Establishment and validation of an early warning model for the failure of PFNA internal fixation in patients with intertrochanteric fracture of femur accompanied by osteoporosis based on clinical data

XU Xiaodan; LU Hanli; LING Yaoguang

doi:10.3969/j.issn.1002-3429.2025.18.013

Abstract：Objective To establish and validate an early warning model for the failure of proximal femoral nail antirotation(PFNA)internal fixation in patients with intertrochanteric fracture of femur(IFF)accompanied by osteoporosis based on clinical data.Methods Sixty patients with IFF accompanied by osteoporosis who underwent PFNA and experienced internal fixation failure(failure group)and sixty patients who did not experience failure(non-failure group)admitted from January 2019 to June 2023 were selected.Additionally,another 60 IFF patients with osteoporosis in different periods of our center were selected as the external validation data set,including 13 patients with internal fixation failure and 47 patients without failure.The general information of patients in the failure group and non-failure group were compared,and univariate and multivariate LASSO-Logistic regression were used to analyze the prognostic factors of patients.Based on the selected predictive variables,the early warning model of PFNA internal fixation failure was constructed by R software and verified.Results The Charlson Comorbidity Index(CCI)score,the proportion of patients with poor spiral blade position,the proportion of patients with lateral wall type Ⅲ,the proportion of patients with unstable internal fixation based on Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association(AO/OTA)classification,the proportion of patients with severe Singh index,the tip-apex distance,and the proportion of patients with poor fracture reduction quality in the failure group were higher than those in the non-failure group,while the thickness of lateral wall was lower than that in the non-failure group(P<0.05,P<0.01).LASSO-Logistic regression analysis showed that elevated CCI score,lateral wall type Ⅲ,unstable AO/OTA classification,severe Singh index,and poor quality of fracture reduction were independent risk factors for internal fixation failure,and increased lateral wall thickness was an independent protective factor for internal fixation failure(P<0.05).The consistency index of the nomogram prediction model for early PFNA internal fixation failure in IFF with osteoporosis based on the above influencing factors was 0.893.The receiver operating characteristic curve and calibration curve showed that the model had high discrimination and accuracy.Conclusion The independent related risk factors for failure of PFNA internal fixation are increased CCI score,lateral wall typeⅢ,unstable AO/OTA classification,severe Singh index,and poor fracture reduction quality,and the independent related protective factor is increased lateral wall thickness.The nomogram prediction model established based on this has high discrimination,accuracy and predictive ability,which can provide help for the formulation of clinical prevention and treatment measures.