Predicting Prolonged Hospital Stays in Elderly Patients With Hip Fractures Managed During the COVID-19 Pandemic in...

Area of Science:

• Geriatric Medicine
• Health Informatics
• Artificial Intelligence in Healthcare

Background:

• Prolonged length of stay (LOS) following hip fracture is a significant risk factor for increased mortality in elderly patients.
• The COVID-19 pandemic introduced unique challenges to healthcare management, potentially impacting patient outcomes like hospital LOS.

Purpose of the Study:

• To develop and validate an artificial neural network (ANN) model for predicting prolonged LOS (≥14 days) in elderly Chilean hip fracture patients.
• To identify key predictors of prolonged LOS within the context of pandemic-related healthcare management.

Main Methods:

• Utilized an official national database of 2686 hip fracture patients managed in 43 Chilean public hospitals during 2020.
• Developed an artificial neural network (ANN), a machine learning model, trained on 80% of the data and tested on 20%.
• Evaluated ANN performance using the area under the curve of the receiver operating characteristic curve (AUC-ROC).

Main Results:

• 30.2% of patients experienced prolonged LOS (≥14 days).
• The ANN achieved fair predictive accuracy, with AUC-ROC values of 0.745 in the training set and 0.742 in the test set.
• The most significant predictors identified were admitting hospital, geographical health service, and surgery within 2 days of admission.

Conclusions:

• An ANN model demonstrated fair accuracy in predicting prolonged LOS for elderly hip fracture patients during the COVID-19 pandemic using national big data.
• Predictors of prolonged LOS were primarily administrative and organizational, highlighting systemic factors over individual patient health.
• Findings suggest opportunities for improving hospital resource allocation and patient flow through data-driven insights.