Auditing patient registration in the Swedish quality register for acute coronary syndrome

AIMS: The present study aims to quantify non-participation in the RIKS-HIA register during 2005 and to compare acute myocardial infarction (AMI) patients registered and not registered in RIKS-HIA, in relation to sociodemographic factors, prevalent disease, and 7-day and 30-day survival. METHODS: We linked information on sociodemographic characteristics, treatments, morbidity, and mortality from the LOMAS (Longitudinal Multilevel Analysis in Scania) database with the RIKS-HIA register. The study population consisted of individuals younger than 85 years living in Scania by 31 December 2004 who had one or more AMI during 2005 (n = 2968). RESULTS: The 70% of the AMI patients included in the register were generally younger, more often men, generally more healthy, more often had AMI as the main diagnosis, and more often underwent revascularisation procedures than AMI patients not included. Among both men (OR(adjusted) = 0.19; 95% CI 0.14-0.27) and women (OR(adjusted) = 0.30; 95% CI 0.20-0.44), registered patients had a lower 30-day mortality than patients not registered in RIKS-HIA. CONCLUSIONS: Even though RIKS-HIA conveys a clear quality improvement for the care of patients with acute coronary syndrome in Sweden, it is important to be aware that the register does not include the entire AMI population, but rather a selected and healthier population of AMI patients. This circumstance decreases the external validity of the information obtained from the RIKS-HIA register. Such an effect might be reduced over time and data from 2006 shows an inclusion rate of 76% among AMI patients aged less than 80 years.