High rate of maintaining self-dependence and low complication rate with a new treatment algorithm for proximal humeral fractures in the elderly population.

Background: We modified our treatment algorithm for proximal humeral fractures in elderly patients in 2013 to a more conservative approach avoiding locking plates. This study assesses the impact of this change on patient self-dependence.

Methods: We carried out an observational comparative study including both retrospectively and prospectively collected data. For the former, 147 isolated proximal humeral fracture patients older than 65 years were treated between 2011 and 2013 at our hospital and included in a historical group. The revised treatment algorithm was applied in a similar non-concurrent, comparative patient group (n = 160) prospectively enrolled between 2015 and 2017. The primary outcome was any loss of self-dependence, with secondary outcomes including documentation of shoulder functional scores, quality of life, and adverse events.

Results: Historical and prospective patients had similar baseline characteristics. Nonoperative treatment was performed in 53 historical patients (36%) and 83 prospective patients (78%). Prospective patients were 1.6 times less likely to lose some level of self-dependence (risk ratio, 0.62; 95% confidence interval, 0.25-1.5; P = .292), and the local adverse event risk dropped from 12.2% to 5.7% (P = .078). Mean shoulder function and quality of life were similar between the 2 groups.

Conclusion: By applying our revised algorithm, a higher proportion of elderly patients maintained their premorbid level of self-dependence and returned to their previous social environment.