Comparison of four screening methods for group B streptococcus

Abstract

Objective To evaluate diagnostic performance of Todd-Hewitt (T-H) broth culture method, direct culture method, liquid chromogenic culture method, and loop-mediated isothermal amplification (LAMP) method for screening group B streptococcus (GBS) during late pregnancy. Methods In the retrospective study, the rectal vaginal secretions samples were collected from pregnant women at 35 to 37 weeks at the obstetrics clinic of Guangzhou Women and Children′s Medical Center affiliated to Guangzhou Medical University during October 2016 to April 2018. For the purposes of clinical evaluation, T-H broth culture was used as the standard reference method, and double-blind trials were used to evaluate diagnostic performance of direct culture method, liquid chromogenic culture method, and LAMP method for screening group B streptococcus during late pregnancy in three research stages. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), coincidence rate and Yoden index for each method were calculated. Also, the level of agreement between each method and T-H broth was assessed using the kappa (k) coefficient. Results A total of 969 specimens were detected by the T-H enrichment culture method, and 90 were positive (9.3%). The sensitivities from high to low were LAMP method [100% (25/25)], direct culture method [81.5% (22/27), 95%CI:65.8%-97.1%], and liquid color culture method [71.1% (27/38), 95%CI:55.9%-86.2%]. Specificities were direct culture method [100% (282/282)], liquid color culture method [98.1% (455/464), 95%CI:96.8%-99.3%], and LAMP method [94.0% (125/133), 95%CI: 89.9%-98.1%]. The coincidence rates were direct culture method [98.4% (22+282)/309], liquid color culture method [96.0% (27+455)/502], and LAMP method [94.9% (25+125)/158]. The Kappa values of the direct culture method (0.889), LAMP method (0.832) and the enrichment culture method were all ≥0.75, and that of the liquid color culture method was 0.708. The false negative rate of direct culture method was 18.5% (5/27), and no false negative case by LAMP method, but its false positive rate was 6.0% (8/133). The false negative rate and false positive rate of liquid color culture method were 28.9% (11/38) and 1.9% (9/464), respectively. Conclusions Of the three screening methods compared in this study, only the LAMP method has the advantages in sensitivity, specificity, and coincidence rate compared with T-H enriched culture method, while the others have a certain degree of false negatives rate. The clinical laboratory can introduce these methods based on laboratory facilities and staffing, or refer to the European and American guidelines and combine the recommended antenatal GBS screening method with intrapartum nucleic acid amplification tests to best meet the clinical demands. Key words: Pregnancy trimester, third; Streptococcus agalactiae; Bacteriological techniques; Culture media; Nucleic acid amplification techniques; Sensitivity and specificity