Respiratory
A Novel, Non-Arbitrary Determination of Amenable Deaths – Exemplified for Chronic Obstructive Pulmonary Disease (COPD) in the 50 US-States Ulrich Mueller* Ulrich Mueller Yesser Falk
Since Rutstein et al. (1976) calculating amenable deaths as proportion of all deaths (general or specific causes) for measuring health system quality found global reception. Typically, a list of most frequent causes of death is selected – the longer the list the less comparable with other populations or over time. Proportion amenable for every cause of death and population is arbitrarily estimated, differences between populations or over time are speculation. Major Proponents from Europe are Mackenbach et al. (2015), from North America Nolte & McKee (2011) with many followers.
Here, we describe non-arbitrary measuring amenable deaths by a best-practice approach: Setting-up cause-of-death specific life-tables (Namboodiri 1990) for subpopulations, deriving a superpopulation-normative-life table (Murray et al. 2012) from minimal age-specific subpopulation mortalities, amenable deaths by age level are determined by the difference between the life-expectancies of real subpopulations and the one in the normative-life table
We illustrate this approach for Chronic Obstructive Pulmonary Disease (COPD). Data – on state level only unisex – are from the US National Center for Health Statistics. Within the US, California has the lowest proportion of amenable COPD deaths, Texas moderate, Kentucky highest. In 2016, for COPD-50-year-olds California’s life expectancy was 21.01, Texas’ 17.82, Kentucky’s 11.94. At age 84, individuals with COPD in the states observed and in normative COPD life tables have less than one year of life expectancy (normative = 0.72 years, California = 0.72 years, Texas = 0.57 years, Kentucky = 0.41 years).
For frequent causes of death and good data availability, our approach with easy-to-understand, informative results, may be superior to all published alternatives. Our concept can also be applied to general mortality, or further divided up in various mutually exclusive and collectively exhaustive clusters of specific mortality / causes-of-death.
