P796 Reassessing the gram stain smear (GSS) polymorphonuclear leukocyte (PMN) cutoff for diagnosing non-gonococcal urethritis (NGU)

Background Recommended cutoffs for PMNs per high-power field (hpf) to define NGU vary. CDC treatment guidelines specify ≥2 PMNs/hpf. Other guidelines recommend ≥5 PMNs/hpf. Methods From 08/2014-08/2018, we enrolled symptomatic and asymptomatic male STD clinic patients ≥16 years with exclusively female partners in the past year. Men with gonorrhea or antibiotic use in the past month were excluded. We collected a urethral swab for GSS and urine for Chlamydia trachomatis(CT) and Mycoplasma genitalium (MG) testing (Aptima, Hologic). We calculated Youden’s Index (J=sensitivity+specificity-1), which maximizes sensitivity and specificity, and calculated the proportions of CT/MG cases missed and cases treated in the absence of CT/MG (test-negative) for three PMN/hpf cutoffs. CT/MG co-infections (N=3) were excluded. Results Among 369 participants, median age was 32 (range 17–71), 53% were white, and 25% were black. Among all men with 0-1, 2-4, 5-9, and ≥10 PMNs/hpf, CT prevalence was 1%, 5%, 11%, and 26%, respectively; MG prevalence was 5%, 3%, 15%, and 17%. J was maximized at ≥5 PMNs/hpf for CT, MG, and CT/MG. Thirteen percent, 17%, and 33% of CT/MG cases were missed at the ≥2, ≥5, and ≥10 PMNs/hpf cutoffs, respectively; 45%, 33%, and 21% of test-negative cases were treated. Among symptomatic men (N=166) with 0-1, 2-4, 5-9, and ≥10 PMNs/hpf, CT prevalence was 0%, 20%, 12%, and 31%, respectively; MG prevalence was 9%, 0%, 18%, and 19%. J was maximized at ≥5 PMNs/hpf for MG, and ≥10 PMNs/hpf for CT and CT/MG. Five percent, 8%, and 25% of CT/MG cases were missed at the ≥2, ≥5, and ≥10 PMNs/hpf cutoffs, respectively; 72%, 64%, and 43% of test-negative cases were treated. Conclusion The increase in missed CT/MG cases between the ≥2 PMNs/hpf cutoff and ≥5 PMNs/hpf cutoff was minimal; the ≥5 PMNs/hpf cutoff treats fewer cases without CT/MG. The ≥5 PMNs/hpf cutoff appears optimal in this population. Disclosure No significant relationships.