Ultrasound Screening for Posterior Shoulder Dislocation in Infants with Persistent Brachial Plexus Birth Palsy

Background: Early detection of posterior shoulder dislocation in infants with brachial plexus birth palsy (BPBP) is essential, but it may be difficult to accomplish with physical examination alone. The aim of this study was to determine the prevalence of shoulder dislocation in patients with BPBP using ultrasound and to identify which physical examination measurements correlated most with dislocation in these patients. Methods: This study was a retrospective review of data obtained in an ultrasound screening program of infants with BPBP born from January 2011 to April 2014. Physical examination included the use of the Active Movement Scale (AMS) and measurement of passive external rotation of the shoulder. Ultrasound measurements included PHHD (percentage of the humeral head displaced posterior to the axis of the scapula) and the alpha angle (intersection of the posterior scapular margin with a line tangential to the humeral head through the glenoid). Shoulder dislocation was defined as both a PHHD of >0.5 and an alpha angle of >30 degrees. Results: Of sixty-six infants who had undergone a total of 118 ultrasound examinations (mean, 1.8; range, 1 to 5), 19 (29%) demonstrated shoulder dislocation with the shoulder positioned in internal rotation; the dislocation was first detected between 2.1 and 10.5 months of age. Infants with a dislocated shoulder demonstrated significantly less mean passive external rotation in adduction (mean, 45.8 degrees versus 71.4 degrees, p < 0.001) and a greater difference between internal rotation and external rotation AMS scores (mean, 5.5-point versus 3.3-point difference, p < 0.001) than those without shoulder dislocation. Passive external rotation in adduction was a better measure for discriminating between dislocation and no dislocation (area under receiver operating characteristic curve [AUC] = 0.89) than was the difference between internal and external rotation AMS scores (AUC = 0.73). A cutoff of 60 degrees of passive external rotation in adduction (<= 60 degrees versus degrees >60) yielded a sensitivity of 94% and a specificity of 69%. Conclusions: Shoulder dislocation is common in infants with BPBP; 29% of the infants presenting to our tertiary care center had a dislocation during their first year of life. Ultrasound shoulder screening is appropriate for infants with BPBP. If passive external rotation in adduction is used to determine which infants should undergo ultrasound, <= 60 degrees should be utilized as the criterion to achieve appropriate sensitivity.