Related Experiment Videos
Transient effects in the Cox proportional hazards regression model
E A Mauger1, R A Wolfe, F K Port
1Department of Biostatistics, University of Michigan, Ann Arbor 48109-2029, USA.
Statistics in Medicine
|July 30, 1995
Summary
This study models mortality rates after treatment switching, finding transplant survival improves over time compared to dialysis for end-stage renal disease patients. The model identifies optimal switching times for improved patient outcomes.
Area of Science:
- Biostatistics
- Medical Statistics
- Survival Analysis
Background:
- Modeling mortality rates is crucial for understanding treatment effectiveness.
- Switching treatments, such as from initial therapy to a second-line option, introduces complex transient effects.
- End-stage renal disease (ESRD) patient outcomes depend significantly on treatment modality, like dialysis versus transplantation.
Purpose of the Study:
- To develop and apply a statistical model for mortality rates considering both short-term and long-term effects of treatment switching.
- To compare the survival experience of end-stage renal disease patients undergoing kidney transplant versus those on dialysis.
- To estimate the time at which survival curves for transplant and dialysis patients cross, providing valuable insights for clinical decision-making.
Main Methods:
- Utilized time-dependent covariates within a hazard function model to capture transient effects of treatment switching.
- Employed an exponential decay model to represent the hazard function, allowing for an initial higher mortality risk post-transplant that decays over time.
- Applied the model to analyze data from the Michigan Kidney Registry (MKR) for end-stage renal disease patients, comparing kidney transplant recipients with dialysis patients.
Main Results:
- The exponential decay model demonstrated an initial higher mortality risk following kidney transplant, which subsequently decreased over time.
- Analysis of the Michigan Kidney Registry data revealed a crossover point where the survival experience of transplant patients surpassed that of dialysis patients.
- Methods for estimating these survival curve crossing times and their variances were successfully developed and applied.
Conclusions:
- The developed model effectively captures the dynamic changes in mortality risk associated with treatment switching in end-stage renal disease.
- Identifying the survival crossover time is critical for informing patients and clinicians about the long-term benefits of kidney transplantation over dialysis.
- The findings provide a quantitative basis for advising patients on the optimal timing of treatment decisions in end-stage renal disease management.