Table 2.
LAMP | dPCR | qPCR | |
---|---|---|---|
Sensitivity | |||
Reitsma | 83·3% (76·9–88·2) | 94·1% (88·9–96·9) | 92·7% (88·3–95·6) |
Latent | 86·2% (20·7–99·9) | 95·8% (54·9–100·0) | 93·4% (60·9–99·9) |
Specificity | |||
Reitsma | 96·3% (93·8–97·8) | 78·5% (57·4–90·8) | 92·9% (87·2–96·2) |
Latent | 94·3% (49·1–100·0) | 73·8% (0·9–100·0) | 93·1% (47·1–100·0) |
Data are % (95% CI). Sensitivities and specificities for each molecular test were estimated using Reitsma's bivariate random-effects model (Reitsma) and Bayesian latent class bivariate model (latent). LAMP=loop-mediated isothermal amplification. dPCR=digital PCR. qPCR=quantitative PCR.