Testing for sexually transmitted infections using vaginal swabs is noninferior to testing using endocervical swabs
Krause A, Manteuffel J, and Miller J. Testing for sexually transmitted infections using vaginal swabs is noninferior to testing using endocervical swabs. Academic Emergency Medicine 2020; 27:S49.
Academic emergency medicine
Background and Objectives: Emergency Department (ED) Testing for Sexually Transmitted Infections (STI) in females has typically been performed with a pelvic exam and an endocervical swab sent for nucleic acid amplification testing (NAAT). The Centers for Disease Control (CDC) recommends vaginal swabs as an appropriate sample type when testing for STI, even when a pelvic exam is performed. For female screening, self-collected vaginal swabs have the same sensitivity and specificity to those collected by a clinician. There is a paucity of data examining the use of vaginal swabs in the ED for STI testing. We implemented a new ED protocol for STI using vaginal swabs across a health system. In this study, we hypothesized that detection rates would be non-inferior to pre-implementation.
Methods: This was a quasi-experimental pre-post study using standardized data from the electronic health record across a health system. We assessed 10 months of pre-implementation data and 10-months of post-implementation data after the intervention. The intervention consisted of a new protocol that introduced vaginal swabs to test for STIs rather than endocervical swabs. Patient eligibility was any female that received testing for STIs. Analysis consisted of non-inferiority testing for proportion with a non-inferiority margin of 1 percentage point absolute difference in detection rates of STI.
Results: The study included 22,291 encounters across 9 EDs within a health system. There were 11,732 encounters during the pre-implementation phase and 10,559 post-implementation. The rate of detection of Chlamydia trachomatis was 7.5% pre-implementation and 7.6% post-implementation (between-group difference, 0.1 percentage points; 90% CI, -0.7 - 0.4; p=0.009 for noninferiority). The rate of detection of Neisseria gonorrhea was 3.1% pre-implementation and 3.6% post-implementation (between-group difference, 0.5 percentage points; 90% CI, -0.8 - 0.04; p=0.011 for noninferiority).
Conclusion: An ED protocol using vaginal swabs for STI testing is non-inferior to the use of endocervical swabs for STI detection rates in female patients.