Related Experiment Videos
Estimating survival for cost-effectiveness analyses: a case study in atherothrombosis
J Jaime Caro1, Khajak J Ishak, Kristen Migliaccio-Walle
1Caro Research Institute, Concord, MA 01742, USA. jcaro@caroresearch.com
Summary
Estimating long-term survival for economic evaluations requires data-driven methods. This study shows how to derive accurate survival estimates for patients with atherothrombotic disease, accounting for comorbidities.
Area of Science:
- Epidemiology
- Health Economics
- Biostatistics
Background:
- Economic evaluations necessitate survival estimates beyond clinical trial follow-up.
- A data-driven approach was developed to generate these crucial estimates.
Purpose of the Study:
- To demonstrate a robust, data-driven methodology for estimating long-term survival.
- To provide accurate survival data for economic analyses of conditions like peripheral arterial disease (PAD), myocardial infarction, and ischemic stroke.
Main Methods:
- Utilized a large observational dataset (over 50,000 patients) with long-term follow-up (to Dec 31, 2000).
- Estimated mean survival by integrating hazard functions to derive full survival curves.
- Employed Cox proportional hazards analyses across defined periods to ensure model proportionality.
Main Results:
- Mean survival varied: 12.1 years post-stroke to 13.6 years post-PAD diagnosis, adjusted for age.
- Comorbidities decreased mean survival by 1-2 years.
- Multiple vascular diseases or subsequent events significantly reduced life expectancy (by 50% or more, or to <5 years).
Conclusions:
- The study validates analytic methods for precise survival estimation.
- Survival is markedly reduced in patients with atherothrombotic disease, especially with comorbidities.
- Estimates derived from younger trial populations may be optimistic compared to the general patient group.