Summary of findings 2. Conventional polymerase chain reaction (PCR) sensitivity analysis, excluding high risk of bias (reference standard domain).

Conventional polymerase chain reaction (PCR) sensitivity analysis, excluding studies at high risk of bias (reference standard domain)
Population: people suspected of leptospirosis in different stages of disease (early to late), excluding those with solely ocular problems or aseptic meningitis
Setting: worldwide, primary to tertiary care facilities, outbreak as well as non‐outbreak settings
Index test: conventional PCR on blood samples (whole blood, serum), all inhouse tests
Reference standard: MAT on serum alone (2‐gate studies), or MAT on serum alongside culturing, or MAT on serum alongside IgM ELISA, or MAT on serum alongside culturing and IgM ELISA (single‐gate studies)
Number of cases/non‐cases (studies): 538/487 (9)	Pooled sensitivity: 87% (95% CI 44% to 98%)	Pooled specificity: 97% (95% CI 60% to 100%)	Consequences in a cohort of 1000
Prevalence:	Positive post‐test probability:	Negative post‐test probability:	Missed diseased	Falsely diagnosed
32.5% (median of all studies)	94% (95% CI 41% to 100%)	94% (95% CI 70% to 99%)	42 (95% CI 5 to 183)	17 (95% CI 1 to 272)
9.7%	78% (95% CI 13% to 99%)	99% (95% CI 91% to 100%)	13 (95% CI 2 to 55)	23 (95% CI 1 to 363)
Positive likelihood ratio: 33.86 (95% CI 1.59 to 719.39)		Negative likelihood ratio: 0.13 (95% 0.02 to 0.85)
Quality of evidence: none of the studies scored 'low risk of bias' on all domains. Risk of bias for the 'reference standard' domain was unclear to low. Risk of bias for the 'patient selection' domain was unclear to high.

CI: confidence intervals; IgM ELISA: immunoglobulin M enzyme‐linked immunosorbent assay; MAT: microscopic agglutination test; NA: not applicable.