REAPPRAISING THE VALUE OF URINE LEUKOCYTE ESTERASE TESTING IN THE AGEOF NUCLEIC-ACID AMPLIFICATION

Background: The leukocyte esterase (LE) test has a limited role in the determination of empiric therapy for male patients screened for ureth ritis because of its poor positive predictive value in low (<5%) preva lence settings. The recent advent of nucleic acid amplification testin g of first-void urine (FVU) has dramatically increased the ease with w hich widespread screening for Chlamydia trachomatis and Neisseria gono rrhoeae can be performed, but the costs of such testing may be prohibi tive. The LE test may therefore have a role in management of urethriti s because of its high negative predictive value. Objectives: To determ ine the sensitivity, specificity, and positive and negative predictive value of LE testing for the diagnosis of N. gonorrhoeae and C. tracho matis in male FVU specimens in a low-prevalence urban setting using a commercial polymerase chain reaction (PCR) as the gold standard. Metho ds: Data were obtained on men presenting to an urban sexually transmit ted disease clinic over a 16-month period, Patients were included if a n FVU had been tested for the presence of LE using a rapid dipstick, r ead by an automated urine analyzer, and the sample (either an FVU or u rethral swab) had then been processed for the detection of N. gonorrho eae and C. trachomatis by PCR. Results: Of 301 assessable patients, th ere were 14 cases of gonorrhoea, 21 cases of chlamydia, and 1 case of dual infection detected by PCR, Most men (245/301; 81.4%) were asympto matic, of whom 12 of 245 (4.9%) had an infection detected compared wit h 24 of 56 (42.9%) in the symptomatic men (P < 0.001), Using a ''great er than or equal to trace'' cutoff, the overall value for the sensitiv ity of the LE test was 77.8% (95% confidence interval, 60.4-89.3), spe cificity 80.8% (75.4-85.2), positive predictive value 35.4% (25.2-47.1 ), and negative predictive value 96.4% (92.8-98.3), Conclusions: The n egative predictive value of the LE test may be of use in determining w hich patients should proceed to specific diagnosis by nucleic amplific ation methods (e.g., PCR or ligase chain reaction). By limiting testin g to patients with positive LE results, costs savings mag be made, ena bling the technology to be used in a wider community setting. The valu e of the LE test in higher prevalence populations with access to nucle ic amplification testing remains to be established.