The Hidden Burden of Mortality Across the Spectrum of ICD-10 Conditions in Australia: A Multiple Cause of Death Analysis

Abstract Background Death certificates record both an underlying cause and contributing conditions, yet mortality statistics predominantly report only the underlying cause. We quantify this “hidden burden” across all ICD-10 conditions in Australian mortality data using the multiple-to-underlying ratio (MUR): total death certificate mentions divided by underlying cause deaths. Methods We analysed Australian Bureau of Statistics Causes of Death 2023 data (N = 187,268 registered deaths) to compute the ratio for all ICD-10 conditions. Three pre-registered confirmatory hypotheses tested sex differences in hypertension and mental health ratios, and geographic variation by preventability, with Holm-Bonferroni correction. Results Death certificates recorded an average of 3.5 causes per decedent, meaning the underlying cause captures only ∼29% of recorded morbidity. Of 663 conditions with ≥10 underlying cause deaths, the ratio ranged from 1.0 (external causes) to 281.1, with a median of 2.5. Among conditions with stable estimates (≥50 underlying deaths), the highest ratio was 94.3 (complications of medical care). Age explained only 10.9% of ratio variation (R 2 = 0.109), and no top-ranked conditions were identified as primarily age-driven, suggesting the ratio ranking is robust to age confounding. However, external validation using US CDC data showed age standardisation materially changed absolute ratio values for 6 of 8 cause groups (divergence 16–34%), with the direction varying by condition rather than following a simple age-concentration pattern. Males showed consistently higher ratios, most strikingly for mental health disorders (62% higher); a counterfactual analysis estimated suicide coding rules explain only 6–15% of this sex difference. Three pre-registered hypotheses were null after correction; H1 and H3 (sex differences) were underpowered (n = 4, n = 8 pairs) with large effect sizes (r = 0.77–0.80), while H2 (geographic variation) showed a clear null. Conclusions The hidden burden of mortality in Australia is substantial and unevenly distributed, with symptom codes, mental health conditions, and hypertension most undercounted. The ratio provides a transparent framework for identifying conditions whose health impact is systematically understated by conventional mortality reporting.