摘要
Background: Multiple hemiarthroplasty (HA) approaches exist for femoral neck fracture (FNF) treatment. However, there is no consensus on the superiority of one approach for postoperative outcomes. This study assessed outcomes among randomized controlled trials (RCTs) comparing the posterior approach (PA), direct lateral approach (DLA), anterolateral approach (ALA), and the direct anterior approach (DAA) in HA for FNF. Methods: PubMed, Ovid/MEDLINE, Scopus, Cochrane Central Registry of Controlled Trials, and Cochrane Database of Systematic Reviews were queried in February 2024. A frequentist model network meta-analysis of eligible prospective RCTs compared outcomes among approaches using P-scores. Results: Of 1,481 retrieved studies, 11 RCTs totaling 1,513 FNF patients who underwent HA through a PA (n = 446; 29.5%), DLA (n = 481; 31.8%), ALA (n = 296; 19.6%), or DAA (n = 290; 19.2%) with median (interquartile range) follow-up of 6 (4.5-12.0) months were included for meta-analysis. DAA was associated with statistically increased operative duration (mean difference [MD], 1.89 minutes; 95% CI, 0.08 to 3.69 minutes; p < 0.001) compared with a PA, whereas an ALA was associated with statistically greater blood loss compared with a PA (MD, 5.81 mL; 95% CI, 4.11 to 7.50 mL; p < 0.001). There were no differences in the incidence of fracture, dislocation, infection, complications, revision, nor 1-year mortality among approaches. There were also no differences in hip nor pain score improvement at latest follow-up, although with high heterogeneity (I 2 = 85.7% and 94.2%, respectively). Conclusion: This study found no clinical difference in improvement in pain, function, nor in revision outcomes and operative characteristics among HA approaches for FNF, and each approach had a similar adverse event profile. Despite limited randomized evidence, these findings suggest comparable short-term efficacy and safety of all approaches. Level of Evidence: Therapeutic, Level I . See Instructions for Authors for a complete description of levels of evidence.