Establishing a Conceptual Framework for the Use of the OECD Indicator “Avoidable Mortality
S Stolpe, K Kunde, HC Vollmar, B KowallAbstract
Introduction
Avoidable mortality is an OECD indicator used to monitor health system performance. It captures premature deaths (<75 years) from causes preventable by public health interventions or amenable to healthcare. Its use implicitly assumes that it enables reliable comparisons of health system performance. However, the validity of this assumption is uncertain, as a comprehensive conceptual framework stating conditions under which avoidable mortality reflects health system performance is lacking.
Methods
We developed a conceptual framework by defining key assumptions required for valid inference from avoidable mortality to health system performance. Potential threats to these assumptions, their empirical relevance and implications on the validity of the indicator were identified based on a literature review.
Results
The proposed framework comprises six core assumptions. Three underpin the causal link between avoidable mortality and health system quality: (1) comparability of baseline morbidity, (2) comparable access to diagnostic, and (3) consistent relationship between disease-specific interventions and mortality outcomes. Three additional assumptions ensure cross-country comparability: (4) uniformly high-quality death certification, (5) consistent and unambiguous attribution of causes of death, and (6) consistent and reliable reflection of health system performance. Evidence indicates that these assumptions are frequently violated. Population-level genetic predispositions (e.g. for Chronic obstructive pulmonary disease), high prevalence of multimorbidity, variation in cause-of-death data quality, country-specific certification preferences, and different disease latency periods compromise validity.
Conclusions
Avoidable mortality has limited validity as an indicator of health system performance, as key underlying assumptions are only partially met. Cross-country comparisons should therefore be interpreted with caution regarding health system performance.