Objective: Sexually transmitted
Chlamydia trachomatis (CT) is often asymptomatic requiring screening with less invasive samples, using nucleic acid amplification tests (NAATs) performed on automated instruments. The objectives were to compare test performance and instrument workflow and maintenance using female first void urine (FVU) and self-collected vaginal swabs (SCVS) by APTIMA Combo 2 (AC2) on Panther (Hologic|GenProbe), to cobas CT/NG on cobas 4800 (Roche) and ProbeTec ET CT/GC Qx on Viper (Becton Dickinson).Methods: SCVS and FVU were tested spiked and unspiked with CT by AC2, cobas and Qx. Comparisons were made using a patient infected status. Pre-analytical interactions, reagent and sample preparation and loading were measured, plus hands-on-time for processing, post-analytical and maintenance. Time to results, automation times and total hands-on-time were calculated for 96 and 192 tests.Results: Probit calculations of levels of detection (LOD50) for SCVS and FVU were 8.1 and 7.9 for AC2, 5.9 and 7.1 for cobas and 6.7 and 5.9 for Qx, respectively. Samples were not inhibitory for amplification. CT prevalence was 9% (52/595). Sensitivities for SCVS were 98.1 for AC2, 84.6 for cobas and 92.3 for Qx, and for FVU 88.0, 81.1 and 75.5, respectively. Specificities ranged from 99.4 to 100. Total hands-on-time for 96 and 192 tests were 21 and 33 min (Panther), 40 and 98 min (cobas) and 105 and 137 min (Qx). Results were reported continuously for 96 and 192 tests at 3 hr 51 min on Panther compared to 2 batched reports at 4 hr 23 min and 6 hr 8 min on cobas and at 3 hr 31 min and 5 hr 8 min on Qx. The Qx required significantly more maintenance.
Conclusions: AC2 demonstrated higher analytical and clinical sensitivities. SCVS detected more infections than FVU. Panther consumed fewer reagents and required significantly less hands on time.