Study objective: To compare the ligase chain reaction (LCR) with culture for the detection of Neisseria gonorrhoeae (GC) and with culture and direct fluorescent antibody (DFA) for identification of Chlamydia trachomatis (CT) in cervical specimens from adolescent women.
Design: A prospective study of test performance.
Setting: Two urban, hospital-based adolescent clinics.
Participants: Adolescent women aged 12-22 yr undergoing pelvic examination for routine sexually transmitted disease (STD) screening or symptoms suggestive of an STD.
Main outcome measures: LCR results were considered to be true positives if confirmed by culture and/or DFA (CT only). Discrepant LCR results were confirmed by testing an alternative locus.
Results: With 538 subjects, LCR for CT had a sensitivity of 98.4% (61/62) and specificity of 96.4% (459/476) prior to resolution and a sensitivity of 98.6% (70/71) and specificity of 99.6% (459/461) after resolution. With 1225 subjects, LCR for GC had a sensitivity of 90.0% (54/60) and specificity of 99.4% (1158/1165) prior to resolution and a sensitivity of 90.6% (58/64) and specificity of 100% (1158/1158) after resolution. CT culture alone identified 80% of the true positives and DFA alone only identified 72%. GC culture alone identified 94% of the true positives.
Conclusions: LCR is an extremely sensitive and specific rapid test, utilizing a single swab and convenient room-temperature storage and transport of specimens. LCR testing of cervical specimens for CT in adolescent women is a better test for detecting CT infection than culture or DFA. LCR testing for cervical GC infection may provide an advantage over culture in circumstances in which optimal transport conditions and viability of the organism cannot be assured.