Key result
Machine learning predicts 1-year hip fracture mortality with ~81% accuracy using basic demographic and lab data.
Why the study?
The study was conducted to determine whether mortality after a hip fracture can be predicted using machine learning models trained on basic demographic and laboratory data, and to identify the key variables associated with 1-, 5-, and 10-year mortality.
Does a machine learning model trained on basic blood, lab test, and demographic data predict 1-, 5-, and 10-year mortality in hip fracture patients?
Observational (n=3,751)
No
Does a machine learning model trained on basic blood, lab test, and demographic data predict 1-, 5-, and 10-year mortality in hip fracture patients?
Effect estimate: AUC 0.79
A machine learning model using basic demographic and lab data can predict 1-year mortality in hip fracture patients with an AUC of 0.79.
May support ML risk stratification in hip fracture patients; leaves open prospective validation before clinical use.
The purpose of this retrospective study was to assess whether mortality following a hip fracture can be predicted by a machine learning model trained on basic blood and lab test data as well as basic demographic data. Additionally, the purpose was to identify the key variables most associated with 1-, 5-, and 10-year mortality and investigate their clinical significance. Input data included 3751 hip fracture patient records sourced from the Medical Information Mart for Intensive Care IV database, which provided records from in-hospital database systems at the Beth Israel Deaconess Medical Center. The 1-year mortality rate for all patients studied was 21% and for those aged 80+ was 29%. We assessed 10 different machine learning classification models, finding LightGBM to have the strongest 1-year mortality prediction performance, with accuracy of 81%, AUC of 0.79, sensitivity of 0.34, and specificity of 0.98 on the test set. The strongest-weighted features of the 1-year model included age, glucose, red blood cell distribution width, mean corpuscular hemoglobin concentration, white blood cells, urea nitrogen, prothrombin time, platelet count, calcium levels, and partial thromboplastin time. Most of these were also in the top 10 features of the LightGBM 5- and 10-year prediction models trained. Testing for these high-ranking biomarkers in new hip fracture patients can aid clinicians in assessing the likelihood of poor outcomes for hip fracture patients, and additional research can use these biomarkers to develop a mortality risk score.
No takes yet. Share an insight, caveat, or question.
Asrian et al. (2023) conducted an observational in Hip fracture (n=3,751). Machine learning model (LightGBM) was evaluated on 1-year mortality prediction performance (AUC 0.79). A LightGBM machine learning model using basic demographic and lab data predicted 1-year mortality in hip fracture patients with 81% accuracy and an AUC of 0.79.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: