Six studies (17 test comparisons) were included in the review and meta-analysis. Quality scores for the six included studies ranged from 25 to 32. The authors stated that they selected tests for analysis for which a small number of studies were distributed as uniformly as possible across the diagnostic tests. They also stated that studies distributed as uniformly as possible across methodological score categories (25 to 27, 28, 30, and 32) when cross-tabulated with diagnostic test were selected. It appeared that some studies that met the stated inclusion criteria were excluded from the review; five tests from six studies were excluded from the review due to the small number of studies available for each test (the authors stated that most had fewer than three).
Estimated sROC curves indicated that the anterior slide test was statistically inferior to the active compression, crank and Speed. No significant differences were found for ROC curves among the active compression, crank and Speed tests. The active compression test was found to have the best ROC curve. Methodological quality and study had no significant effects in the model.
Active compression test: Reported sensitivity estimates ranged from 0.470 to 0.778 and specificity estimates ranged from 0.111 to 0.730 (six studies).
Anterior slide test: Reported sensitivity estimates ranged from 0.05 to 0.100 and specificity estimates ranged from 0.815 to 0.930 (three studies).
Crank test: Reported sensitivity estimates ranged from 0.125 to 0.580 and specificity estimates ranged from 0.560 to 0.826 (five studies).
Speed test: Reported sensitivity estimates ranged from 0.040 to 0.478 and specificity estimates ranged from 0.674 to 0.990. (three studies).