Background: Evaluation and reporting of the heterogeneity in diagnostic-test accuracy (DTA) reviews is not standardised. A recent evaluation of diagnostic reviews identified through EMBASE and MEDLINE revealed serious flaws in heterogeneity reporting. The review proposes a guidance on testing for and quantifying variability in reviews of diagnostic test accuracy following 5 steps: 1: visualise total variability; 2: judge whether there is more variability in sensitivity and specificity than can be expected due to chance (sampling error) alone; 3: measure the total between-study variability; 4: attribute some of the between-study variability to the threshold effect and 5: explore what study features might explain some of the variability.
Objectives: To describe: 1) the methods used to evaluate heterogeneity in DTA review protocols and full reviews and to assess the degree of completeness in heterogeneity analysis; 2) how the results of between-study variability were used to guide the analytical approach; and, 3) how heterogeneity was interpreted regarding implication for clinical practice.
Methods: We will include all DTA review protocols and DTA Cochrane reviews published in the Cochrane Library from its inception to March 2017. To assess the degree of completeness in heterogeneity analysis we will use the guidance propose by Naaktgeboren et al. We will classify the results in three categories: i. complete heterogeneity evaluation: completed the 5 steps; ii. Incomplete evaluation: completed 3 steps; insufficient evaluation: less than 3 steps.
Results: We will provide a detailed description and discussion of the review results.