The Devil Is in the Data: Can Regional Variation in Amenable Mortality Help to Understand Changes in Health System...

The contribution of health systems to health is commonly assessed using levels of amenable mortality. Few such studies exist for Poland, with analyses of within-the-country patterns being particularly scarce. The aim of this paper is to analyse differences in amenable mortality levels and trends across Poland's regions using the most recent data and to gain a more nuanced understanding of these differences and possible reasons behind them. This can inform future health policy decisions, particularly when it comes to efforts to improve health system performance. We used national and regional mortality data to construct amenable mortality rates between 2002 and 2019. We found that the initially observed decline in amenable mortality stagnated between 2014 and 2019, something not seen elsewhere in Europe. The main driver behind this trend is the change in ischemic heart disease (IHD) mortality. However, we also found that there is a systematic underreporting of IHD as a cause of death in Poland in favour of heart failure, which makes analysis of health system performance using amenable mortality as an indicator less reliable. We also found substantial geographical differences in amenable mortality levels and trends across Poland, which ranged from -3.3% to +8.1% across the regions in 2014-2019. These are much bigger than variations in total mortality trends, ranging from -1.5% to -0.2% in the same period, which suggests that quality of care across regions varies substantially, although some of this effect is also a coding artefact. This means that interpretation of health system performance indicators is not straightforward and may prevent implementation of policies that are needed to improve population health.