The specific aims of this project are to investigate the treatment-specific differences in changes in post-operative behavior of patients who have been treated for coronary artery disease (CAD) with surgical intervention and the effects of these differences in behavior change on patient health outcomes and cost of treatment. Percutaneous coronary intervention (PCI)--the newer and less invasive procedure-- and coronary artery bypass graft (CABG) surgery are the surgical interventions for CAD considered. The behavioral changes at issue pertain to cigarette smoking, the absence of a regular exercise regimen, caloric intakes [reflected in part by body mass index (BMI) and overweight/obese], and excessive alcohol consumption. Can differences in these behavior changes between PCI and CABG patients account for evidence suggesting that long-term survival outcomes are no better for the former than for the latter? Is there a measurable impact on health outcomes, particularly post-operative medical expenditures, the need for reintervention, and mortality outcomes, related to these patient behaviors? If so, how do outcomes differ between CABG and PCI patients once changes in behavior are held constant? The hypothesis is that there are differences in post-operative behavior changes, and that those treated for CAD via the relatively less invasive surgery, PCI, will invest in health improvement (through improved behavior) at lower rates than those treated via the more invasive CABG surgery. This hypothesis is rooted in the idea that the two different surgical interventions convey different information about the seriousness of having CAD to the patient being treated. Due to more serious physical and psychological reminders, those treated with the less invasive technology (PCI) will on average reallocate less effort to health investments than those treated with the more invasive surgery (CABG). The project will employ data from the Health and Retirement Survey/Assets and Health Dynamics Among the Oldest Old, the National Long-Term Care Survey, and the National Health Interview Survey. Each survey will be linked to Medicare claims files and to the National Death Index. The multivariate analysis will be based on difference-in-differences and propensity score matching methods.
If the hypothesis of offsetting behavior (less investment in health) by PCI patients turns out to be correct, there are relatively inexpensive and easy ways to implement complementary technologies: targeted physician counseling, behavior-specific patient follow-up, and behavior modification programs. In turn, improved patient health outcomes for those treated with PCI could result: higher survival rates, a decreased need or delayed need for reintervention, and decreases in medical care spending. Overall all these developments would lead to further realization of the potential welfare improvement associated with the introduction of less invasive surgical technology in the treatment of CAD.