<![CDATA[Modeling patient recovery rates has often been conducted using the Cox proportional hazards model. However, this model assumes that independent variables satisfy the proportional hazards assumption, which is not always the case. To address such violations, the Cox stratified regression model offers an alternative by stratifying variables that fail to meet the assumption. This study aims to develop a Cox stratified model for the recovery rate of stroke patients and identify factors significantly affecting the model. A descriptive quantitative study design was employed, using medical records of stroke inpatients at Dadi Hospital from January to December 2023. The analysis revealed that the hemiplegia variable violated the proportional hazards assumption at a 5% significance level, necessitating the use of Cox stratified regression. Cholesterol was identified as the only factor significantly affecting the model at the same significance level. The final model, based on interaction testing, was a Cox stratified model without interaction, with cholesterol as the key variable influencing the recovery rate]]>
Copyrights © 2025
