This study assessed whether race, sex, and insurance status predict underuse of coronary-artery bypass surgery (CABG) and percutaneous transluminal coronary angioplasty (PCTA) procedures among patients who need revascularization, and, if so, whether type of hospital ownership or availability of revascularization facilities are related to this underuse. The study sample included all patients who had coronary angiography for suspected atherosclerosis in 1992 in New York City; oversampling was conducted in order to obtain a sufficient number of minority patients. Data were obtained from record abstractions. Patients with significant coronary artery disease with need of revascularization (as determined by study protocol) were retained for the study.

Of the 631 patients who met criteria, 27% were African American, 29% were Hispanic and 44% were white. Overall, 74% of the patients had cardiac revascularization (95% confidence interval=71% to 77%). The rates did not differ significantly by race, sex, or insurance status. However, rates did vary by hospital type, from 21% to 87%. Rates were lower in hospitals where revascularization was conducted off-site (59%, 95% confidence interval=56% to 65%) than in hospitals with revascularization facilities (76%, 95% confidence interval = 74% to 79%). These results did not change when analyses were limited to those patients with left main or three-vessel disease. It is noteworthy that rates of recommendation for revascularization also varied substantially among hospitals and followed the same patterns as rates of revascularization procedures (85% in onsite hospitals versus 75% in offsite hospitals).

After adjusting for patient and hospital characteristics, the only factors that significantly explained either receiving revascularization or receiving a recommendation for revascularization were being treated in a hospital with on-site revascularization facilities and having more severe disease. When analyses were restricted to hospitals without on-site facilities, there were lower rates of revascularization for women (48% for women versus 64% for men) and higher rates for African Americans (74% for African Americans versus 42% for Hispanics and 59% for whites). Neither of these two patterns reached statistical significance, probably due to the restricted sample size. Similar patterns were observed for recommendation in offsite hospitals.

The reasons given for lack of revascularization at onsite hospitals were no recommendation (15% of cases), patient refusal (5% of recommended cases), death (1% of recommended cases), and unknown (2% of recommended cases). The reasons given for offsite hospitals were no recommendation (27% of cases), patient refusal (10% of recommended cases), and unknown (21% of recommended cases).

The authors conclude that, while 26% of patient who met the criteria for revascularization failed to receive this procedure, the rate of underuse was not higher in women, in African Americans or Hispanics, or in uninsured patients. However, the underuse was associated with not having revascularization facilities onsite. The authors suggest that, “in addition to the lack of availability of services, the lack of provision of CABG surgery and PTCA may be a proxy for constrained resources that lead to failure to follow up on the results of angiography to achieve needed revascularization. This is supported by this study in which three of the four hospitals without onsite facilities were municipal hospitals, institutions known to be chronically underfinanced. In contrast, all of the hospitals with onsite facilities but one were private referral hospitals.”

The difference in recommendation rates between onset and offsite hospitals accounted for half of the total difference in revascularization rates. Several factors that may be responsible include the possibilities that more patients in offsite hospitals than in onsite hospitals had mitigating circumstances (drug addiction, old age, low ejection fraction); offsite physicians used stricter criteria for recommendations (severe cases were recommended at similar rates to onsite hospitals, but non-acute cases had lower rates of recommendation at offsite hospitals than onsite hospitals); and offsite physicians made recommendations less often to those who were uninsured, which was not true at onsite hospitals. Furthermore, the authors note that the lower procedure rates, in addition to lower recommendation rates, could be due to problems in scheduling, transfer, and compliance.

In an effort to explain the cases for which the reasons for not performing revascularization were unknown, the authors suggest that “cultural, financial, and institutional barriers to the receipt of necessary procedures can be substantial. Many patients in the offsite hospitals are poor, members of ethnic minority groups, and recent immigrants for whom cultural and communication barriers to care can be significant. These factors and the lack of a primary care physician who provides counseling, understanding, and follow through have been shown to result in the failure of patients to understand the need for treatment and to obtain them.” They also add that lack of ability to pay might also be a significant barrier that might explain the observation reported from this study.

Because these results differ from previous studies with regard to racial patterns, it is necessary to consider the potential selection biases that might have resulted in artifactual findings in this study.