Mortality data, based on a single underlying cause of death (derived from multiple conditions on the death certificate), has provided critical evidence for policy and practice for over a century. However, radical shifts in patterns of death over recent decades, to older ages and predominantly chronic disease, mean that only the minority of deaths truly have only one causal condition. Moreover, it is likely to cause bias (including under- and over-representation of certain causes of death), misrepresent the certainty of data and lead to loss of highly informative data, including on cause-specific mortality rates and burden of disease, undermining basic evidence for policy and practice.
Despite global recognition of the urgent need to better integrate data on multiple causes of death (MCoD) into mortality statistics, statistical methods for their analysis remain patchy and underdeveloped. The 2018 introduction of ICD-11, enriching data on comorbidity, confers further urgency and a unique opportunity to prioritise devising pragmatic tools to optimise use of MCoD in mortality reporting, the overarching aim of this project.
This project brings together outstanding research, clinical and policy expertise to develop and test novel statistical methods, guided by an audit of the global evidence to: (i) identify current MCoD methods and gaps; (ii) characterise MCoD in Australia; (iii) develop and test statistical methods for MCoD quantification to fill the most important methodological gaps; (iv) quantify the impact of using MCoD vs conventional methods; and (vi) develop an MCoD framework and toolkit for implementation in practice. It will make use of the National cause of Death collection and large-scale cohort study with linked health data. The findings will optimise data for policy and practice and enable a more complete picture of disease processes at the end of life.