Forty-five studies were included in the review. Twenty two of these met the authors' criteria for high quality, but only two were considered to have adequate sample size (number of participants ranged from 12 to 585).
Impingement: Six studies assessed OSTs for impingement. Meta-analysis gave pooled sensitivity and specificity values for the Neer test of 0.79 (95% CI 0.75 to 0.82; four studies) for sensitivity and 0.53 (95% CI 0.48 to 0.58; four studies) for specificity. Pooled estimate of sensitivity for Hawkins-Kennedy test was 0.79 (95% CI 0.75 to 0.82; four studies) and the pooled estimate of specificity was 0.59 (95% CI 0.53 to 0.64; four studies). For both tests, the pooled diagnostic odds ratio had confidence intervals that crossed one. For other tests of impingement, reported sensitivities ranged from 4% to 88% and reported specificities ranged from 28% to 100%.
Rotator cuff integrity: Fifteen studies assessed 20 OSTs for rotator cuff integrity. None of the 10 OSTs for rotator cuff pathology that were examined in more than one study had consistent diagnostic performance. Reported sensitivity values ranged from 10% to 98% and reported specificity values ranged from 9% to 98%. Data for tests not included in meta-analyses suggested that hornblowers’s sign and external rotation lag sign may be useful for tears of the rotator cuff.
Glenoid labrum integrity: Twenty one studies assessed OSTs for labral pathologies. Meta-analysis gave a pooled estimate for sensitivity of Speed test for superior labral anterior to posterior (SLAP) lesions of 0.32 (95% CI 0.24 to 0.42; five studies) and a pooled estimate of specificity of 0.61 (95% CI 0.54 to 0.68; five studies). Pooled diagnostic odds ratio had confidence intervals that crossed one. Other OSTs for labral pathologies reported sensitivities that ranged from 5% to 99% and specificities that ranged from 11% to 98%. Data for tests not included in meta-analyses suggested that biceps load II may be useful for SLAP lesions.
Five studies assessed nine OSTs for various aspects of instability. Reported sensitivity values ranged from 28% to 100% and reported specificity values ranged from 44% to 99%. Data for tests not included in meta-analyses suggested that apprehension, relocation and anterior release may be useful tests for anterior instability.
Acromioclavicular joint pathology: Three studies assessed 10 OSTs for acromioclavicular joint pathology. Reported sensitivity values ranged from 16% to 96% and reported specificity values ranged from 10% to 96%.