When Is Any Test Better Than No Test for Chlamydia?

Abstract

T he suboptimal performance of many antigen-based chlamydia detection assays had long been a concern of Dr Julius Schachter, an innovator in the field of chlamydia diagnostics. He had reviewed the submitted article and was in the process of writing a commentary to accompany the accepted article when he was suddenly struck by the SARS-CoV-2 virus and subsequently died of complications of this disease. His shoes are too large to fill, but in his memory, I have written this commentary to include the thoughts he and I discussed as well as my own perspective.The article by Su and colleagues 1 in this issue of Sexually Transmitted Diseases once again raises the issue of the minimal performance requirements for a point-of-care (POC) test for Chlamydia trachomatis. More than 20 years ago, Gift et al. 2 described the implications of test and treat facilitated by a theoretical POC with low sensitivity. In Gift's model, in a population with an 8% prevalence of chlamydia and a 74% return-for-treatment rate, a POC with sensitivity as low as 65% would result in an increase in cases treated if compared with using a 95% sensitive laboratory-based test when no cases were treated in the absence of diagnostic test results. The potential public health benefit in terms of rapid and accurate treatment is a function of population prevalence, return for treatment rates, and assay sensitivity. The "population" in question may be the general population or the specific population served by a particular health care setting. In a more recent model developed by Rönn and colleagues, 3 additional variables were considered, including test turnaround time, testing setting, and treatment delay. In this model, the authors used very conservative parameter estimates and still described public health benefits to the general population (prevalence <3%) with a theoretical POC that was 10% less sensitive than laboratory-based nucleic acid amplification tests (NAATs). Looking at this model of reduced incidence and reduced pelvic inflammatory disease to achieve general population benefits, a test with 65% sensitivity would likely not be useful.Both of these analyses were based on considerations in resource wealthy settings. However, most treatable sexually transmitted infections (STIs) occur in resource-limited settings. Syndromic management is often the standard of care in such settings, and Wi et al., 4 estimated that as many as 3.6% to 21.6% of infections are undertreated and 30.1% to 40.9% of patients are overtreated when using this management strategy. With syndromic management, people are assessed for STIs exclusively based on symptoms and clinical signs, so asymptomatic people are excluded from management by definition and therefore undertreated. The proportion undertreated as estimated by Wi et al. does not include these infections. For chlamydia, this is a serious concern because more than half of infections are estimated to be asymptomatic, and when left untreated, these infections can lead to pelvic inflammatory disease, tubal-factor infertility...