275 Validation of The Radiographic Union Score for Humeral Fractures (RUSHU) For Prediction of Humeral Shaft Nonunion

Abstract

Abstract Aim To validate the use of RUSHu score in prediction of humerus non union. Method All patients having radiographs of humerus performed between Jan 2016 to December 2018 were assessed based on inclusion and exclusion criteria. The RUSHu scoring system as published was used to score each 6-week radiograph, separately by 2 blinded observers. 6 months was used as end point to assess outcome. Cohort of 188 observations were used to assess utility of scoring system to predict non union. Results 94 suitable fractures were identified. Union rate of 72.3% was observed. Mean score in union group was 9.6, 6.4 for non-unions. There was substantial inter-observer reliability with an ICC of 0.73. Rate of union progressively increases with increasing RUSHu scores. ROC curve analysis identifies 8 as most suitable for use as threshold. Area under the curve is high (0.9) Conclusions A low RUSH score at 6 weeks is a reliable predictor of non union down the line. If a score 7 or lower is observed, it should trigger a discussion with the patient and review of correctable factors contributing to development of non union. Consideration of surgical fixation should be made at this stage if instability is felt to be a major contributing cause. A patient with score of 8 or higher is more likely to go on to union. Routine use of RUSHu score can aid in clinical decision making and introduce an element of objectivity in clinical assessment. It has potential to prompt earlier intervention and reduce morbidity duration.