Predicting the outcome of patients with proximal humeral fracture by radiographic findings : a retrospective study

Callaghan, M and Ball, E ORCID: 2014, 'Predicting the outcome of patients with proximal humeral fracture by radiographic findings : a retrospective study' , Health & Quality of Life Outcomes . (Submitted)

PDF - Accepted Version
Download (232kB) | Preview
Official URL:


Background: A retrospective study was undertaken to establish if the outcome of patients with proximal humeral fracture could be predicted by radiographic findings. 12 patients were reviewed with a proximal humeral fracture and were treated conservatively with sling immobilisation, followed by physiotherapy. After each treatment was completed, patient’s radiographs, taken immediately after the initial trauma, were reviewed and the fracture was classified according to Neer and arbeitsgemeinschaft orthopaedie (AO). There were two crucial time points. Methods: Reviewer one, blinded to the patients’ final functional outcome, attempted to forecast patient functionality by placing each patient into one of 3 broad groups described as ‘good’, ‘moderate’ or ‘poor’ based on the radiographs. Six months after finishing physiotherapy all patients were reassessed functionally by another researcher (reviewer 2), blinded to radiographic classification and prediction. Multiple regression analysis was used to measure outcomes at both interval points. Results: Multiple regression analysis revealed a significant correlation (R2 0.613, p = 0.014) between the radiographic predicted outcome and the immediate post treatment outcome when adjusting for age. At 6 months review, there was a significant correlation between predicted radiographic outcome and functional outcome when adjusted for patients’ age (R2 0.765 P = 0.001); patients remained in their predicted groups. Six months after physiotherapy only one patient reported being pain free. Poisson linear regression analysis revealed that compared to the patients in the ‘poor’ and ‘moderate’ groups those in the ‘good’ group had 3 treatment sessions wasted (P < 0.001). Those in the ‘poor’ group accounted for 50 out of 76 treatments after their progress had plateaued. Conclusions: This study indicates that knowledge of the types of proximal humeral fracture classification and accurate radiological diagnosis may help planning conservative treatment realistically and reduce the number of unnecessary physiotherapy treatments.

Item Type: Article
Schools: Schools > School of Health and Society > Centre for Applied Research in Health, Welfare and Policy
Journal or Publication Title: Health & Quality of Life Outcomes
Publisher: BioMed Central
Refereed: Yes
ISSN: 1477-7525
Funders: Non funded research
Depositing User: S Rafiq
Date Deposited: 13 Nov 2014 17:39
Last Modified: 15 Feb 2022 18:49

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)


Downloads per month over past year