Shoulder extension impairment with residual neonatal brachial plexus injury.

[BACKGROUND] Impairment of both shoulder extension and behind-the-back function are common in patients with residual neonatal brachial plexus injury (NBPI), but have scarcely been studied or reported in the literature. Behind-the-back function is classically evaluated using the hand-to-spine task used for the Mallet score. Angular measurements of shoulder extension with residual NBPI have generally been studied utilizing kinematic motion laboratories. To date, no validated clinical examination method for this has been described.

[METHODS] Intraobserver and interobserver reliability analyses of 2 shoulder extension angles-passive glenohumeral extension (PGE) and active shoulder extension (ASE)-were performed. Afterwards, a retrospective clinical study was conducted on prospectively collected data on 245 children with residual BPI treated from January 2019 through August 2022. Demographic characteristics, level of palsy, previous surgical procedures, modified Mallet score, and bilateral PGE and ASE data were analyzed.

[RESULTS] All inter- and intraobserver agreements were excellent, ranging from 0.82 to 0.86. The median patient age was 8.1 years (3.5-21). Among the 245 children, 57.6% had Erb's palsy, 28.6% extended Erb's palsy, and 13.9% global palsy. One hundred sixty-eight (66%) of the children could not touch their lumbar spine, among whom 26.2% (n = 44) had to swing the arm to reach it. Both the degrees of ASE and PGE achieved correlation significantly with the hand-to-spine score, the ASE strongly (r = 0.705) and the PGE weakly (r = 0.372) (both P < .0001). Significant correlations also were found between lesion level and the hand-to-spine Mallet score (r = -0.339; P < .0001) and ASE (r = -0.299; P < .0001), and between patient age and the PGE (P = .0416, r = -0.130). A statistically significant decrease in PGE and incapacity to reach the spine were found in patients who underwent glenohumeral reduction, shoulder tendon transfer, or humeral osteotomy, relative to those who had microsurgery or no surgery. Receiver operating curves showed that, for both PGE and ASE, the minimum extension angle required to successfully perform the hand-to-spine task was 10°, with sensitivity levels of 69.9 and 82.2, and specificity levels of 69.5 and 87.8 (both P < .0001), respectively.

[CONCLUSIONS] Glenohumeral flexion contracture and lost ASE are extremely common in children with residual NBPI. Both the PGE and ASE angles can be measured reliably with a clinical exam, with at least 10° of PGE and ASE necessary to perform the hand-to-spine Mallet task.