P5246Nationwide linked administrative data study to explore the variations in statistical classification, management and 30-day mortality of acute myocardial infarctions in Latvia

Abstract

Abstract Background According to statistics of Organisation for Economic Cooperation and Development (OECD) Latvia has one of the highest 30-day mortality for acute myocardial infarction (AMI). The output data used in the calculation of this indicator was analysed at a hospital level to explain the causes of it. Methods Linked to Causes of death registry administrative reimbursement system data from 2014 to 2017 were analysed. Defined exclusion criteria were applied and 11675 emergency AMI admissions were indexed. As evaluation of the inter-hospital differences in non-ST elevation myocardial infarction (NSTEMI) revealed significant inconsistency in statistical classification of non-ST elevation acute cardiac events, the 30-day mortality and the factors influencing it, was exclusively analysed for 9168 ST elevation myocardial infarction (STEMI) cases. The outcomes of different reperfusion scenarios were analysed for five groups of hospitals of size, location, the availability of percutaneous coronary interventions (PCI) and cardiac surgery - local hospitals without PCI, n=10 (non-PCI local), regional hospitals without PCI, n=5 (non-PCI regional), regional hospitals with part-time PCI, n=2 (PCI regional), university hospital with PCI, n=1 (PCI general univ.), university cardiology tertiary hospital with PCI, n=1 (PCI tertiary univ.). The multivariable logistic regression was employed to adjust the results for potential explanatory variables as patient age, gender, comorbidities (measured as Charlson index (Ch. index)), the distance between their residence and hospitals. Results 30-day mortality for all indexed AMI admissions (17,4%) was in line with the value of the indicator in OECD reports. However only 21,5% of all cases have been coded as NSTEMI (12,2% mortality). The mortality of the patients with STEMI have reached 18,8% composed by mortality in range from 13,7% for patients directly admitted to hospitals having both PCI and cardiac surgery units to 23,7% for patients initially admitted to local hospitals. Significant variations in mortality between reperfusion scenarios as well as the variations in the frequency of application reperfusion methods between hospital groups were observed. Some variation in mortality between hospital groups persisted also after adjustment to the reperfusion therapy scenario and other potential confounders. Distribution of STEMI patients Conclusion The improvement of statistical classification practice for NSTEMI is needed to measure the actual AMI mortality in Latvia. Although the 30-day mortality for AMI may be lower than internationally reported because of under-accounting of prognostically more favourable NSTEMI cases, it is still very high. Among patients with STEMI, there are wide inter-hospital variations in care and mortality. There is a substantial room for improvements in all types of hospitals to increase reperfusion rates and to reduce delays in initiation of it Acknowledgement/Funding University of Latvia/Centre for Disease Prevention and Control joint grant for project “Transparency for health care quality and efficiency”