including anxiety, worry, and distress. Other potential harms of serologic screening include false-positive results that lead to psychosocial distress and costs of confirmatory testing.
Quiz Ref IDThe estimates of the accuracy of serologic tests are generally applicable to populations with a higher prevalence of HSV-2 infection than general primary care populations in the United States. The majority of studies assessing the accuracy of HerpeSelect enrolled a population with HSV-2 prevalence greater than 40% based on Western blot. Use of HerpeSelect in a population with lower prevalence, similar to that of US adults, would greatly increase the number of false-positive test results. For example, in a hypothetical population of 100 000 persons with a prevalence of asymptomatic HSV-2 of 50% (similar to the prevalence in included studies), with test sensitivity of 99% and specificity of 81%, there would be an estimated 49 500 true-positive results and 9500 false-positive results (positive predictive value, 84%). If the prevalence instead were 16% (similar to the seroprevalence in the general US adult population among people with unknown symptom status), the number of true-positive results would be estimated at 15 840, and the number of false-positive results would be estimated at 15 960 (positive predictive value, 50%). True-positive results would decrease further, and false-positive results would increase further if the prevalence were less than 16%.
If sensitivity were unchanged, screening a lower-prevalence population would reduce the number of false-negative test results, although the negative predictive value would change little. It is possible, however, that the sensitivity of the screening tests could be lower in a lower-prevalence population, owing to such factors as lower antibody levels, thus increasing the number of false-negative results per 1000 persons tested. The direction of these changes with prevalence would be similar regardless of which cutpoints were used.
There was limited evidence evaluating preventive interventions for asymptomatic adults who screen positive for HSV-2. No studies enrolled pregnant women or adolescents, and none assessed behavioral counseling interventions. Two RCTs (total of 129 participants) assessed the benefit of preventive antiviral medications for reducing HSV-2 viral shedding and symptomatic occurrences among adults seropositive for HSV-2 who reported no prior genital herpes symptoms. Evidence from these 2 trials does not allow an accurate estimate of the benefit of preventive antiviral medications for improving health outcomes. The 2 trials differed in several ways. They assessed different medications (valacyclovir and acyclovir), recruited from different sources, and used different tests to establish HSV-2 infection. Both assessed outcomes over a short time (6-8 weeks) and relied on self-report to ascertain symptom occurrence. This duration is likely inadequate to evaluate whether antiviral medications reduce symptom incidence among populations who have been asymptomatic. Results were inconsistent and imprecise; 1 trial found benefit for valacyclovir compared with placebo for reducing viral shedding and symptom occurrences,26 and the other found no statistically significant differences between groups.27
Similarly, the 2 RCTs assessing preventive antiviral medications for reducing HSV-2 transmission between serodiscordant partners were heterogeneous and found inconsistent results. One enrolled immunocompetent couples from primarily industrialized countries,28 while the other enrolled couples discordant for both human immunodeficiency virus and HSV-2 from sub-Saharan Africa.32 The trials assessed different medications (valacylovir and acyclovir) and over different durations (8 months and a median of 18 months). One trial found benefit for valacyclovir compared with placebo for reducing symptomatic HSV-2 infection and HSV-2 seroconversion in the susceptible partner; however, the magnitude of benefit for symptom reduction was modest, and results were imprecise (0.5% vs 2.2%, respectively; hazard ratio, 0.25 [95% CI, 0.08-0.75]).
One trial assessed harms of medications; adverse events were similar between groups randomized to valacyclovir and placebo.26 Other review studies have concluded that there are few harms in nonpregnant adults.34
This review had limitations. Studies were required to compare an FDA-approved currently available serologic screening test with Western blot. We did not evaluate other comparisons, such as a serologic test compared with a viral polymerase chain reaction swab or culture or with another commercially available serologic test, to determine concordance. We limited the assessment to studies enrolling people with no current or prior symptoms and found only 2 trials. For people with frequent symptomatic recurrences of genital herpes (more than 4 episodes per year), antiviral medications have been shown to reduce the frequency of recurrences; however, the magnitude of effect is uncertain, and the quality of evidence is low. A Cochrane review published in 2014 evaluated the efficacy of antiviral medications (acyclovir, famciclovir, and valacyclovir) to suppress genital herpes outbreaks in nonpregnant adults34; 22 trials were included, and the risk of bias was considered high for half of the studies and unclear for the other half. The authors concluded that there was low-quality evidence that the risk of having at least 1 clinical recurrence was reduced with acyclovir (9 parallel-group trials, n = 2049; RR, 0.48 [95% CI, 0.39-0.58]), valacyclovir (4 trials, n = 1788; RR, 0.41 [95% CI, 0.24-0.69]), or famciclovir (2 trials, n = 732; pooled RR, 0.57 [95% CI, 0.50-0.64]).34 It is unclear whether these results would apply to people who have less frequent recurrences (or who are asymptomatic).
Serologic screening for genital herpes is associated with a high rate of false-positive test results and potential psychosocial harms. Evidence from RCTs does not establish whether preventive antiviral medication for asymptomatic HSV-2 infection has benefit.
Corresponding Author: Cynthia Feltner, MD, MPH, Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, 725 Martin Luther King Jr Blvd, CB#7295, Chapel Hill, NC 27599 (email@example.com).
Author Contributions: Dr Feltner 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: Feltner, Grodensky, Ebel, Ashok, Jonas.Acquisition, analysis, or interpretation of data: Feltner, Grodensky, Ebel, Middleton, Harris, Ashok, Jonas.Drafting of the manuscript: Feltner, Grodensky, Ebel, Middleton, Ashok, Jonas.Critical revision of the manuscript for important intellectual content: Feltner, Grodensky, Ebel, Harris, Ashok, Jonas.Statistical analysis: Feltner.Obtained funding: Feltner, Jonas.Administrative, technical, or material support: Grodensky, Ebel, Middleton, Jonas.Supervision: Feltner, Jonas.
Conflict of Interest Disclosures: All authors have completed and submitted ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This research was funded under contract HHSA-290-2012-00015-I, Task Order No. 6, from the Agency for Healthcare Research and Quality (AHRQ), US Department of Health and Human Services, under a contract to support the USPSTF.
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. 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: David Meyers, MD (AHRQ), Evelyn Whitlock, MD, MPH (Kaiser Permanente Research Affiliates EPC), and Meera Viswanathan, PhD; Carol Woodell, BSPH; Rachel Weber, PhD; Christiane Voisin, MSLS; Sharon Barrell, MA; and Loraine Monroe (RTI International–University of North Carolina EPC).
Additional Information: A draft version of the full evidence report underwent external peer review from 4 content experts (Rhoda A. Morrow, PhD, University of Washington, Seattle; David Kimberlin, MD, University of Alabama at Birmingham; Jeffrey Peipert, MD, MPH, MHA, Washington University, St Louis, MO; and 1 anonymous content expert) and 4 partner reviewers from the Eunice Kennedy Shriver National Institute of Child Health and Human Development; National Institute of Allergy and Infectious Diseases; National Center for HIV, Hepatitis, STD, and TB Prevention (Centers for Disease Control and Prevention); and Office of the Surgeon General, G-3/5/7, Health and Wellness. Comments from reviewers 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.