In most sexually transmitted disease (STD) clinics in the United States, physical examination is a routine part of care for clients, regardless of risk and symptoms. To meet the increasing demand for STD services, testing-only visits, in which specimens for testing of Chlamydia trachomatis (CT) and Neisseria gonorrhoeae (GC) are self-obtained by patients and the clinical encounter does not include a full physical examination by a clinician, have been used to increase the efficiency of STD services.1–3 Testing-only visits capitalize on the opportunity created by nucleic acid amplification testing (NAAT) for CT and GC because these tests can be performed on self-obtained urine and vaginal swab specimens by men and women, respectively.4–7 During testing-only visits, tests for syphilis and human immunodeficiency virus (HIV) can also be offered.
When a significant proportion of STD patients are served through testing-only visits, clinics can potentially accommodate a larger number of clients without additional personnel cost.2,3 Patients can also benefit from testing-only visits because the duration of clinic visit can be significantly reduced.2 When physical examination by qualified health care professionals limits the access to STD services, offering testing-only visits to specific patient groups can be an effective strategy to free up the capacity of STD clinics. However, testing-only visits almost certainly will lead to missed opportunities to diagnose STDs, STD syndromes, and infections other than CT, GC, syphilis, and HIV.
We conducted a self-administered survey in 3 urban STD clinics and abstracted the patients’ medical charts to identify diagnoses patients had received during their standard care visits. Using data from the chart abstraction, we estimated the prevalence of selected diagnoses that could become “missed diagnoses” if different client groups had used testing-only visits instead of standard care. We also examined client characteristics associated with common missed diagnoses.
Study Population and Data Collection
All clients who attended the STD clinic in New Orleans, Louisiana; Jackson, Mississippi; and St Louis County, Missouri, between June and September 2008 were asked to participate. All 3 STD clinics are designated public clinics for STD services. While waiting to be seen, participants completed a self-administered questionnaire in private. The survey contained questions about demographics, reason(s) for visit, specific STD-related symptoms, and risk behaviors in the previous 3 months.
Unless the patients were referred to the clinic for STD treatment or HIV testing only, patients received an examination per standard STD care protocol,8 regardless of symptoms and risk behaviors.
Diagnostic procedures were similar at all 3 clinics. In women, samples collected during examination included a vaginal or cervical swab for GC and CT combo–NAAT assays (Aptima Combo 2; Gen-Probe, San Diego, CA), an endocervical swab for gram stain, and a vaginal swab for wet-mount microscopy to detect trichomoniasis, bacterial vaginosis (BV), and vulvovaginal candidiasis (VVC). Bacterial vaginosis was diagnosed according to the criteria of Amsel et al.9 In men, examinations included urethral swabs for methylene blue stain at clinics in New Orleans and Jackson,10 or gram stain at St Louis clinic to diagnose urethritis (nongonococcal urethritis), and urethral swabs or urine for GC and CT combo–NAAT assays (Aptima Combo 2; Gen-Probe). In both men and women, blood was collected for Treponema pallidum and HIV serologic testing. The diagnosis of genital herpes was by clinical impression or viral culture, whereas the diagnosis of genital warts was by clinical impression. Participants’ medical charts including laboratory records were reviewed about 2 weeks after clinic visit. A chart abstraction form was used to collect information on clinical findings, diagnoses, and treatment during the clinic visit. After the survey questionnaire and the abstraction form were matched by the date of survey and patients’ medical record number, we entered the patients’ data into a database for analyses.
Missed Diagnoses of Testing-Only Visits
Because only CT, GC, syphilis, and HIV are tested during testing-only visits, such visits may result in various missed diagnoses. Of all the STDs that may be missed by testing-only visits, genital herpes, genital warts, pharyngeal or rectal GC, trichomoniasis, and pelvic inflammatory diseases (PIDs) in women are of greatest health risk and public health concern, and their prevalence was included in this assessment. By definition, testing-only visits will not include vaginal swabs for wet-mount microscopy or urethral swabs for methylene blue stain, which would lead to missed diagnoses of BV and VVC in women and nongonococcal urethritis in men.
Although a clinician visit may be valuable also because it provides an opportunity for patient education, counseling, reassurance, and so on, this study was not designed to assess possible missed opportunities in these areas. Our study also did not consider other services at an STD clinic, such as Papanicolaou test or family planning services.
Self-Reported Symptoms and Risk Behaviors in the Past 3 Months
Many possible STD-related symptoms are vague and not specific. In this survey, participants’ possible responses for STD-related symptoms included the following: (1) discharge from the penis or vagina (discharge), (2) painful or burning urination (dysuria), (3) sores or ulcers on the penis or vagina, (4) rashes, (5) itching or tingling around the penis or vagina (itching), (6) foul odor in the penis or vagina, (7) having no any symptoms, and (8) other symptoms (specify). In the final analyses, symptoms 1 and 2 above were combined as “discharge/dysuria” and 3 and 4 as “sore/ulcer/rash.” If a participant reported a similar symptom in their own words, such as “drip” or “red bumps” under “other symptoms,” these symptoms were recorded as discharge/dysuria and sore/ulcer/rash, respectively. Survey participants’ possible responses for risk behaviors included the following: (1) new partner, (2) sex with more than 1 partner (multiple partners), (3) my partner may have other partner(s) (nonexclusive partner), and (4) other risk factors such as men who have sex with men. Survey participants were instructed to check all symptoms or risk behaviors they had, so the responses were not mutually exclusive.
All analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC). The differences in the prevalence of missed diagnoses were tested using the χ2 test. The correlates that were statistically significant at P < 0.05 level from bivariate analyses were entered into logistic regression models to identify variables that are independently associated with different missed diagnoses. We used the backward selection approach with criterion of selection at P = 0.05 to yield parsimonious models for each undesirable consequence. The Hosmer-Lemeshow statistic was used to measure models’ goodness of fit. The Hosmer-Lemeshow statistic measures whether model prediction is significantly different from observed values, but a model that has passed the goodness-of-fit test does not necessarily explain much of the variance in the dependent variable.
To evaluate the discrimination ability of models in predicting the missed diagnoses, we used the Herrell concordance index (C-index), which is an extension of the receiver operating characteristic curve area and describes the performance of a given model applied to the population under consideration.11 The C-index ranges from 0.5 (no discrimination) to 1 (perfect discrimination).
Characteristics of Study Participants
We estimated that the survey response rate was high in all 3 clinics (approximately 95%). Most patients (>90%) were African Americans. A total of 2887 unique individuals from the 3 STD clinics participated in the survey. We excluded 305 individuals because of the following: (a) pregnancy (n = 88), (b) medical charts not locatable (n = 49), (c) treatment-only or HIV testing-only visits (n = 100), or (d) participant’s sex was not known (n = 68). The sample size for the final analyses was 2582 (89.4% of surveyed); the number of participants from each study site was similar.
The median age was 24 years, and 50% of participants were women (Table 1). Some participants were not tested for all STDs; we defined the “prevalence” of a condition as the percentage of patients with an infection or a syndrome diagnosed and recorded in the chart. Both CT and GC were the most common infections in the study population, and the prevalence was higher in men than in women (Table 1). Syphilis and viral STDs (genital herpes and genital warts) were relatively uncommon. We identified 4 cases of PID (2 of the 4 cases also had GC infection); there was no rectal or pharyngeal GC recorded in patients’ charts.
Prevalence of Possible Missed Diagnoses by Client Characteristics
We estimated the prevalence of possible missed diagnoses in men and women separately. We first analyzed the overall prevalence of these possible missed diagnoses and then compared the prevalence by client characteristics, namely, age, self-reported STD-related symptoms, and risk behaviors in the previous 3 months (Tables 2 and 3).
In women, 45.3% were younger than 25 years (Table 2). The most common STD-related symptom was discharge or dysuria. Only 25.9% of women reported “do not have any symptoms.” Among the risk behaviors in the past 3 months, no-exclusive partner(s) was most common, reported by 34.6% of women. About one third of women (31.7%) reported no risk behavior.
The diagnosis of genital herpes or genital warts was uncommon in women, with an overall prevalence of 3.2%. Among the client characteristics, only self-reported STD-related symptoms were significantly associated with the diagnoses of viral STDs. Among women who reported genital sores or ulcers or rashes (sore/ulcer/rash), 19.9% were diagnosed as having genital herpes/warts, compared with 1.2% in those who reported no such symptoms (P < 0.05). Among clients who reported itching or tingling (itching/pruritus), 4.9% were diagnosed as having genital herpes/warts (Table 2).
Overall, 9.6% of women received the diagnosis of trichomoniasis (Table 2). Consistent with previous reports,12–14 the prevalence of trichomoniasis was higher in older women than in younger women (11.9% vs. 7.7%, P < 0.05). Sexually transmitted disease–related symptoms of discharge/dysuria and foul odor were associated with significantly higher prevalence of trichomoniasis, but the prevalence was still high in those who reported no such symptoms. If women who reported “Do not have any symptoms” were triaged to a testing-only visit, trichomoniasis would have been missed in 6.3%, or 1 in 16 clients (Table 2).
We defined symptomatic BV as a diagnosis of BV plus the reporting of any symptoms of discharge/dysuria, itching, or foul odor. Although 42.1% of women received a diagnosis of BV, the prevalence of symptomatic BV, for which treatment is indicated, was only 29.6%. In addition, 14.8% of women received a diagnosis of VVC. Together, 41.0% of women received a diagnosis of VVC or symptomatic BV (Table 2). The prevalence of symptomatic BV or VVC was high, regardless of client characteristics. The lowest prevalence (9.3%) was observed in women who reported “Do not have any symptoms.”
We similarly assessed the prevalence of diagnoses in men that could become missed diagnoses if they had used testing-only visits (Table 3). Compared with women, men were less likely to report STD-related symptoms: only 27.2% reported discharge/dysuria compared with 47.6% in women; nearly 40% of men reported “Don’t have any symptoms,” higher than the 25.9% in women (both P < 0.0001). In contrast, higher percentages of men reported having had a new partner or multiple partners in the past 3 months than did women (both P < 0.0001).
In men, the overall prevalence of genital herpes/genital warts (6.6%) was significantly higher than that in women (P < 0.0001). Overall, 19.3% of men received a diagnosis of non-GC/CT urethritis, which we defined as urethritis based on urethral smear with methylene blue stain and had negative test results for GC and CT. The prevalence of non-GC/CT urethritis was high, regardless of symptoms, and ranged from 15.7% in those who reported no symptoms to 32.6% in those who reported malodor. It should be noted that the prevalence of non-GC/CT urethritis did not differ by risk behaviors (Table 3).
Of note, men who have sex with men was reported by 5.1% of male participants. Owing to the small number, we did not estimate the prevalence of missed diagnoses in this population.
Attempt to Develop Triage Algorithms
We conducted multivariate logistic regression analyses to identify independent predictors for each diagnosis that could become a missed diagnosis if testing-only visits were used. Only statistically significant predictors from bivariate analyses (variables associated with bold numbers in Tables 2 and 3) were included in the logistic regression models.
In men and women, only age and STD-related symptoms were significant predictors of various diagnoses (Table 4). For example, trichomoniasis was less common in women younger than 25 years and was more common in those who reported symptoms of discharge/dysuria or foul odor. In contrast, trichomoniasis was less common in women who reported the symptoms of sore/ulcer/rash. However, the discriminatory power of these predictors was low, with a C-index of only 0.64. Further analyses showed that among women who reported symptoms of both discharge/dysuria and foul odor, the prevalence of trichomoniasis was quite high, at 18.4%. However, in women younger than 25 years who did not report discharge/dysuria nor foul odor, the prevalence of trichomoniasis was still 6.1%, indicating limited discriminatory power even when age and multiple symptoms are combined.
The C-indexes for all models were far below 1 (perfect discrimination), with the highest index at 0.79. The model for non-GC/CT urethritis was barely above 0.5 (no discrimination) (Table 4).
With the advent of testing methods that use self-collected urine or vaginal swab specimens for screening of the most common bacterial STDs, testing-only visits have a great potential to improve the efficiency of STD clinics. This innovative approach is particularly important for clinics that have to turn patients away due to budget cuts or reduced personnel resources. The potential of using testing-only visits to increase clinic capacity to control serious STDs should be evaluated. In our study population with high prevalence of GC and CT infections in both men and women, if the clinics have to turn patients away, on average, 8.5% (approximately 1 in 12) would leave the clinic with an untreated GC infection and 15.3% (approximately 1 in 7) with an untreated CT infection.
Our study also shows that testing-only visits can lead to substantial missed diagnoses such as trichomoniasis in women. Obviously, our findings about missed diagnoses depend on “standard of care” during a typical evaluation. For example, in clinics where routine visits do not include wet-mount examination for asymptomatic women, the number of missed diagnoses will be lower. When women attending the study clinic forgo routine physical examination and traditional saline wet-mount examination, the opportunity to diagnose trichomoniasis would be missed. Even if testing-only visits are restricted to women who do not report any STD-related symptoms, 6.3% of women with asymptomatic or subclinical trichomoniasis would remain undiagnosed and untreated, potentially resulting in an increased risk for HIV infection, PID, and other health consequences.15–17 New advances in testing technology, such as NAATs for the detection of trichomoniasis based on self-collected specimens, could potentially be incorporated into testing-only visits.18–21 At present, some STDs or syndromes would only be revealed by physical examination, and some laboratory testing can only be done on specimens collected by a clinician during physical examination. Therefore, some lost opportunities for diagnosing conditions such as BV and VVC are unavoidable with testing-only visits. Another disadvantage of testing-only visits is the potential to miss same-day treatment of GC, CT, or syphilis. Because of the risk of continued transmission, which can be substantial given the high prevalence of risk behaviors in the study population and the risks associated with delayed treatment, same-day treatment should be an important goal for STD care among patients attending STD clinics.
It has been shown that patients’ characteristics and behaviors can be used, to some extent, to help minimize the missed diagnoses of testing-only visits.1–3 Conventionally, physicians have to examine patients with symptoms when there is a triage system. Although predictors such as age, STD-related symptoms, and risk behaviors were significantly associated with 1 or more diagnoses (Tables 2 and 3), it is clear that these predictors including the various STD-related symptoms have limited discriminatory power in the high-risk population in our study, even when used in combinations. Furthermore, the predictors vary for different missed diagnoses. Our failed attempt to develop suitable algorithms demonstrated again that it is unrealistic to use a single algorithm to eliminate or minimize all the missed diagnoses of testing-only visits. Our study has several limitations. First, the prevalence of STDs and STD syndromes was based on diagnoses and laboratory reports recorded in patients’ charts. Per clinic policy, all patients receiving standard care should have been tested for GC and CT; we assumed that patients who were not tested or those for whom testing results were unknown/indeterminate all had negative test results. Second, this was a hypothetical assessment. We did not consider patients’ acceptance of testing-only visits, which can be relatively low without patient education,22 and we included in our assessment some diagnoses that may or may not be important to miss for clinical reasons such as symptomatic vaginosis and VVC. Third, the tradeoff between missed diagnoses and having to turn away patients is context specific and may depend on local disease burden and whether other clinical services are available in the area. Last, due to the relatively small sample size from individual sites, we did not compare and contrast the differences between clinics.
With the changes and advances in the laboratory detection of common STDs, testing-only visits may become increasingly popular as an important strategy to improve the efficiency of STD clinics. In populations with high burden of diseases, the use of such visits may result in missed opportunities for diagnosing some STDs. Our results strongly suggest that the magnitude of missed diagnoses from the testing-only strategies should be considered before implementing such approaches in an STD clinic.
1. Heijman TL, Van der Bij AK, De Vries HJ, et al.. Effectiveness of a risk-based visitor-prioritizing system at a sexually transmitted infection outpatient clinic. Sex Transm Dis 2007; 34: 508–512.
2. Shamos SJ, Mettenbrink CJ, Subiadur JA, et al.. Evaluation of a testing-only “express” visit option to enhance efficiency in a busy STI clinic. Sex Transm Dis 2008; 35: 336–340.
3. Paneth-Pollak R, Schillinger JA, Borrelli JM, et al.. Using STD electronic medical record data to drive public health program decisions in New York City. Am J Public Health 2010; 100: 586–590.
4. Hobbs MM, van der Pol B, Totten P, et al.. From the NIH: Proceedings of a workshop on the importance of self-obtained vaginal specimens for detection of sexually transmitted infections. Sex Transm Dis 2008; 35: 8–13.
5. Hoebe CJ, Rademaker CW, Brouwers EE, et al.. Acceptability of self-taken vaginal swabs and first-catch urine samples for the diagnosis of urogenital Chlamydia trachomatis and Neisseria gonorrhoeae with an amplified DNA assay in young women attending a public health sexually transmitted disease clinic. Sex Transm Dis 2006; 33: 491–495.
6. Schachter J, Chernesky MA, Willis DE, et al.. Vaginal swabs are the specimens of choice when screening for Chlamydia trachomatis and Neisseria gonorrhoeae: Results from a multicenter evaluation of the APTIMA assays for both infections. Sex Transm Dis 2005; 32: 725–728.
7. Chernesky MA, Martin DH, Hook EW, et al.. Ability of new APTIMA CT and APTIMA GC assays to detect Chlamydia trachomatis and Neisseria gonorrhoeae in male urine and urethral swabs. J Clin Microbiol 2005; 43: 127–131.
8. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines, 2006. MMWR 2006; 55 (No. RR-11).
9. Amsel R, Totten PA, Spiegel CA, et al.. Nonspecific vaginitis: Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983; 74: 14–22.
10. Taylor SN, DiCarlo RP, Martin DH. Comparison of methylene blue/gentian violet stain to gram’s stain for the rapid diagnosis of gonococcal urethritis in men. Sex Trans Dis 2011; 38: 995–996.
11. Harrell FE Jr, Califf RM, Pryor DB, et al.. Evaluating the yield of medical tests. JAMA 1982; 247: 2543–2546.
12. Sutton M, Sternberg M, Koumans EH, et al.. The prevalence of Trichomonas vaginalis infection among reproductive-age women in the United States, 2001–2004. Clin Infect Dis 2007; 45: 1319–1326.
13. Owusu-Edusei K, Tejani MN, Gift TL, et al.. Estimates of the direct cost per case and overall burden of trichomoniasis for the employer-sponsored privately insured women population in the United States, 2001 to 2005. Sex Transm Dis 2009; 36: 395–399.
14. Wendel KA, Erbelding EJ, Gaydos CA, et al.. Trichomonas vaginalis polymerase chain reaction compared with standard diagnostic and therapeutic protocols for detection and treatment of vaginal trichomoniasis. Clin Infect Dis 2002; 35: 576–580.
15. McClelland RS, Sangare L, Hassan WM, et al.. Infection with Trichomonas vaginalis increases the risk of HIV-1 acquisition. J Infect Dis 2007; 195: 698–702.
16. Van Der Pol B, Kwok C, Pierre-Louis B, et al.. Trichomonas vaginalis infection and human immunodeficiency virus acquisition in African women. J Infect Dis 2008; 197: 548–554.
17. Moodley P, Wilkinson D, Connolly C, et al.. Trichomonas vaginalis is associated with pelvic inflammatory disease in women infected with human immunodeficiency virus. Clin Infect Dis 2002; 34: 519–522.
18. Wiesenfeld HC, Lowry DL, Heine RP, et al.. Self-collection of vaginal swabs for the detection of chlamydia, gonorrhea, and trichomoniasis: Opportunity to encourage sexually transmitted disease testing among adolescents. Sex Transm Dis 2001; 28: 321–325.
19. Joyner JL, Douglas JM Jr, Ragsdale S, et al.. Comparative prevalence of infection with Trichomonas vaginalis among men attending a sexually transmitted diseases clinic. Sex Transm Dis 2000; 27: 236–240.
20. Van Der Pol B, Kraft CS, Williams JA. Use of an adaptation of a commercially available PCR assay aimed at diagnosis of chlamydia and gonorrhea to detect Trichomonas vaginalis in urogenital specimens. J Clin Microbiol 2006; 44: 366–373.
21. Nye MB, Schwebke JR, Body BA. Comparison of APTIMA Trichomonas vaginalis transcription-mediated amplification to wet mount microscopy, culture, and polymerase chain reaction for diagnosis of trichomoniasis in men and women. Am J Obstet Gynecol 2009; 200: 188.e1–188.e7.
22. Howard EJ, Xu F, Taylor SN, et al.. Screening methods for Chlamydia trachomatis and Neisseria gonorrhoeae infections in STD clinics: what do patients prefer? Sex Trans Infect 2011; 87: 149–151.