Comparison of diagnostic performance between conventional and ultrasensitive rapid diagnostic tests for diagnosis of malaria: A systematic review and meta-analysis

Abstract

Background

Successful malaria treatment, control and elimination programs require accurate, affordable, and field-deployable diagnostic tests. A number of studies have directly compared diagnostic performance between the new ultrasensitive rapid diagnostic test (us-RDT) and conventional rapid diagnostic test (co-RDT) for detecting malaria. Thus, we undertook this review to directly compare pooled diagnostic performance of us-RDT and co-RDT for detection of malaria.

Methods

PubMed, Web of Science, Scopus, Embase, and ProQuest were searched from their inception until 31 January 2021 accompanied by forward and backward citations tracking. Two authors independently assessed the quality of included studies by RevMan5 software (using the QUADAS-2 checklist). Diagnostic accuracy estimates (sensitivity and specificity and others) were pooled using a random-effect model and 95% confidence interval (CI) in Stata 15 software.

Results

Fifteen studies with a total of 20,236 paired co-RDT and us-RDT tests were included in the meta-analysis. Molecular methods (15 studies) and immunoassay test (one study) were used as standard methods for comparison with co-RDT and us-RDT tests. The pooled sensitivity for co-RDT and us-RDT were 42% (95%CI: 25–62%) and 61% (95%CI: 47–73%), respectively, with specificity of 99% (95%CI: 98–100%) for co-RDT, and 99% (95%CI: 96–99%) for us-RDT. In asymptomatic individuals, the pooled sensitivity and specificity of co-RDT were 27% (95%CI: 8–58%) and 100% (95%CI: 97–100%), respectively, while us-RDT had a sensitivity of 50% (95%CI: 33–68%) and specificity of 98% (95%CI: 94–100%). In low transmission settings, pooled sensitivity for co-RDT was 36% (95%CI: 9 76%) and 62% (95%CI: 44 77%) for us RDT, while in high transmission areas, pooled sensitivity for co RDT and us RDT were 62% (95%CI: 39 80%) and 75% (95%CI: 57–87%), respectively.

Conclusion

The us-RDT test showed better performance than co-RDT test, and this characteristic is more evident in asymptomatic individuals and low transmission areas; nonetheless, additional studies integrating a range of climate, geography, and demographics are needed to reliably understand the potential of the us-RDT.