Influence of regional cardiovascular mortality on the use of angiography after acute myocardial infarction

Utilization of angiography after acute myocardial infarction (AMI) treated with thrombolytics has been shown in large clinical trials to be related primarily to the availability of the procedure and not individual clinical circumstances. This study evaluated the regional influence of overall population cardiovascular mortality on utilization of angiography in the United States participants of the Global Utilization of Streptokinase and t-PA for Occluded Arteries (GUSTO-1) trial. Published summary statistics from GUSTO-1 and U.S. Census Bureau 1991 delta were evaluated using simple and multiple linear regression with analysis for outliers. Region predictor variables (age adjusted) included mean total cardiovascular deaths/100,000/year (ICD/9 codes 390 to 459), mean coronary artery disease deaths/100,000/year (ICD/9 codes 410 to 414), and mean stroke deaths/100,000/year (ICD/9 codes 430 to 438), with the major outcome being regional proportion of GUSTO-1 patients undergoing angiography during the hospital stay after treatment with thrombolysis. All 3 cardiovascular death rates varied significantly by region (p < 0.00002) with no significant difference in GUSTO-I mortality by region (p = 0.25). Simple linear regression analysis revealed associations between regional death rates and angiography use (r = 0.60, p = 0.12; r = 0.39, p = 0.33; r = 0.81, and p = 0.015). Multiple stepwise linear regression analysis found regional death rate due to stroke as the strongest predictor of angiography use with 65.86% of the variation explained by the model. New England was found to be a consistent outlier with reduced angiography use because of its background regional disease burden. This study confirms regional bias in the use of angiography in GUSTO-1. This form of operator bios appears to be due to more aggressive practice patterns in regions, except New England, where the overall cardiovascular disease burden is greater in terms of lives lost per 100,000 per year. (C) 1997 by Excerpta Medica, Inc.