Please login to view this media
- Talk
- 21/09/2021
- UK
Validation of the Radiographic Union Score for Humeral fractures (RUSHU): A Retrospective Study in an Independent Centre
Description
This presentation features William Fordyce, a core trainee at Derriford Hospital, discussing an external validation study focused on the Radiographic Union Score for Humeral fractures (RUSHU). Fordyce explains how the study aims to identify humeral fractures that are unlikely to unite early, potentially mitigating long-term patient morbidity and economic independence issues. RUSHU, proposed in 2019, is a scoring system based on x-rays taken 6 weeks post-injury, evaluating four cortices for potential healing. A score of 8 or higher suggests a high likelihood of union while a score below indicates greater risk of non-union, necessitating further surgical consideration.
The study sought to validate the scoring system within an external population, examining the reliability and generalizability of RUSHU through statistical methods and drawing comparisons with previous studies. It involved a cohort of 57 patients managed from 2005-2012, determining union vs non-union outcomes based on clinic letters and x-rays. Reviewing doctors, blinded to the outcomes, provided individual scores for analysis.
Results indicated 51 patients united while six did not, demonstrating good agreement among reviewers with significant differences in scores reported between the union and non-union groups. Although a good area under the ROC curve (0.87) suggests solid discrimination ability of the RUSHU score, the positive predictive value (PPV) raised some concerns, indicating lower efficacy in applying early surgery based on RUSHU scores compared to the original study.
Fordyce concludes that RUSHU can effectively reduce morbidity by allowing for quicker interventions for fractures at risk of non-union and expresses the need for further validations across diverse populations to enhance predictive performance. The discussion invites additional inquiries regarding study statistics and treatment methodologies employed within the study setting.