Chronic Acromioclavicular Joint and Subacromial Pain: Reliability and Diagnostic Accuracy of Clinical Tests

Abstract

Shoulder pathologies are the third most common musculoskeletal complaint globally. In New Zealand they are responsible for substantial disability-adjusted life years and healthcare expenditure. Chronic shoulder pathologies are often misdiagnosed due to the poor validity of tests and an over-reliance on imaging. A correct diagnosis of shoulder pain is integral to forming a prognosis and management plan.

The aim of this thesis was to investigate the validity of commonly used clinical tests for identifying acromioclavicular joint (ACJ) and subacromial pain in patients with chronic shoulder pathologies.

A systematic review was carried out to investigate the reliability and diagnostic accuracy of orthopaedic special tests (OSTs) for ACJ and subacromial pain. Estimates of inter-rater reliability were poor to moderate for the majority of tests and the outcomes between the studies varied considerably. The diagnostic accuracy of OSTs was also poor, with no test consistently demonstrating adequate diagnostic accuracy to be a valid tool in clinical practice. When combined, clusters of clinical tests, other than OSTs, have shown adequate diagnostic accuracy to influence the probability of an ACJ or subacromial diagnosis.

A prospective inter-rater reliability study was completed with 20 participants presenting with shoulder pain. A standardised physical assessment was conducted twice for each participant by two blinded clinicians. The test results of each clinician were compared to calculate percentage agreement, prevalence-adjusted and bias-adjusted kappa (PABAK) and kappa scores. The outcomes demonstrated poor to moderate agreement for most tests (≤0.60 PABAK). The exception to this, with high inter-rater reliability, was observed muscle wasting, bilateral ACJ deformity, pseudoparalysis and the horizontal adduction with external rotation test (Add/ER).

A diagnostic accuracy study was conducted with 38 participants with ACJ or subacromial pain. Each participant underwent a standardised interview and physical assessment followed by anaesthetic injection into the subacromial space +/- the ACJ. A positive anaesthetic response threshold was set at ≥65% improvement on the numeric pain rating scale. None of the clinical variables reached statistical significance for diagnostic accuracy, suggesting that they cannot be used as stand-alone tools to either rule in or out ACJ or subacromial pain in this cohort. A multivariate analysis identified a combination of 10 positive and negative variables that have the potential to be clinically useful to predict a subacromial diagnosis. These were: difficulty with overhead tasks, a strain injury onset, lowest pain ≥5/10, presence of muscle wasting, onset of pain from a repetitive activity, worst pain ≥8/10, the primary pain site over or above the clavicle, painful Add/ER, painful passive internal rotation and painful resisted flexion at 10 degrees. The area under the receiver operating characteristic curve for these variables was 0.73, with a misclassification error of 35%.

This thesis’s findings support the current research that clinical variables as stand-alone tests are not able to accurately identify an ACJ or subacromial diagnosis. It is recommended that clinicians place more emphasis on a combination of findings from the wider clinical assessment. Future research should focus on examining the validity of combinations of tests with anaesthetic blocks as a reference test.