To PCI or not to PCI? Researchers develop predictive tool for stable CAD

Using data from the randomized COURAGE trial, a team of researchers developed a model to predict one-year angina and quality-of-life outcomes among patients with stable coronary artery disease (CAD).

The authors said their model showed “modest/good” discrimination, with a C statistic of 0.72-0.82, and suggested percutaneous coronary intervention (PCI) provides a small but consistent benefit in these areas compared to medical therapy alone. Senior study author John A. Spertus, MD, MPH, and colleagues believe their tool could promote shared decision-making between physicians and patients, particularly in the setting of stable heart disease where treatments may only be separated by incremental improvements to quality-of-life.

“Developing tools to share with patients their one-year angina and health status to improve shared medical decision-making should be prospectively tested to illuminate their benefit in improving the patient-centeredness and value of PCI,” Spertus et al. wrote in Circulation: Cardiovascular Quality and Outcomes.

The model was developed using data from 2,287 stable ischemic heart disease patients from the COURAGE trial, which randomized participants to either PCI plus medical therapy or optimal medical therapy alone. All patients took the Seattle Angina Questionnaire (SAQ) multiple times during follow-up to assess the frequency of angina, physical limitations related to chest pain and quality of life scores.

Each portion of the SAQ is graded on a 0 to 100 scale, with higher scores representing better health.

PCI was associated with SAQ scores an average of 2.39 points higher. There were no significant interactions between PCI and other factors, suggesting the treatment maintains this modest benefit across all baseline characteristics.

Half of the patients in the medical therapy group were predicted to be free of angina at one year, compared to 57 percent of patients who received PCI plus medical therapy. Fifty-eight and 66 percent of patients in the PCI group were classified as having very good/excellent physical function and quality of life, respectively, compared to 55 and 59 percent of individuals receiving only medical therapy.

Baseline angina frequency, age, diabetes status and medication use were more strongly associated with these outcomes than whether a patient received PCI or not, the researchers found.

 “Given a growing recognition of the need to engage patients in shared decision-making, there is a need to develop tools to provide individualized estimates of outcomes so that patients and their physicians can better understand how different treatment approaches might affect a particular patient,” the authors wrote. “Although there were substantial variations in one-year health status across different domains that could be predicted by patient characteristics, PCI offered only a very small improvement over OMT (optimal medical therapy) alone. These results from COURAGE can help clinicians support shared decision-making with their patients by directly estimating the modest improvements in outcomes with PCI.”

One limitation of the model is it relied on the intensive medical treatment given to COURAGE participants 15 years ago. If treatments were different in current practice, it may cause the model to be less accurate, the researchers noted. In addition, SAQs would need to be collected from patients for the tool to be employed.

Spertus and colleagues also pointed out certain factors such as socioeconomic status weren’t measured and could further improve the prediction model.