Meta-epidemiologic analysis indicates that MEDLINE searches are sufficient for diagnostic test accuracy systematic reviews

Abstract

Objectives: To investigate how the summary estimates in diagnostic test accuracy (DTA) systematic reviews are affected when searches are limited to MEDLINE.

Study design and setting: A systematic search was performed to identify DTA reviews that had conducted exhaustive searches and included a meta-analysis. Primary studies included in selected reviews were assessed to determine whether they were indexed on MEDLINE. The effect of omitting non-MEDLINE studies from meta-analyses was investigated by calculating the summary relative diagnostic odds ratio (RDORs) = DORMEDLINE only/DORall studies. We also calculated the summary difference in sensitivity and specificity between all studies and only MEDLINE-indexed studies.

Results: Ten reviews contributing 15 meta-analyses met inclusion criteria for quantitative analysis. The RDOR comparing MEDLINE-only studies with all studies was 1.04 (95% confidence interval [CI], 0.95, 1.15). Summary estimates of sensitivity and specificity remained almost unchanged (difference in sensitivity: -0.08%; 95% CI -1% to 1%; difference in specificity: -0.1%; 95% CI -0.8% to 1%).

Conclusion: Restricting to studies indexed on MEDLINE did not influence the summary estimates of the meta-analyses in our sample. In certain circumstances, for instance, when resources are limited, it may be appropriate to restrict searches to MEDLINE. However, the impact on individual reviews cannot be predicted.

Keywords: Comprehensive searches; Diagnostic test accuracy; Dissemination bias; MEDLINE; Meta-analyses; Systematic review.