Construction and validation of a frailty risk prediction model for geriatric hematologic neoplasms patients: A cross-sectional study

Abstract

Background

With the increasing incidence of malignant hematological neoplasms in the elderly population, debilitating issues have gradually become an important challenge for patients.

Objectives

To construct a prediction model, draw a nomogram, and perform internal validation of the model.

Methods

505 elderly patients with hematological neoplasms were included in the study. The survey was conducted using research tools such as a general information questionnaire, the Chinese version of the Geriatric 8. A risk prediction model was established and a line chart was drawn to visualize the model after univariate and multivariate Logistic regression analysis. Internal validation of the model was performed using Boot strap bootstrap sampling, calibration curve, receiver operating characteristic curve and area under curve, decision curve analysis to internally validate the model.

Results

After constructing the model and resampling, it was shown that the calibration curve matched the ideal curve well, and the decision analysis curve showed good calibration, discrimination, and clinical benefit within the 0.0–1.0 threshold range.

Conclusion

The prediction model constructed in this study has good predictive effects and can help clinical medical staff to identify the risk of frailty in geriatric hematologic neoplasms patients at an early stage.