Objective: We convened a multinational panel to develop appropriateness criteria for percutaneous transluminal coronary angioplasty (PTCA) and coronary artery bypass graft (CABG). To assess the applicability of these criteria, we applied them to patients referred for coronary revascularization. Finally, to understand how multinational criteria may differ from criteria developed by a panel of physicians from one country, we compared the appropriateness ratings using the multinational panel's criteria and those made using similar criteria previously developed by a panel of Dutch physicians.
Methods: We conducted a prospective survey and review of the medical records of 2363 consecutive patients presenting with chronic stable angina or following a myocardial infarction who were referred for PTCA (n=1137) or CABG (n= 1226) at ten Dutch hospitals performing coronary revascularization. Appropriateness was measured using two sets of criteria developed by: (1) a Dutch panel of cardiologists and cardiothoracic surgeons in 1991; and (2) a similarly composed European panel in 1998.
Results: More PTCA referrals were rated inappropriate by Dutch criteria compared with multinational criteria among both patients with chronic stable angina (34.8 versus 6.1%; P< 0.001) and those with a recent myocardial infarction (28.1 versus 0.9%; P< 0.001). Among those patients referred for bypass surgery, the Dutch criteria judged a greater proportion of cases inappropriate than multinational criteria did for patients with chronic stable angina (3.7 versus 1.5%, P< 0.001). The proportion of cases rated inappropriate for bypass surgery among patients following a myocardial infarction was similar between the two panels (3.9 versus 2.4%, respectively; P=0.40). After reclassifying the data for two of the clinical factors used in the appropriateness criteria (lesion morphology and intensity of medical therapy) based on evidence that appeared in the literature after the Dutch panel met, we found no significant differences between the Dutch and multinational panels' appropriateness ratings.
Conclusions: While fewer cases were judged inappropriate using the multinational criteria compared with the Dutch criteria, the differences in ratings were related primarily to the clinical factors used by each panel. These findings support the review of appropriateness criteria, and other forms of clinical guidelines, to ensure that they are current with the clinical evidence before using them to assess clinical care. Developing such criteria using a multinational panel, in contrast to multiple single country panels, would be a more efficient use of resources.