Sensitivity, specificity, and predictive values of the Limulus lysate assay for detection of exclusion of gonococcal cervicitis.

Abstract

The Limulus amoebocyte lysate (LAL) assay was evaluated for its ability to detect or exclude gonococcal cervicitis in two groups of women. The first (positive) group consisted of 100 untreated women who were referred to the venereal disease clinic with culture-proven gonococcal cervicitis. The second (negative) group consisted of 50 normal volunteers who were evaluated on two separate occasions. In the first group, Gram stains and repeat cervical cultures were 53 and 93% sensitive, respectively. In the second group, Gram stains and cultures were negative. For the LAL assay, ectocervical mucus was removed with a sponge, and a depyrogenated cotton-tipped swab was then used to collect endocervical specimens. The swab was placed in 1 ml of diluent (1:1 dilution), and serial twofold dilutions were made and tested for endotoxin by the LAL assay. Incubation was carried out at 37 degrees C for 30 min; positive or negative results were indicated by gelation or lack of gelation, respectively. At a dilution of 1:256, sensitivity and specificity of the LAL assay were 57 and 99%, respectively. The positive predictive values ranged from 36.5 to 97.4% for theoretical prevalence rates of 1 to 40%. At a dilution of 1:8, the sensitivity and specificity were 100 and 78%, respectively. At this dilution, the negative predictive value was 100% regardless of the prevalence rate. Thus, these preliminary results show that at the higher dilution, the LAL assay was comparable to Gram stain in diagnostic accuracy of gonococcal cervicitis, and if used as a screening test at the lower dilution, a negative LAL assay would exclude women without gonococcal cervicitis.