 | The 131st Annual Meeting (November 15-19, 2003) of APHA |
|
| The 131st Annual Meeting (November 15-19, 2003) of APHA |
Sunny Kim, PhD, Health Psychology, Ohio University, 236 Porter Hall, Athens, OH 45701, (740)597-1963, kims2@ohio.edu
Objectives: The purpose of this study is to illustrate the consequences of ignoring the time dependency of a covariate in a longitudinal study. It is called a fixed covariate if the values measured on each individual at the beginning of a follow-up remain fixed over time. Sex, treatment group, type of surgery, and ethnicity are examples of fixed covariates. On the other hand, it is a time-dependent if the values of the covariates are changing over time such as marital status, blood pressure, and the presence of depression. Method: Using real data containing time dependent covariate, the problem of ignoring the time of occurrence in a proportional hazard model was demonstrated. Originally, data were collected from renal transplanted patients to study the impact of post transplant diabetes on the overall survival. The occurrence of diabetes is a time dependent covariate. Some patients developed diabetes right after renal transplantation, while others developed diabetes many years after the kidney transplantation. Result: When the time of diabetes development was ignored, diabetes appeared to be not associated with a patients’ survival time (Hazard ratio =1.1; 95% confidence interval 0.95-1.17). However, when the time of diabetes occurrence was considered, the occurrence of diabetes was a significant risk factor of a patients’ survival (Hazard ratio=1.8; 95% confidence interval 1.35-2.42). Conclusion: The study results could be misleading if a time dependent covariate is treated as a fixed covariate. The correct use of statistical procedures is critical for valid conclusions of a study.
Learning Objectives:
Keywords: Statistics, Methodology
Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.