Hot off the press: Self‐obtained vaginal swabs for sexually transmitted infection testing

Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT) are the two most common sexually transmitted infections (STIs) reported in the United States. Emergency departments (EDs) are diagnosing an increasing percentage of NG/CT cases compared to STI clinics.1, 2 The criterion standard for NG/CT diagnosis is nucleic acid amplification testing (NAAT), with the collection method being provider-performed endocervical sampling. Provider-performed sampling is uncomfortable for patients and has numerous other limitations, including the need for exam rooms, an available provider, and often a female chaperone. These limitations can delay sample collection and can decrease efficiency in a busy ED. The need for universal pelvic examination in the ED to perform STI testing has also come under increasing scrutiny.3 Vaginal sample collection with self-obtained swabs was first developed and researched in non-ED settings. These studies demonstrated comparable sensitivity for NG/CT diagnosis when compared to provider performed swabs and high patient acceptability.4-8 However, those studies were performed in a wide range of non-ED settings and mostly examined asymptomatic screening rather than STI testing in an acute care environment. This study compares the sensitivity of provider performed and self-obtained swabs in an ED setting and explores patient preference. This is a prospective observational cohort study performed in a single ED. Participants were female patients 18 years of age or older who were judged by the treating practitioner to need NG/CT testing. Treatment arm participants obtained vaginal swabs for NG/CT themselves, while control arm participants had provider-performed endocervical swabs. The primary outcome was noninferiority of self-obtained swabs for NG/CT, with noninferiority being defined if sensitivity was 90% or above. Self-obtained swabs were highly sensitive in the diagnosis of NG/CT, and patients generally found self-swabbing to be acceptable. This was a good-quality study that met almost all of the elements of a quality checklist for observational trials. There were still some limitations to the study, including it being a single-center study, the selection bias introduced by convenience sampling (no overnight patients), and the trial narrowly missing the preset enrollment goal of 615 patients (due to the COVID pandemic suspending research activities). Additionally, the study used a rapid turnaround NAAT swab, which may not be available in many EDs. The biggest methodologic point to consider is that, although the conclusion of the paper states that “self-obtained vaginal swabs were found to be noninferior to provider-performed endocervical sampling,” the results did not actually meet the study's definition of noninferiority. The 95% confidence intervals (CIs) indicate that self-swabs could have a sensitivity as low as 88%, which is below the threshold of 90% that the authors set for noninferiority. Therefore, the conclusion should technically have been that self-swabs were not noninferior. However, that does not mean that self-swabs are definitely worse than those performed by providers. The point estimate is that self-swabs were 95% sensitive, and the 95% CI means that they could be even more accurate than that. Had the authors met their sample size, the CIs would have been smaller, and it is possible this study would have reached the threshold for noninferiority. It is reasonable to make clinical decisions based on the reported point estimate of 95% sensitivity, but we must keep in mind the uncertainty that remains given the CIs around that result. Overall, 515 patients completed both self-swabs and provider-performed swabs. The mean age was 30.7 years and 52% were of Hispanic ethnicity. The overall prevalence for either NG or CT or both was 17% (86/515). Of these patients with STIs, 34% (29/86) had infection with NG, 54% (47/86) with CT, and 12% (10/87) had a coinfection. For the primary outcome, self-swabs had a sensitivity of 95% (95% CI = 88% to 99%) for the detection of NG/CT when compared to provider performed swabs. Secondary outcomes revealed an excellent kappa of 93%, and self-swab sensitivities for NG and CT were 97% and 94%, respectively. A total of 93% of patients felt that collecting a self-sample is acceptable, but 28% of patients were concerned about doing their own swabs incorrectly. A total of 26% of patients refused self-swabs due to reasons clearly related to the collection procedure (e.g., “uncomfortable performing self-swab,” “worried might do it incorrectly,” “prefers physician”) rather than the consent process. Although this study did not technically demonstrate that self-swabs were noninferior to provider-performed swabs, the results are promising, and we feel that self-swabs are a reasonable option for most patients who need STI testing in the ED. This is provided there is not another specific reason for which pelvic examination is needed and the patient can understand the instructions and correctly perform a self-swab. I offer this all the time, as long as the patient doesn't have any pelvic/abdominal pain or other symptoms that make me think they need an exam. And if I do think they need an exam and they refuse it, I still offer this, I just also clearly explain the risks of skipping the exam. I have also run into resistance from patients to do a vaginal swab when they know there is a urine PCR test available. I have found usually they are more open to a self-swab than provider performed endocervical swab but still some are hesitant. Thank you Brian for the reference to Sean Wilson's article with stats to support my reasoning behind recommending the self-swab. Speaking of urine samples, despite what the data showed in the company-sponsored studies, Sean Wilson did a nice ED-based study showing that the rapid Cepheid assay had only an 87% sensitivity for gonorrhea, and 85% for chlamydia from urine samples. The prospect of missing 1 in 7 chlamydia and 1 in 8 gonorrhea is not good. That is part of what spurred use to examine self-swabs rather than urine, and why we are now working on having patients collect them very early in the triage process. Although this study failed to definitively demonstrate the noninferiority of self-swabs, the high sensitivity means that they remain a reasonable alternative to pelvic examination for ED patients in whom you suspect an STI. The authors have no potential conflicts to disclose.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.