Do operator volumes relate to clinical outcomes after percutaneous coronary intervention in the Canadian health care system?

Abstract

Many US studies have documented an association between operator volume and outcomes after percutaneous coronary intervention (PCI). No study has assessed whether this relationship exists in Canada, where PCI is performed only at a limited number of regional centers and operator volumes are higher. All PCI procedures performed in the province of Ontario from 1995 to 2001 were analyzed using administrative databases. The outcomes of interest were coronary artery bypass graft during the same hospitalization, mortality at 30 days, or the combined end point. A total of 38,561 PCI procedures were performed by 65 physicians at 8 centers. Over the study period, risk-adjusted coronary artery bypass graft rates fell from 2.0% in 1995 to 0.7% in 2000 (P < .0001) with no change in mortality. The median annual PCI volume was 132 (25th, 75th percentile: 81, 182) cases. After stratifying operators by average annual PCI volume into low (< 155 cases), intermediate (155-195 cases), and high (> 195 cases) volume, there were no significant linear relationships between risk-adjusted outcomes and operator terciles. No significant correlations were seen between individual PCI volume and risk-adjusted rates of mortality, bypass surgery, or the combined end point (P = .2, P = .35, and P = .95, respectively). In contrast to US studies, there does not appear to be an association between PCI volume and outcomes in Ontario. These findings may be related to the high annual volumes of most operators and institutions within Ontario.