Prognosis and risk prediction of bone impaction grafting through femoral head-neck fenestration: a retrospective cohort study.

The bone impaction grafting through femoral head-neck fenestration was a favorable hip preservation procedure but without prognosis estimation. This study retrospectively reviewed 79 patients' clinical data (114 hips) with osteonecrosis of the femoral head (ONFH) who underwent this procedure from June 2009 to June 2019. By the end of June 2022, the median survival time of the hip was (74.13 ± 44.88) months, and the success rate of hip preservation was 68.42%. Lateral reserved angle (LPA) and combined reserved angle (CPA) had statistically significant differences ( < 0.001) both in univariate analysis and a multivariate logistic regression model. The multivariate logistic regression model of area under curve (AUC) area of the receiver operating characteristic (ROC) curve was 0.931(sensitivity = 95.00%, specificity = 88.40%, log-rank test:  < 0.01), and the calibration curve indicated good prediction accuracy. The ROC analysis and Cox proportional hazards regression model revealed that the cutoff point of LPA was 50.95° (sensitivity = 95.00%, specificity = 72.09%, log-rank test:  < 0.05) and the cutoff point of CPA was 90.51° (sensitivity = 90.00%, specificity = 90.70%, log-rank test:  < 0.05). A nomogram plot to predict the risk of failure (C-index = 0.873, 95% CI: 0.785 to 0.961) and nomograms for predicting the survival probability at 1, 2 or 3 years whose calibration curves showed excellent prediction accuracy were available for the clinician. Preserved angles (PAs) are valuable in the prediction of prognosis in surgical treatment. The bone impaction grafting through femoral head-neck fenestration can achieve better clinical efficacy, especially for patients with LPA >50.95° and CPA >90.51°.