Development and Validation of a Diagnostic Model for Predicting the Probability of Nontraumatic Osteonecrosis of the Femoral Head

Qiang Xu,Hangjun Chen,Sihai Chen,Jing Shan,Guoming Xia,Zhiyou Cao,Xuqiang Liu,Min Dai

DOI: https://doi.org/10.2139/ssrn.3502343

IF: 4.6

2020-01-01

Scientific Reports

Abstract:Background Although corticosteroids and alcohol are two major risk factors for nontraumatic osteonecrosis of the femoral head (NONFH), the effects of other factors have rarely been studied, thereby making early diagnosis and treatment of NONFH difficult. This study aimed to develop and validate a nomogram to estimate the probability of NONFH using clinical risk factors other than corticosteroids and alcohol consumption. Methods A training cohort of 790 patients (n=434, NONFH; n=356, femoral neck fractures [non-NONFH]) diagnosed in our hospital from January 2011 to December 2016 was used for model development. A least absolute shrinkage and selection operator (lasso) regression model was used for date dimension reduction and optimal predictor selection. A predictive model was developed from univariate and multivariate logistic regression analyses. Performance characterisation of the resulting nomogram included calibration, discriminatory ability, and clinical usefulness. After internal validation, the nomogram was further evaluated in a separate cohort of 300 consecutive patients included between January 2017 and December 2018. Results The simple prediction nomogram included five predictors from univariate and multivariate analyses, including gender, total cholesterol levels, triglyceride levels, white blood cell count, and platelet count. Internal validation showed that the model had good discrimination (area under the receiver operating characteristic curve [AUC]=0.80) and calibration. Good discrimination (AUC=0.81) and calibration were preserved in the validation cohort. Decision curve analysis showed that the predictive nomogram was clinically useful. Conclusions The simple diagnostic nomogram, which combines demographic data and laboratory blood test results, was able to quantify the probability of NONFH in cases of early screening and diagnosis.