Neer, Hawkins, and Yocum tests are used in clinical practice to detect sub-acromial impingement. Their reliability has been few investigated. The goal of the study was to investigate reliability of Neer, Hawkins and Yocum tests. We also assessed an original manoeuver.

Patients with shoulder pain due to degenerative rotator cuff disease were prospectively included. They were twice assessed at one week interval by two independent observers. Results of Neer, Hawkins and Yocum tests were recorded as positive or not. Results of an original manoeuver (OM) including elevation in medial rotation followed by lateral one were also. Intra- and inter-observer reliability was assessed. Criteria of reliability were the percentage of agreement and the kappa coefficient of concordance with 95% confidence interval (CI). Kappa coefficient indicated moderate concordance if>0.4, good>0.61.

Thirty-four Patients were included: age 60±11, ratio F/M 26/8, pain duration 27±67 months. For intra-observer reliability, agreement was 80–88% with Neer, Hawkins and Yocum tests. It was 80% with OM. Using kappa coefficient, intra-observer reliability was poor with Neer test (0.34 [−0.1;1]) and moderate with Hawkins (0.56 [0;0.9]), Yocum (0.48 [0;0.8]) and OM (0.6 [0.2;0.9]). For inter-observer reliability, agreement was 73–88% with Neer, Hawkins and Yocum tests. It was 79% with OM. Using kappa coefficient, inter-observer reliability was poor with Yocum test (0.35 [0;0.7]), moderate with Hawkins (0.54 [0.2;0.8]) and OM (0.58 [0.2;0.8]), and good with Neer (0.64 [0.2;0.9]).

Neer test was impaired by poor intra-observer reliability and Yocum test was impaired by poor inter-observer reliability. Hawkins test and OM had a balanced profile of moderate reliability. There was a trend to higher values with OM. We underlined there an interesting profile of reliability of an original clinical manoeuver to detect sub-acromial impingement.