The normalised mortality, a new measure to ensure valid conclusions from cardiovascular mortality

Abstract Background The quality and efficacy of health systems is assessed by comparing age-standardised mortality (ASM) from cardiovascular diseases (CVD), as they are regarded avoidable by health care. Differences in CVD mortality in West-Europe are surprising and hard to explain: in 2018, ASM for CVD in Germany was much higher (women: 102, men: 155/100,000) than in the Netherlands (65 and 97). Yet, these differences do not root in health system quality, but in different shares of CVD deaths among all deaths (DE: 39%, NL: 25%), which strongly effect CVD mortality. We developed a mortality measure less sensible to differences in shares of CVD deaths to increase plausibility of comparisons and validity of conclusions. Methods WHO data on sex-specific all and CVD deaths, ASM for CVD and all-causes for 2000-2018 for Western-European countries was used to calculate the share of CVD deaths among all deaths. ASM for CVD was divided by the share of CVD deaths resulting in a mortality measure normalized for 1% of CVD deaths. ASM and normalized mortality for CVD were used in comparative analyses. Strength of linear association between total and (normalized) ASM for CVD was estimated (adj. R²). Results Normalised CVD mortality ranked from 2.0 (CHE in women) to 4.9 (SVN in men) in 2018. While ASM for CVD was lowest in Danish women and Dutch men, Switzerland leads both rankings using the normalised CVD mortality. While ASM for CVD was much higher in Germany than in the Netherlands, the countries had the same normalized CVD mortality (2.6). ASM for CVD was moderately associated with all-cause mortality only in men (adj. R² 0.33; in women: -0.06). Using the normalized CVD mortality yielded an adj R² of 0.66 in women and 0.94 in men. The normalised CVD mortality ensures reliable comparisons of CVD mortality eliminating the effect of different shares of CVD deaths. It yields plausible results and enables valid conclusions about health system quality and better decisions for public health. Key messages • Age-standardised CVD mortality rates are effected by different shares of CVD deaths in countries and hinder valid conclusions on health systems efficacy and quality of health systems. • The normalised mortality rate eliminates the effect of different shares of CVD deaths in countries and allows to validly compare CVD mortality.