Comparative efficacy and safety of locking intramedullary nails versus locking plates for displaced proximal humeral fractures

Aims The optimal fixation method for displaced proximal humeral fractures (PHFs) remains debated. Locking intramedullary nails (INs) are considered theoretically less invasive than locking plates (LPs), yet randomized evidence comparing their safety and functional outcomes is inconsistent. This study compares IN and LP using trial sequential analysis (TSA) and complexity-stratified subgroups to determine whether current randomized evidence is sufficient to draw firm conclusions. Methods A systematic search of PubMed, Scopus, and Web of Science was conducted from database inception through to January 2026. Randomized controlled trials (RCTs) comparing IN and LP in adults with displaced PHFs were included. Primary outcomes were patient-reported outcome measures (PROMs) across early (three months), mid-term, and long-term intervals, alongside complication rates. TSA was applied to control for random errors and estimate the required information sizes. Results Six RCTs involving 382 participants were included. At early follow-up, no significant differences in PROMs or pain scores were observed, refuting the expected immediate muscle-sparing advantage of IN. At 12 months, no significant differences were observed between IN and LP in Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire score (p = 0.42), Constant-Murley Score (p = 0.55), or American Shoulder and Elbow Surgeons score (p = 0.34). At long-term follow-up ≥ 24 months, LP demonstrated statistically lower DASH scores (p = 0.03); however, the difference did not exceed the minimal clinically important difference. Crucially, complexity-stratified analyses revealed a directional signal favouring LP specifically in three- and four-part fractures. TSA showed conclusive evidence that IN does not reduce reoperation rates (risk ratio 1.45; p = 0.58) or screw-related complications compared with LP. Conclusion While pooled evidence demonstrates overarching equivalence, implant selection should be guided by fracture complexity. For two-part surgical neck fractures, either implant is highly appropriate. For displaced three- and four-part fractures, a directional signal from stratified analyses, and the clinical burden of reoperations suggest a potential advantage for LPs. This signal should inform clinical decision-making pending adequately powered, fracture-stratified RCTs. Cite this article: Bone Jt Open 2026;7(11):1419–1428.