Comparison between arthroplasty and non-operative treatment for proximal humeral fractures: a systematic review and meta-analysis

Background: The clinical efficacy of reverse shoulder arthroplasty (RSA), hemiarthroplasty (HA), and non-surgical management in the treatment of proximal humeral fractures (PHFs) is inconclusive. This systematic review and meta-analysis compared the clinical outcomes of arthroplasty and non-surgical management of PHFs. Methods: The databases of PubMed, Embase, Web of Science, and Cochrane Library were searched on 5 May 2023 for studies comparing arthroplasty and non-surgical treatment of PHFs. Both randomized controlled trials (RCTs) and non-randomized controlled trials (nRCTs), were included. Standard methodological quality assessments were conducted for both types of studies. The primary outcome was the Constant-Murley Score (CMS) after surgical or non-surgical treatment. Secondary study outcomes included the visual analog scale (VAS), range of motion, and complications. All functional scores and complications were subjected to subgroup and sensitivity analyses. Results: A total of four RCTs and six nRCTs were included in this study, which provided 508 patients in total for meta-analysis: 238 treated with arthroplasty and 270 treated non-surgically, of which 83 were treated with HA and 155 with RSA. All relevant information was collected, including functional scores, VAS, range of motion, and complications. The study found no significant difference in functional outcomes (mean difference, 2.82; 95% confidence interval, -0.49 to 6.14; P = 0.10; I-2 = 77%) and complications (mean difference, 1.08; 95% confidence interval, 0.51-2.25; P = 0.85; I-2 = 47%) between arthroplasty and non-surgical treatment. Both RCTs and nRCTs showed the same results. However, VAS scores were significantly lower in surgical treatment compared to non-surgical treatment. Subgroup and sensitivity analyses showed that RSA could achieve better functional scores than non-surgical treatment (mean difference, 6.00; 95% confidence interval, 1.97-10.03; P = 0.004; I-2 = 0%), while the results for HA were not significant (P > 0.05). Conclusion: There were no significant differences in complications between arthroplasty and non-surgical treatment for PHFs. RSA could achieve better functional results than non-surgical treatment, while HA could only achieve better forward flexion.