Radiologic restoration inaccuracy increases postoperative dislocation in primary total hip arthroplasty: a retrospective study with propensity score matching

Linbo Peng,Yi Zeng,Yuangang Wu,Haibo Si,Fuxing Pei,Bin Shen
DOI: https://doi.org/10.1007/s00402-021-04263-7
Abstract:Background: Dislocation is a complicated process and associated with multivariate etiology and risk factors. The purpose of this study was to verify if radiologic restoration inaccuracy increases postoperative dislocation in primary total hip arthroplasty (THA) individually or synergistically. Methods: From 2017 to 2020, we identified 76 (1.9%) patients who suffered postoperative dislocations from 3926 THAs in our institution. After excluded patients with previously proven patient-related and implant-related risk factors, the remaining patients were used to match a 1:1 control patients who were without dislocation. The cup position (inclination and anteversion angles), hip offset (HO), leg length discrepancy (LLD), and abductor lever arm (ALA) were analyzed by univariate analyses and multivariate logistic regressions. Results: Measurements on radiographs showed excellent interobserver agreement (intraclass correlation coefficient (ICC) 0.922-0.952) and intraobserver agreements (ICC 0.933-0.967). HO restoration inaccuracy (without ± 5 mm) was associated with higher dislocation risk (OR 4.241 95% CI 1.440-12.492, P = 0.009). The restoration inaccuracy of the cup position, LLD, or ALA could not increase the dislocation risk individually. When combining the radiologic restoration inaccuracy factors, the HO + LLD restoration inaccuracy and HO + ALA restoration inaccuracy increased the odds of postoperative dislocation (OR 12.056, 95% CI 1.409-103.127, P = 0.023; OR 4.770, 95% CI 1.336-17.028, P = 0.016, respectively). Combining the four risk factors of cup safe zone, HO, LLD, and ALA, patients with 3 or 4 radiologic restoration inaccuracy factors showed a higher risk of dislocation (OR 13.500, 95% CI 1.340-135.983, P = 0.027). Conclusion: Hip offset (HO) restoration inaccuracy increased the risk of dislocation following primary THA. Multivariate radiologic restoration accuracy is critical for the prevention of dislocation and needs to be valued by surgeons.
What problem does this paper attempt to address?