Cholera diagnosis in human stool and detection in water: A systematic review and meta-analysis.

Background: Cholera continues to pose a problem for low-resource, fragile and humanitarian contexts. Evidence suggests that 2.86 million cholera cases and 95,000 deaths due to cholera are reported annually. Without quick and effective diagnosis and treatment, case-fatality may be 50%. In line with the priorities of the Global Task Force on Cholera Control, we undertook a systematic review and meta-analysis of diagnostic test accuracy and other test characteristics of current tests for cholera detection in stool and water.

Methods: We searched 11 bibliographic and grey literature databases. Data was extracted on test sensitivity, specificity and other product information. Meta-analyses of sensitivity and specificity were conducted for tests reported in three or more studies. Where fewer studies reported a test, estimates were summarised through narrative synthesis. Risk of Bias was assessed using QUADAS-2.

Results: Searches identified 6,637 records; 41 studies reporting on 28 tests were included. Twenty-two tests had both sensitivities and specificities reported above 95% by at least one study, but there was, overall, wide variation in reported diagnostic accuracy across studies. For the three tests where meta-analyses were possible the highest sensitivity meta-estimate was found in the Cholera Screen test (98.6%, CI: 94.7%-99.7%) and the highest specificity meta-estimate in the Crystal VC on enriched samples (98.3%, CI: 92.8%-99.6%). There was a general lack of evidence regarding field use of tests, but where presented this indicated trends for lower diagnostic accuracy in field settings, with lesser-trained staff, and without the additional process of sample enrichment. Where reported, mean test turnaround times ranged from over 50% to 130% longer than manufacturer's specification. Most studies had a low to unclear risk of bias.

Conclusions: Currently available Rapid Diagnostic Tests can potentially provide high diagnostic and detection capability for cholera. However, stronger evidence is required regarding the conditions required to secure these levels of accuracy in field use, particularly in low-resource settings.

Registration: PROSPERO (CRD42016048428).