The problem of attributing deaths of nonadherers: the VA coronary bypass experience

Area of Science:

• Cardiology
• Clinical Trials
• Health Services Research

Background:

• The VA Coronary Artery Surgery Study (CASS) investigated medical versus surgical treatment for coronary artery disease.
• A common concern in such trials is whether patients switching treatments (crossover) are at higher risk than those who adhere to their assigned strategy.

Purpose of the Study:

• To analyze the characteristics and risks associated with patients crossing over from medical to surgical treatment in the CASS trial.
• To determine if crossover patients had different baseline risk factors compared to those who adhered to medical therapy.
• To evaluate the validity of survival analysis methods that account for treatment crossover.

Main Methods:

• Analysis of patient data from the VA Coronary Artery Surgery Study (CASS).
• Examined baseline risk factors (angiographic, noninvasive, clinical) for patients who crossed over from medical to surgical treatment.
• Compared risk factors between patients who crossed over and those who adhered to medical treatment.

Main Results:

• 84 out of 354 patients randomized to medical treatment underwent subsequent bypass surgery.
• Crossover rate was approximately 4% per year for patients without left main coronary artery disease.
• Crossover was not associated with the number of diseased vessels or other baseline risk factors, but was linked to angina severity.

Conclusions:

• The assumption that crossover patients are at greater risk was not supported by baseline data.
• Angina severity, though a reason for crossover, was not a significant risk factor for mortality in this study.
• Findings support the validity of analyzing survival data using the crossover method, treating nonadherers as lost to follow-up at the time of treatment change.