High Interrater Reliability for Arm Span and Ulnar Length Measurements

Abstract

INTRODUCTION: Pulmonary function tests (PFTs) are normalized based on patient height. Arm span has been routinely collected as a surrogate measure for height, as severe scoliosis can result in reduced standing height. However, upper extremity contractures and limited patient cooperation may render accurate and reproducible measurement of arm span and sitting height challenging. Ulnar length has also been used as a measure of stature, and may be more reproducible in the early onset scoliosis patient population. We sought to determine the interrater reliability of ulnar length compared to arm span and sitting height among early onset scoliosis (EOS) patients. METHODS: Ninety-seven EOS patients enrolled in a multicenter registry underwent measurement of arm span and ulnar length by two individuals on the same occasion. Height, weight, sitting height, presence of instrumentation, diagnosis, and radiographic measures were also recorded. Mean age was 10 (range, 1.6 to 20 years). Mean height was 135 cm, and mean weight was 37 kg. In total, 64% of patients had spinal instrumentation. Some 32% of patients had idiopathic scoliosis, 32% had congenital deformity, 9% neuromuscular, and remainder syndromic/other. Interclass correlation coefficient was assessed to determine the reproducibility of arm span and ulnar length measurements, with 1.0 representing perfect agreement. RESULTS: There was high interrater reliability with an ICC approaching 1.0 for arm span, ulnar length, and sitting height. Excluding patients over age 10, ICC dropped, especially for ulnar length, but still remained in the acceptable range (Table). Arm span correlated well with height (R square 0.94, p<0.0001) better than ulnar length (R square 0.89, p<0.0001) (Figure). T1-T12 and T1-S1 length correlated poorly with standing height, which is reflective of underlying scoliosis (R square 0.53, 0.29, respectively). DISCUSSION AND CONCLUSION: Arm span has good interrater reliability and is reflective of standing height. Prospective EOS registries should continue to collect arm span in favor of ulnar length for outcome measures and analysis of PFT results