Effects of Longitudinal Changes in Charlson Comorbidity on Prognostic Survival Model Performance Among Newly Diagnosed Patients with Hypertension

Date
2015-05-01
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Objectives: To assess methods of defining comorbidities by comparing risk adjusted mortality predictive model fit and performance among newly diagnosed hypertensive population. Methods: We included nearly all patients 18 year and older with an incident diagnosis of hypertension from one Canadian Province. We compared prognostic model performance for Cox regression models using Charlson comorbidities as time-invariant covariates (TIC) at baseline and time-varying covariates (TVC). Cox regression was used to calculate hazard ratios. Model fit and performance was based on the comparison of the AIC and Likelihood Ratio. Results: All Cox regression time-varying covariate models (TVCMs) outperformed time-invariant covariate (TIVMs) baseline models, based on a comparison of AIC and Likelihood Ratio, regardless of the method used to adjust for individual risk using the Charlson Comorbidities. TVCMs included all 17 Charlson comorbidities as individual independent variables showed the best fit and performance compared with similar baseline models, AIC (1,670,491 to 1,720,126) and Likelihood Ratio (112,941.72 to 63,239.78) respectively. Conclusion: Accounting for changes in patient comorbidity status over time more accurately capture a patient’s health risk and improves predictive model fit and performance over longer follow-up periods than traditional baseline method.
Description
Keywords
Epidemiology, Health Care Management, Medicine and Surgery
Citation
Rymkiewicz, P. (2015). Effects of Longitudinal Changes in Charlson Comorbidity on Prognostic Survival Model Performance Among Newly Diagnosed Patients with Hypertension (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/25827