Hospitalization for acute myocardial infarction – trends in case fatality, and the impact of changing definition on number of events, subtypes and mortality

Abstract

Background: Few studies of acute myocardial infarction (AMI) among hospitalized patients have had direct estimates of long-term case fatality after AMI, and the literature is conflicting regarding gender-related prognosis following AMI. Several revisions of the definition of AMI have complicated the analyses of trends in the number of AMI events and mortality. Aims: To provide age and gender specific estimates of trends in short- and long-term case fatality for a first AMI of patients hospitalized, and to analyse the impact of applying five different definitions of AMI on number of AMI events, types and long-term mortality. Methods: Paper I included patients hospitalized with a first AMI at Haukeland University Hospital during 1979-2001. Data were retrieved from The Western Norway Cardiovascular Registry and included additional data on long-term and all-cause death. The study period was divided into three periods and these were compared (p-trend). The clinicians used the WHO definition of AMI in the study period. Paper II and Paper III examined a different patient cohort, which was hospitalized for AMI (815 patients) during 1 March 2002 and 28 February 2003. The cohort also included 679 patients hospitalized in the same period with at least one measurement of elevated cardiac troponin I (cTnI) during the hospital stay, but not diagnosed with AMI. Results: Paper I: The short- and long-term case fatality declined substantially during 1979-2001 in 11878 patients hospitalized with a first AMI. The unadjusted 28-day case fatality declined from 31.1% to 19.8% in men and from 37.3% to 26.8% in women from the first period (1979-1985) compared with last period (1994-2001) (p-trend <0.0001). Landmark analysis showed continued decline in 1-10 year case fatality. Case fatality rates were significantly lower in women than men in patients >60 years. Paper II: The WHO 1979 definition of AMI with CK-MB mass as biomarker was used as reference and resulted in 566 definite AMIs among the 1494 total cohort of patients. When applying the ESC/ACC 2000, the AHA 2003, the Universal 2007 and the Universal 2012 definition of AMI with troponin I as biomarker we observed approximately 30% more AMI events. The short- and long-term mortality were moderately higher applying the newer definitions. Paper III: Applying the Universal 2012 definition compared with the Universal 2007 definition of AMI resulted in a minimal decrease in number of AMI events from 769 to 760, with numbers of patients classified with Type 1, 2, 3, 4a, 4b and 5 AMI according to the Universal 2012 definition being 685, 27, 28, 13, 3 and 4 patients respectively. Conclusions: There has been a substantial decline in short-term and long-term case fatality in patients hospitalized for a first AMI during 1979-2001. Women >60 years fare better than men do when we compare age-adjusted case-fatality rates. The number of AMI events depends on the definition of AMI and biomarker. When the WHO 1979 definition of AMI was used as reference, we found that applying three newer definitions of AMI resulted in approximately 30% more patients diagnosed with definite AMI. Few patients classified as Type 2 AMI is specifically noticed.