AbstractBackgroundSexual history does not accurately identify those with extragenital Neisseria gonorrhoeae (NG) and Chlamydia trachomatis (CT), so universal extragenital sampling is recommended. Nucleic acid amplification tests (NAATs) are expensive. If urogenital, plus rectal and pharyngeal, samples are analyzed, the diagnostic cost is trebled. Pooling samples into 1 NAAT container would cost the same as urogenital samples alone. We compared clinician triple samples analyzed individually with self-taken pooled samples for diagnostic accuracy, and cost, in men who have sex with men (MSM) and females.MethodsThis was a prospective, convenience sample in United Kingdom sexual health clinic. Randomized order of clinician and self-samples from pharynx, rectum, plus first-catch urine (FCU) in MSM and vulvovaginal swabs (VVS) in females, for NG and CT detection.ResultsOf 1793 participants (1284 females, 509 MSM), 116 had NG detected (75 urogenital, 83 rectum, 72 pharynx); 276 had CT detected (217 urogenital, 249 rectum, 63 pharynx). There was no difference in sensitivities between clinician triple samples and self-pooled specimens for NG (99.1% and 98.3%), but clinician samples analyzed individually identified 3% more chlamydia infections than pooled (99.3% and 96.0%; P = .027). However, pooled specimens identified more infections than VVS/FCU alone. Pooled specimens missed 2 NG and 11 CT infections, whereas VVS/FCU missed 41 NG and 58 CT infections. Self-taken pooled specimens were the most cost-effective.ConclusionsFCU/VVS testing alone missed many infections. Self-taken pooled samples were as sensitive as clinician triple samples for identifying NG, but clinician samples analyzed individually identified 3% more CT infections than pooled. The extragenital sampling was achievable at no additional diagnostic cost to the FCU/VVS.Clinical Trials RegistrationNCT02371109.