The Effect of Molecular Rapid Diagnostic Testing on Clinical Outcomes in Bloodstream Infections: A Systematic Review and Meta-analysis.

Abstract

Previous reports on molecular rapid diagnostic testing (mRDT) do not consistently demonstrate improved clinical outcomes in bloodstream infections (BSIs). This meta-analysis seeks to evaluate the impact of mRDT in improving clinical outcomes in BSIs. We searched PubMed, CINAHL, Web of Science, and EMBASE through May 2016 for BSI studies comparing clinical outcomes between mRDT and conventional microbiology methods. Thirty-one studies were included with 5920 patients. The mortality risk was significantly lower with mRDT than with conventional microbiology methods (odds ratio [OR], 0.66; 95% confidence interval [CI], .54-.80), yielding a number needed to treat of 20. The mortality risk was slightly lower with mRDT in studies with antimicrobial stewardship programs (ASPs) (OR, 0.64; 95% CI, .51-.79), and non-ASP studies failed to demonstrate a significant decrease in mortality risk (0.72; .46-1.12). Significant decreases in mortality risk were observed with both gram-positive (OR, 0.73; 95% CI, .55-.97) and gram-negative organisms (0.51; .33-.78) but not yeast (0.90; .49-1.67). Time to effective therapy decreased by a weighted mean difference of -5.03 hours (95% CI, -8.60 to -1.45 hours), and length of stay decreased by -2.48 days (-3.90 to -1.06 days). For BSIs, mRDT was associated with significant decreases in mortality risk in the presence of a ASP, but not in its absence. mRDT also decreased the time to effective therapy and the length of stay. mRDT should be considered as part of the standard of care in patients with BSIs.