gms | German Medical Science

Introduction: International comparisons of health outcomes often use OECD indicators [1]. The indicator ‘AMI 30-day-mortality using unlinked data’ describes hospital mortality after acute myocardial infarction (AMI, ICD-10: I21-I22) and shall reflect quality of acute care. According to this indicator, AMI hospital mortality is higher in Germany than in most European countries. Yet, in 2020 in Europe, Germany ranked high on health expenditure, availability of cardiologists and catheterization facilities [2]. We discuss, whether the OECD indicator allows valid conclusions regarding quality of acute care and whether more reliable sources exist.

Methods: International literature and public health reports on AMI mortality, patient registration and health care systems were reviewed. Experts from European countries were contacted to share their experiences and knowledge. OECD indicators ‘AMI 30-day-mortality using unlinked data’ and ‘average length of stay after AMI’ for European countries were downloaded from the OECD database (https://stats.oecd.org) for 2000-2019. Changes in AMI mortality were graphically described. The influence of average length of stay on AMI mortality was estimated by linear regression.

Results: Since 2000, OECD AMI hospital mortality decreased strongly in Europe. In 2019, it was 2.9% in the Netherlands and 3.2% in Norway, but 8.5% in Germany. Rules for registration of in-patients differ in Europe, so that day cases are not always included. Different organization of health care equally influences calculated AMI hospital mortality. Average length of stay in hospital after AMI is closely associated with reported AMI mortality (R²=0.58).

AMI hospital mortality generally is influenced by the number of patient transfers. In countries with centralized catheterization facilities, patients are more often transferred between hospitals than in countries with de-centralized facilities. Each patient transfer increases the number of patient cases, lowering calculated AMI hospital mortality.

AMI hospital mortality reported from European registries and cohorts differs greatly – even compared to OECD data of the same country – due to diverse inclusion criteria and AMI definitions.

Discussion: The OECD indicator for AMI hospital mortality is highly influenced by factors that are not related to patient care, but to health care organization, patient registration [3] and national re-imbursement strategies. AMI-30 day mortality including a patient follow-up after hospital discharge is less influenced by these factors and would allow more valid conclusions. Yet, due to restrictions in data linkage, it is not universally available. AMI hospital mortality and AMI 30-day mortality reported from registries and cohort studies differ from OECD reported AMI mortality and are similarly difficult to compare [3], [4]. Registries differ in populations included. Coverage of all ages, all types of acute care units or all types of hospitals is often lacking. Quality of acute care and outcome after AMI can only be compared with registries in sentinel regions in Europe with coordinated inclusion criteria.

Conclusion: The OECD indicator for AMI hospital mortality does not mirror quality of acute care, but foremost organisational differences in health systems. Conclusions about quality of care from registry and cohort data are similarly limited valid. European sentinel registries are needed to compare patient care and outcome in AMI.

Die Autorin wurde im Rahmen eines Projekts von der Deutschen Herzstiftung e.V. finanziert.

The authors declare that an ethics committee vote is not required.