Reframing disease burden: validation of DALY-per-case as a per-diagnosis severity metric.

Introduction

The global burden of disease, spanning communicable and non-communicable conditions, reached 2.8 billion DALYs in 2021, with projections of increase due to demographic shifts [ 1 ]. This underscores the need for complementary metrics that better quantify individual-level severity alongside population burden to guide prevention and research. Current indicators such as incidence, mortality, DALYs, and MIR, focus on population trends but often mask per-case severity across diverse diseases [ 2 ].

Established metrics like incidence and mortality highlight high-burden conditions [ 2 ], while DALYs integrate YLLs and YLDs [ 1 ]. MIR, calculated as mortality divided by incidence, proxies survival and healthcare performance but lacks YLD detail [ 3 ]. These population-weighted measures can obscure individual severity, necessitating a new approach.

Because these measures are population-weighted and dominated by high-prevalence conditions, per-case severity may be obscured [ 1 ]. As an illustration from oncology, rarer cancers can impose substantial individual burden despite modest aggregate DALYs [ 1 , 4 , 5 ]. MIR, while a crude survival proxy, omits disability detail and struggles to reflect prognosis heterogeneity [ 2 , 3 ]. Clinicians’ observed severity gradients (e.g., rapid vs. indolent courses) underscore the need for a standardized, prevalence-independent severity measure across diseases [ 6 ].

Building on prior uses in infectious disease epidemiology [ 7 , 8 ], this study validates DALY-per-case, defined as total DALYs divided by incidence, estimating per-diagnosis healthy life years lost by integrating mortality and disability. Although YLDs are commonly presented as prevalence-based quantities, the Global Burden of Disease (GBD) estimation framework links incidence, duration, remission, and mortality through compartmental disease modelling to ensure internal consistency between incidence and prevalence (9,30,36,31). Within this structure, DALYs are computed as the sum of YLLs and YLDs, representing the total health loss attributable to disease. Under these modelling assumptions, particularly in relatively stable epidemiologic conditions, total DALYs approximate the expected lifetime health loss attributable to incident cohorts. Consequently, dividing total DALYs by incident cases provides an estimate of the average expected healthy life-years lost per new diagnosis [ 9 – 11 ]. Unlike frequency-based incidence and mortality, prevalence-weighted DALYs, or MIR’s survival proxy, DALY-per-case quantifies individual burden, enabling severity comparisons across diseases.

Importantly, DALY-per-case is not introduced as a fundamentally new burden construct, but as a standardized per-incident normalization of established DALY components within the GBD framework. Whereas DALYs per capita reflect societal burden and indicators such as case-fatality risk (CFR) or MIR approximate lethality, DALY-per-case quantifies the average expected healthy life-years lost per new diagnosis at the population level by integrating both fatal (YLL/I) and non-fatal (YLD/I) components. Its contribution lies in formal validation, theoretical positioning within the GBD structure, and development of an interpretive framework that separates intrinsic per-diagnosis severity from overall population burden.

Comparative epidemiology increasingly requires metrics beyond infectious settings [ 12 , 13 ]. Traditional indicators like case-fatality rate or YLL per case focus on mortality, while DALYs, combining fatal and non-fatal outcomes, can still obscure per-case severity due to population-weighting [ 14 , 15 ]. This drives the need for individual-level measures, with DALY-per-case integrating YLD and YLL per diagnosis, aligning with GBD calls for unified survival-morbidity metrics across conditions [ 1 , 14 , 16 ].

The significance of DALY-per-case lies in complementing existing indicators, extending burden-of-disease concepts from foodborne and infectious diseases [ 17 , 18 ]. Epidemiologically, it ranks conditions by severity and impact on life expectancy and quality, highlighting high-severity, lower-prevalence entities that contribute disproportionate individual burden [ 19 , 20 ]. It can also track therapeutic progress sensitively, revealing per-case burden reductions obscured by aggregate DALYs [ 21 ]. Clinically, it aligns epidemiological data with clinicians’ severity observations, aiding research advocacy [ 22 ]. For policy, it identifies conditions with high lethality and disability [ 23 ].

By filling the gap in per-case severity quantification, DALY-per-case reframes epidemiology from a purely population-driven to a multidimensional perspective, integrating both prevalence and individual severity. This disease-agnostic approach has the potential to sharpen research priorities, align clinical and epidemiological perspectives, and inform more equitable strategies to address the projected rise in global disease burdens.

Methods

Study design

This population-based descriptive study evaluates DALY-per-case as a per-incident severity indicator defined as the average number of disability-adjusted life years (DALYs) lost per new diagnosis. The metric was computed globally across cancer sites, as a validation group, and compared with established measures including incidence, mortality, total DALYs, and the mortality-to-incidence ratio (MIR). DALY-per-case represents an aggregate, population-based estimate of expected health loss per incident case and does not imply measurement of individual patient-level severity.

Data sources

Both incidence and DALY estimates were obtained from the Global Burden of Disease (GBD) 2021 study conducted by the Institute for Health Metrics and Evaluation (IHME), (14). Incidence represented new cancer cases, and DALYs were the sum of years of life lost (YLLs) and years lived with disability (YLDs). For each cancer type, global totals (all ages, both sexes) for incidence, DALYs, YLLs, and YLDs with 95% uncertainty intervals (UIs) were extracted. Cancer categories followed GBD classifications aligned with ICD-10 definitions.

Statistical framework

Within the Global Burden of Disease (GBD) framework, disability-adjusted life years (DALYs) are defined as the sum of years of life lost (YLL) and years lived with disability (YLD): 1 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$DALY=YLL+YLD$$\end{document}

Years of life lost (YLL) are calculated as the number of deaths multiplied by the standard life expectancy at the age of death. 2 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$YLL=DXL$$\end{document}

where D denotes the number of deaths and L the standard life expectancy at the age of death.

Years lived with disability (YLD) are estimated within the GBD modeling framework using incidence- or prevalence-based approaches that incorporate duration and disability weights. 3 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$YLD=IXDWXDuration$$\end{document}

where I represents incident cases, DW the disability weight, and Duration the average time lived with the condition.

Based on these definitions, DALY-per-case is derived as a per-incident normalization of total DALYs: 4 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{D}\mathrm{A}\mathrm{L}\mathrm{Y}\mathrm{S}-\mathrm{p}\mathrm{e}\mathrm{r}-\mathrm{c}\mathrm{a}\mathrm{s}\mathrm{e}=\frac{DALY}{Incidence}=\frac{\left(YLL+YLD\right)}{Incidence}$$\end{document}

Total DALYs, YLL, and YLD are expressed in absolute years. Incidence is expressed as number of new cases. Accordingly, DALY-per-case is expressed in units of years lost per incident case. No rate standardization (per 100,000 population) was applied in the derivation of DALY-per-case, as the metric represents an average per diagnosis rather than a population rate.

This formulation permits decomposition into mortality and morbidity components: 5 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$DALY-case=YLL/I+YLD/I$$\end{document}

where “I” represents incident cases. Temporal change in DALY-per-case can therefore be expressed as: 6 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varDelta DALY-case=\varDelta\left(YLL/I\right)+\varDelta\left(YLD/I\right)$$\end{document}

enabling decomposition of per-case burden changes into survival-related versus morbidity-related contributions.

Uncertainty propagation

Uncertainty intervals for DALY-per-case were derived from published 95% uncertainty bounds for DALYs and incidence from GBD 2021. Because posterior draw-level data were not available for recomputation of derived ratios, conservative bound division was applied.

Lower bound of DALY-per-case: 7 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${DALY}_{case}^{L}=\frac{{DALY}^{L}}{{I}^{U}}$$\end{document}

Upper bound of DALY-per-case: 8 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${DALY}_{case}^{U}=\frac{{DALY}^{LU}}{{I}^{L}}$$\end{document}

where DALY L and, DALY U denote lower and upper bounds of total DALYs; I L and I U denote lower and upper bounds of incidence.

This method preserves conservative coverage and may yield wider intervals than simulation-based propagation using posterior draws.

Metric definition and scope of application

Using IHME GBD 2021 global totals (all ages, both sexes), DALY-per-case was computed across cancer sites as the average expected healthy life-years lost per incident diagnosis at the population level. The interpretation of DALY-per-case relies on the assumption that, within the GBD modeling framework, total DALYs approximate the expected lifetime health loss attributable to incident cases under relatively stable epidemiologic conditions. This assumption reflects the internal consistency of GBD’s compartmental modeling linking incidence, mortality, duration, and disability weights.

DALY-per-case is intended for comparative population-level epidemiology within the standardized GBD framework. It is appropriate for cross-disease severity ranking, temporal trend analysis, and policy prioritization at national or global scales. It does not represent individual-level clinical severity and should not be interpreted as a prognostic measure for specific patients. Application outside standardized GBD case definitions or in data-sparse settings requires cautious interpretation and, where possible, local validation.

Validation data

To assess construct and convergent validity, we used indicators from the SEER program (National Cancer Institute, USA). Age-adjusted incidence (2018–2022) and mortality (2019–2023) from SEER Explorer informed mortality-to-incidence ratios (MIR = mortality ÷ incidence). Five-year relative survival by cancer site was extracted from SEER [ 24 ], and when available, checked against CONCORD-3 [ 25 ]. Median age at diagnosis from SEER captured demographic variation across sites. These SEER-based parameters were aligned with global DALY and incidence estimates from IHME GBD 2021 to ensure consistency between burden- and survival-based severity frameworks. Validation used cancer as the primary exemplar due to availability of independent survival benchmarks enabling construct validation. To assess generalizability, five non-cancer diseases were selected to represent distinct epidemiologic and clinical patterns, including communicable and non-communicable conditions, acute and chronic trajectories, and varying mortality–morbidity compositions. This structured selection allowed evaluation of DALY-per-case performance across heterogeneous disease profiles within the GBD framework.

This disease-agnostic formulation extends prior work by introducing (i) a quadrant-based 2D framework with median thresholds for total DALYs and DALY-per-case to visualize/classify burdens, and (ii) regression models testing independence from MIR and survival. Validation used cancer as the exemplar condition given availability of survival benchmarks. To assess generalizability beyond oncology, five non-cancer diseases were selected to represent distinct epidemiologic and clinical profiles: tuberculosis (communicable, potentially curable infection with substantial YLL contribution), HIV/AIDS (chronic infectious disease with prolonged disability and mortality), ischemic heart disease (high-mortality non-communicable condition), type 2 diabetes mellitus (high-prevalence chronic metabolic disease with dominant YLD contribution), and Alzheimer’s disease and other dementias (neurodegenerative condition characterized by prolonged disability and lower acute mortality). This spectrum enabled evaluation of DALY-per-case performance across diseases with differing mortality-morbidity compositions.

Analysis

DALY-per-case was computed for major cancers as mean years lost per case with 95% UIs. Cancers were ranked by DALY-per-case and compared with rankings by incidence, mortality, and total DALYs. To improve visualization of the highly skewed distribution of population burden, total DALYs were log-transformed prior to plotting. Log-transformed total DALYs (population burden; x-axis) were plotted against DALY-per-case (individual severity; y-axis), yielding four quadrants: high-high, high-low, low-high, and low-low. Median-based thresholds (total DALYs: 10 million; DALY-per-case: 18) were derived empirically from the distribution and stress-tested (± 20%) for stability. For uncertainty interpretation and categorical comparison, severity bands were defined on the continuous DALY-per-case scale (years per incident case). Based on distributional clustering of observed values, three interval domains were specified: low severity (< 10 years per case), moderate severity (10–20 years per case), and high severity (> 20 years per case). These thresholds were empirically derived to facilitate interpretability rather than reflect externally imposed clinical cutoffs, consistent with established methodological approaches for categorizing continuous epidemiologic measures when no predefined severity standards exist [ 26 , 27 ].

Stability classification was determined by the position of the 95% uncertainty interval (UI) relative to these predefined interval domains. A cancer was classified as “stable” within a severity category when the point estimate and most of the UI remained within a single interval domain. Cancers were classified as “variable” when the UI substantially crossed adjacent severity thresholds, indicating instability in categorical placement. Interpretation of uncertainty intervals followed established Global Burden of Disease methodological standards [ 9 , 11 ]. Relative UI width was calculated as (upper-lower)/mean and expressed as percent variability; across cancers, relative UI widths.

Construct validity was assessed using Pearson’s correlation between DALY-per-case and 5-year relative survival (SEER). Because DALY-per-case values were right-skewed, logarithmic transformation was applied prior to correlation analysis to better approximate normality and satisfy parametric assumptions.

To assess convergent validity and redundancy with mortality-based indices, we fit multiple linear regressions with DALY-per-case as the outcome and MIR, 5-year survival, and median age at diagnosis as predictors. Multicollinearity was evaluated via variance inflation factors (VIFs). We then ran partial correlations to test independence: (a) DALY-per-case with survival controlling for MIR, and (b) DALY-per-case with MIR controlling for survival.

Results

Two-dimensional burden analysis

A scatter plot of DALY-per-case vs. total DALYs classified 34 cancers into four quadrants via median thresholds for population burden (10 million total DALYs; x-axis) and individual severity (18 DALYs/case; y-axis) (Fig. 1 ). The upper-right quadrant (high burden-high severity) included tracheal, bronchus, and lung cancer (46.5 million; 20.4), stomach (22.8 million; 18.5), liver (12.9 million; 24.4), and pancreatic (11.3 million; 22.3). The lower-right quadrant (high burden-moderate severity, 6–11 DALYs/case) comprised breast (20.6 million; 9.7), colorectal (24.4 million; 11.1), and prostate (8.1 million; 6.1). The upper-left quadrant (low burden < 10 million; high severity) included brain and central nervous system (CNS) cancers (8.9 million; 24.9), leukemia (10.9 million; 23.8), malignant neoplasms of bone and articular cartilage (2.5 million; 27.6), and mesothelioma (0.7 million; 21.6). The lower-left quadrant (low burden; low severity < 6 DALYs/case) included thyroid (1.2 million; 5.0), testicular (0.6 million; 6.1), and non-melanoma skin cancers (1.2 million; 0.2). This quadrant-based 2D framework reveals divergences in disease burdens, illustrated by brain/CNS (high severity, low population impact and extends as a visualization toolkit for comparative epidemiology across conditions, as shown for tuberculosis (5.6 DALYs/case) and diabetes (3.2).

Cancer rankings by burden and severity

Cancers were ranked by total DALYs and by DALY-per-case to clarify population vs. individual burden (Table 1 ). By total DALYs, tracheal/bronchus/lung ranked 1 st (46.5 M), followed by colorectal (24.4 M), stomach (22.8 M), and breast (20.6 M), reflecting high incidence and mortality. In contrast, DALY-per-case emphasized per-case severity: malignant neoplasms of bone/articular cartilage (27.6; rank 1), neuroblastoma/other peripheral nervous cell tumors (26.3; rank 2), and brain/CNS (24.9; rank 3), despite lower total-DALY positions (23rd, 34th, 10th). Notable divergences included breast (4th by total DALYs; 26th by DALY-per-case) and prostate (11th; 28th), indicating high population burden but lower per-case severity, versus brain/CNS (10th; 3rd), indicating high individual severity. Lung (1st in total DALYs; 10th in DALY-per-case), liver (6th; 4th), pancreatic (7th; 7th), and stomach (3rd; 11th) ranked highly across both dimensions, underscoring dual burden.

Fig. 1 Relationship between population-level burden and individual-level severity of cancers

Scatterplot of total disability-adjusted life years (DALYs; x-axis) versus DALY-per-case (y-axis) for major cancer sites, based on IHME GBD 2021 estimates. Each point represents a cancer type, positioned according to its overall global DALY burden (population impact) and average per-case severity (years of healthy life lost per new diagnosis. The plot is divided into four quadrants: (i) high burden-high severity such lung, liver, pancreas, stomach, (ii) high burden-low severity such as breast, colorectal, prostate, (iii) low burden-high severity such as brain/CNS, bone, mesothelioma, and (iv) low burden-low severity such thyroid, testicular, non-melanoma skin cancer.

Table 1 Cancer rankings by total DALYs and DALY-per-case (IHME GBD 2021) Cancer Total DALYs (millions) DALY-per-case Rank by total DALYs Rank by DALY-per-case Tracheal, bronchus, lung 46.5 20.4 1 10 Colorectal 24.4 11.1 2 24 Stomach 22.8 18.5 3 11 Breast 20.6 9.7 4 26 Esophageal 13.0 22.5 5 6 Liver 12.9 24.4 6 4 Pancreatic 11.3 22.3 7 7 Leukemia 10.9 23.8 8 5 Cervical 9.9 14.9 9 20 Brain/CNS 8.9 24.9 10 3 Prostate 8.1 6.1 11 28 Non-Hodgkin lymphoma 7.8 12.8 12 23 Other malignant neoplasms 6.5 15.3 13 19 Lip/oral cavity 5.9 13.9 14 21 Ovarian 5.2 17.3 15 15 Bladder 4.4 8.1 16 27 Kidney 4.0 10.4 17 25 Gallbladder/biliary 3.7 17.2 18 16 Larynx 3.1 15.6 19 18 Other pharynx 2.8 16.7 20 17 Multiple myeloma 2.6 17.4 21 13 Uterine 2.6 5.4 22 31 Bone/articular cartilage 2.5 27.6 23 1 Nasopharynx 2.5 20.9 24 9 Malignant skin melanoma 1.7 5.5 25 30 Soft tissue sarcomas 1.7 17.4 26 14 Thyroid 1.2 5.0 27 32 Other neoplasms 1.2 0.0 28 34 Non-melanoma skin 1.2 0.2 29 33 Hodgkin lymphoma 1.2 18.4 30 12 Mesothelioma 0.7 21.6 31 8 Testicular 0.6 6.1 32 29 Eye 0.5 13.8 33 22 Neuroblastoma 0.3 26.3 34 2

Uncertainty analysis

Incorporation of 95% uncertainty intervals (UIs) confirmed the stability of DALY-per-case rankings (Table 2 ). High-severity cancers, including malignant neoplasms of bone and articular cartilage (DALYS-per-case 27.6 years; 95% uncertainty interval (UI): 20.1–38.5), neuroblastoma (26.3; 16.8–41.2), and brain/CNS cancers (24.9; 18.7–33.4), consistently ranked among the top across UI ranges. Low-severity cancers, such as thyroid (DALYS-per-case 5.0 years; uncertainty interval (UI): 4.0–6.2.0.2), prostate (6.1; 5.1–7.2), and breast (9.7; 8.5–11.1), remained stable at the lower end. Mesothelioma (21.6; 18.7–25.0) and pancreatic cancer (22.3; 19.1–26.3) also maintained high-severity positions. Cancers with wider UIs, such as neuroblastoma and Hodgkin lymphoma (18.4; 11.0–28.8.0.8), showed variability but retained high severity. These results demonstrate the robustness of DALY-per-case rankings to statistical uncertainty.

Table 2 DALY-per-case with 95% Uncertainty Intervals (IHME GBD 2021) Cancer DALY’s/incidence Lower bound Upper bound Stability rank Bladder cancer 8.1 7.0 9.7 Stable (low) Brain and central nervous system cancer 24.9 18.7 33.4 Stable (high) Breast cancer 9.7 8.5 11.1 Stable (low) Cervical cancer 14.9 12.5 17.6 Stable (mid) Colon and rectum cancer 11.1 9.6 13.1 Stable (low-mid) Esophageal cancer 22.5 17.8 28.7 Stable (high) Eye cancer 13.8 8.7 21.3 Variable (wide UI) Gallbladder and biliary tract cancer 17.2 12.7 23.7 Stable (mid-high) Hodgkin lymphoma 18.4 11.0 28.8 Variable (moderate-high) Kidney cancer 10.4 9.4 11.6 Stable (low) Larynx cancer 15.6 13.5 18.1 Stable (mid) Leukemia 23.8 17.9 30.9 Stable (high) Lip and oral cavity cancer 13.9 11.8 16.3 Stable (mid) Liver cancer 24.4 19.7 30.1 Stable (high) Malignant bone & cartilage 27.6 20.1 38.5 Stable (high) Malignant skin melanoma 5.5 4.6 6.5 Stable (low) Mesothelioma 21.6 18.7 25.0 Stable (high) Multiple myeloma 17.4 14.0 21.9 Stable (mid-high) Nasopharynx cancer 20.9 16.3 26.5 Stable (high) Neuroblastoma 26.3 16.8 41.2 Variable (but high) Non-Hodgkin lymphoma 12.8 11.0 15.2 Stable (mid) Non-melanoma skin cancer 0.2 0.2 0.2 Stable (very low) Other malignant neoplasms 15.3 12.3 18.6 Stable (mid) Other neoplasms 0.0 0.0 0.0 Stable (very low) Other pharynx cancer 16.7 14.6 19.2 Stable (mid) Ovarian cancer 17.3 14.5 20.7 Stable (mid–high) Pancreatic cancer 22.3 19.1 26.3 Stable (high) Prostate cancer 6.1 5.1 7.2 Stable (low) Soft tissue sarcomas 17.4 12.3 25.4 Variable (mid-high) Stomach cancer 18.5 13.9 24.8 Stable (mid-high) Testicular cancer 6.1 5.5 6.8 Stable (low) Thyroid cancer 5.0 4.0 6.2 Stable (low) Tracheal, bronchus, lung cancer 20.4 16.7 24.8 Stable (high) Uterine cancer 5.4 4.5 6.6 Stable (low)

Correlation with 5-year survival

To assess alignment with cancer prognosis, DALY-per-case estimates were compared with 5-year relative survival rates from population-based cancer registries (SEER) across available cancer types, excluding categories lacking survival data (other malignant neoplasms, other neoplasms, other pharynx cancer, neuroblastoma) (Table 3 ). A strong inverse Pearson correlation was observed between DALY-per-case and 5-year survival ( r = − 0.72) (Fig. 2 ), indicating that cancers associated with lower survival tended to incur higher average healthy life-years lost per diagnosis.

Cancers with poor survival, such as pancreatic cancer (22.3 DALYs per case; 13.5% survival), esophageal cancer (22.5; 23.5%), liver cancer (24.4; 22.8%), and lung cancer (20.4; 32.6%), exhibited high DALY-per-case values. Conversely, cancers with favorable survival, including thyroid cancer (5.0; 98.9%), prostate cancer (6.1; 97.9%), testicular cancer (6.1; 94.9%), and malignant skin melanoma (5.5; 96.1%), showed lower DALY-per-case values.

Table 3 DALY-per-case and 5-Year Relative Survival Cancer DALYS per case 5 years survival Bladder cancer 8.1 78 Brain and central nervous system cancer 24.9 33 Breast cancer 9.7 91.7 Cervical cancer 14.9 68 Colon and rectum cancer 11.1 65.4 Esophageal cancer 22.5 23.5 Eye cancer 13.8 85.3 Gallbladder and biliary tract cancer 17.2 21 Hodgkin lymphoma 18.4 90.5 Kidney cancer 10.4 79.9 Larynx cancer 15.6 59.2 Leukemia 23.8 68.2 Lip and oral cavity cancer 13.9 70.6 Liver cancer 24.4 22.8 Malignant neoplasm of bone and articular cartilage 27.6 71.2 Malignant skin melanoma 5.5 96.1 Mesothelioma 21.6 23.1 Multiple myeloma 17.4 62.4 Nasopharynx cancer 20.9 65 Neuroblastoma and other peripheral nervous cell tumors 26.3 N/A Non-Hodgkin lymphoma 12.8 74.2 Non-melanoma skin cancer 0.2 95 Other malignant neoplasms 15.3 N/A Other neoplasms 0.0 N/A Other pharynx cancer 16.7 N/A Ovarian cancer 17.3 51.6 Pancreatic cancer 22.3 13.5 Prostate cancer 6.1 97.9 Soft tissue and other extraosseous sarcomas 17.4 66.9 Stomach cancer 18.5 43.2 Testicular cancer 6.1 94.9 Thyroid cancer 5.0 98.9 Tracheal, bronchus, and lung cancer 20.4 32.6 Uterine cancer 5.4 68

Fig. 2 Relationship between DALY-per-case and 5-year relative survival across cancer sites

Scatterplot of DALY-per-case (x-axis) against 5-year relative survival (y-axis) for 30 cancer types, based on IHME GBD 2021 estimates for DALYs and population-based registry survival. Each point represents one cancer site. A strong inverse association is visible, Pearson correlation coefficient ( r = −0.72), with highly lethal cancers such as pancreas, liver, esophagus, and lung clustering at the lower right (high DALY-per-case, low survival), while favorable-prognosis cancers such as thyroid, prostate, testicular, and melanoma appear at the upper left (low DALY-per-case, high survival). Intermediate cancers occupy central positions, highlighting the gradient between individual severity and prognosis.

Cross-disease illustration

Supplementary analyses computed DALY-per-case for selected non-cancer conditions to assess generalizability, including tuberculosis (5.2 DALYs per case, 95% UI: 3.76–7.22), HIV/AIDS (21.9, 95% UI: 17.45–27.09), ischemic heart disease (15.1, 95% UI: 11.9–18.9), Alzheimer’s disease and other dementias (3.8, 95% UI: 1.97–7.02), and type 2 diabetes mellitus (3.5, 95% UI: 3.18–3.95).

In relation to cancer severity patterns, HIV/AIDS (21.9) approached the per-case severity levels observed in high-severity cancers such as pancreatic and liver cancer, aligning with the high-severity quadrant of the 2D framework. Ischemic heart disease (15.1) clustered within the moderate severity range, comparable to colorectal and ovarian cancers. In contrast, type 2 diabetes (3.5) and Alzheimer’s disease (3.8) demonstrated per-case severity levels similar to lower-severity cancers such as thyroid, melanoma, and prostate, positioning them within the lower-severity quadrants despite substantial population burden. Tuberculosis (5.2) also aligned with lower-severity ranges on a per-case basis. (Table S1).

Multivariate and partial correlation validation

To assess convergent validity and potential redundancy with established mortality-based indicators, DALY-per-case was regressed on 5-year survival, median age at diagnosis, and mortality-to-incidence ratio (MIR). The overall model was statistically significant (F(3,24) = 11.98, p < 0.001) and explained 60.0% of the variance (R² = 0.60; adjusted R² = 0.55).

In the fully adjusted model (Table 4 ), MIR demonstrated the largest standardized coefficient (β = 0.52) and approached statistical significance ( p = 0.06), whereas 5-year survival ( p = 0.20) and median age ( p = 0.23) were not independently significant after adjustment. These findings indicate that DALY-per-case is strongly influenced by mortality burden but is not entirely explained by survival or age structure.

Reduced and partial correlation analyses further showed attenuation of the survival association after controlling for MIR, suggesting shared variance between mortality-based indicators and DALY-per-case. However, the persistence of residual variance in the multivariable model indicates that DALY-per-case captures additional information beyond MIR alone, consistent with its integration of both fatal (YLL) and non-fatal (YLD) components.

Table 4 Multiple linear regression analysis of DALY-per-case with 5-year survival, median age at diagnosis, and MIR across 28 cancer sites Predictor B SE β T p VIF Constant 23.40 9.99 -- 2.34 0.028 -- 5-year survival (%) − 0.090 0.069 − 0.357 − 1.31 0.202 4.43 Median age at diagnosis (years) − 0.124 0.100 − 0.185 − 1.24 0.228 1.34 Mortality-to-incidence ratio (MIR) + 15.08 7.66 + 0.517 1.97 0.061 4.14 Model statistics: R = 0.774 R² = 0.600 Adjusted R² = 0.549 Std. Error = 4.48 F (3, 24) = 11.98, p < 0.001. DALY-per-case = mean disability-adjusted life years lost per new cancer diagnosis; MIR = mortality-to-incidence ratio. Coefficients are unstandardized (B) with standardized (β) estimates reported for comparison. Model includes 28 cancer sites (both sexes combined)

Discussion

This study demonstrates that DALY-per-case robustly quantifies per-diagnosis severity, showing a strong inverse correlation with 5-year survival ( r = −0.72), meaning that cancers associated with poorer prognosis correspond to substantially higher average healthy life-years lost per new diagnosis. This alignment with survival gradients supports construct validity, indicating that DALY-per-case behaves consistently with established prognostic patterns while additionally incorporating morbidity burden. Multivariable analyses further demonstrated partial independence from the mortality-to-incidence ratio (MIR), validated across 34 cancer types. Extension to tuberculosis, diabetes, HIV/AIDS, ischemic heart disease, and Alzheimer’s disease within the validated 2D burden-severity framework supports its applicability across diverse diseases within the standardized GBD framework.

It is important to clarify that DALY-per-case is not proposed as a new burden construct, but as a standardized per-incident normalization of established DALY components within the GBD framework. Its contribution lies in formal validation against survival benchmarks, demonstration of partial independence from MIR, and the development of an interpretive framework separating intrinsic per-diagnosis severity from overall population burden. Unlike case-fatality risk (CFR) or MIR, which approximate lethality, and unlike DALYs per capita, which reflect societal burden, DALY-per-case integrates both fatal (YLL/I) and non-fatal (YLD/I) components to quantify expected healthy life-years lost per new diagnosis.

Epidemiologically, DALY-per-case distinguishes high-prevalence conditions with moderate per-case severity from high-severity conditions with disproportionate individual burden, insight not obtainable from incidence or survival alone [ 21 , 28 ]. Compared with existing indicators, DALY-per-case offers clear advantages, unlike MIR, criticized for weak theoretical grounding and sensitivity to registry completeness [ 3 ], it rests on YLL + YLD and is consistently estimable across countries via GBD data [ 4 , 21 , 28 , 29 ]. Case-fatality and median survival, though clinically intuitive, are often unavailable outside high-income registries; DALY-per-case enables severity comparisons in data-limited settings, aligning with WHO equity priorities [ 23 ]. By incorporating YLD, it captures morbidity among survivors, effects often missed by survival endpoints [ 19 , 30 ]. Across epidemiology, clinical practice, and policy, DALY-per-case supplies a standardized severity scale for cross-disease and cross-regional comparisons and supports burden decomposition to distinguish incidence-driven changes from per-case severity shifts [ 4 , 21 , 28 ].

In clinical practice, from a methodological perspective, DALY-per-case bridges classical measures of case fatality and disability weighting, providing a per-incident metric consistent with GBD’s standardized structure. It quantifies the combined loss of life quantity and quality per diagnosis, complementing survival endpoints such as overall and progression-free survival in oncology, by flagging conditions where therapeutic advances could yield substantial per-patient gains [ 4 , 19 ]. It also highlights survivorship and long-term morbidity such as cardiotoxicity, chronic pain, requiring proactive follow-up [ 19 , 30 ]. In policy, pairing total DALYs (population burden) with DALY-per-case (individual severity) in a 2D framework categorizes conditions into high-burden/high-severity, high-burden/low-severity, low-burden/high-severity, and low-burden/low-severity groups, operationalizing WHO’s call to balance broad-reach interventions with targeted efforts for high-severity conditions [ 4 , 23 , 31 ]. Extension beyond cancer illustrates the metric’s disease-agnostic applicability. HIV/AIDS demonstrated per-case severity within the high-severity range (> 20 DALYs per case), comparable to intrinsically severe cancers such as pancreatic and liver cancer and positioning it within the high-severity quadrant of the 2D framework [ 10 , 32 ]. In contrast, type 2 diabetes and Alzheimer’s disease exhibited lower per-case severity (< 10 DALYs per case), aligning with lower-severity cancer categories and occupying the low-severity quadrants despite substantial total population burden [ 32 ]. Ischemic heart disease occupied an intermediate severity range (10–20 DALYs per case), corresponding to moderate-severity placements within the framework [ 32 ]. These patterns reinforce the distinction between population burden and per-diagnosis severity and demonstrate consistent interpretability across communicable and non-communicable diseases.

Interpretation in low- and middle-income countries (LMICs) requires caution. Elevated DALY-per-case values may reflect delayed diagnosis, limited treatment access, and systemic healthcare constraints rather than intrinsic biological severity alone [ 4 , 23 ]. In addition, GBD estimates in data-sparse settings rely more heavily on statistical modelling where registry coverage is incomplete, potentially widening uncertainty intervals. Accordingly, DALY-per-case should be interpreted cautiously for subnational resource allocation and ideally validated using country-specific survival, stage-at-diagnosis, or registry-linked data before informing local policy decisions [ 4 , 23 ].

Integrating DALY-per-case into comparative risk assessment and burden decomposition clarifies temporal shifts in disease severity [ 16 , 21 , 28 ]. As a per-diagnosis complement to population-weighted measures (total DALYs, MIR) [ 3 ],, it helps distinguish incidence-driven changes from true reductions in individual burden, refining trend interpretation across regions and time [ 4 , 31 ]. The metric is mathematically compatible with GBD procedures for YLL and YLD [ 10 ], and with the standardized disability-weight system [ 14 ]. Thus, DALY-per-case serves as a unifying per-incident metric for comparative epidemiology, enabling decomposition of mortality and morbidity across diseases and periods [ 21 , 28 , 31 ].

Statistical validation shows DALY-per-case quantifies a distinct severity construct integrating fatal and non-fatal outcomes. Its strong inverse correlation with 5-year survival and positive association with MIR establish convergent validity. Multivariable and partial-correlation analyses indicate it is not a surrogate for MIR or survival: when both entered, only MIR was borderline significant, and the DALY-per-case/MIR association persisted after adjusting for survival, supporting partial independence and discriminant validity. This mirrors comparative epidemiology, where composite burden measures combining mortality and morbidity outperform single-dimension indicators in explaining disease impact [ 11 ].

Although DALY-per-case correlates with the mortality-to-incidence ratio (MIR), the two measures differ conceptually. MIR estimates the probability of death among newly diagnosed cases and has been criticized for sensitivity to registry completeness and limited theoretical grounding as a survival proxy [ 3 , 33 ]. In contrast, DALY-per-case integrates both fatal (YLL/I) and non-fatal (YLD/I) components within the GBD framework [ 9 , 10 , 32 , 34 ], providing a continuous, health-adjusted severity scale rather than a lethality ratio. Its persistence after adjustment for survival indicates that it is not redundant with MIR but extends it by incorporating morbidity among survivors.

From an interpretive standpoint, these results position DALY-per-case between survival and population burden. Its close alignment with survival validates it as a prognostic proxy, while residual independence from MIR shows added information from disability weighting. This dual correspondence supports DALY-per-case as a composite severity index for both high- and low-data settings. Prior work has called for per-incident indicators to disentangle gains in survival from reductions in morbidity [ 4 ]. DALY-per-case advances measurement by providing a unified severity scale, harmonized with GBD architecture and empirically calibrated against survival, bridging clinical outcomes and population health metrics.

Policy implications

From a policy perspective, DALY-per-case complements population-weighted measures by distinguishing intrinsic per-diagnosis severity from aggregate burden. While total DALYs prioritize high-prevalence conditions, per-case normalization highlights diseases associated with disproportionate individual health loss [ 9 , 32 ]. Pairing total DALYs with DALY-per-case aligns with WHO priority-setting principles that balance efficiency with severity considerations [ 9 , 35 ]. In LMICs, interpretation requires caution because modelled inputs and healthcare access gaps may influence estimates; local validation is recommended before subnational resource allocation.

Limitations

The principal limitation of DALY-per-case is that it is derived from modeled population-level DALY estimates within the GB) framework rather than from individual-level longitudinal clinical data. Accordingly, it reflects average expected health loss per diagnosis at the population level and does not represent patient-specific clinical severity. DALY estimates depend on modeled mortality rates, disability weights, disease duration assumptions, and adjustments for under-ascertainment [ 9 , 36 , 37 ], and DALY-per-case necessarily inherits these assumptions.

Age structure may influence per-case estimates. Conditions affecting younger populations generate larger years of life lost (YLL) contributions due to higher remaining life expectancy, potentially elevating DALY-per-case relative to late-onset diseases. Additionally, disability weights may incompletely capture psychosocial or context-specific burdens, particularly in low-resource settings [ 38 ].

Cross-disease comparability is conditional rather than absolute. Although the GBD framework standardizes case definitions, duration modeling, and disability weights across conditions [ 32 ], communicable and non-communicable diseases differ in natural history, surveillance completeness, and case ascertainment. In low- and middle-income countries, greater reliance on modeled inputs may widen uncertainty intervals and influence per-case estimates. Therefore, DALY-per-case should be interpreted as a standardized comparative metric within the GBD architecture rather than a universal biological severity constant.

Finally, the ecological design limits patient-level inference. However, the strong correlation with survival supports construct validity at the population level.

Conclusion

DALY-per-case is a standardized per-incident severity indicator that complements traditional population-weighted burden metrics. It aligns with survival gradients, integrates fatal and non-fatal components, and enables structured differentiation between intrinsic per-diagnosis severity and aggregate population burden.

Implications for future work

Future research should validate DALY-per-case using patient-level longitudinal cohorts, explore age-standardized per-case formulations, and assess performance in subnational and registry-linked datasets. Simulation-based uncertainty propagation using GBD posterior draws may further refine interval estimation. Application of DALY-per-case in time-series analyses could also help distinguish therapeutic improvements in survival from reductions in morbidity burden. Integration into national reporting systems may enable real-time monitoring of per-diagnosis severity trends across diseases.

Supplementary Information

Supplementary Material 1.

Supplementary Material 2.

Supplementary Material 3.

Supplementary Material 4.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Author contributions

Single author paper.

Funding

No fund was received for this work.

Data availability

The data used in this work is attached in the supplementary. Its also available here:1. https://www.healthdata.org/2. https://seer.cancer.gov/.

Declarations

Competing interests

The authors declare no competing interests.

References

1. GBD 2021, Global Burden of Disease Study 2021 (GBD 2021) Results, Seattle, WA, 2021. Accessed: Mar. 21, 2025. [Online]. Available: https://vizhub.healthdata.org/gbd-results

2. Bray F Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries CA Cancer J Clin 2024 74 3 229 263 10.3322/caac.21834 38572751 Bray F, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. 10.3322/caac.21834. 38572751 10.3322/caac.21834

3. Asadzadeh Vostakolaei F Karim-Kos HE Janssen-Heijnen MLG Visser O Verbeek ALM Kiemeney LALM The validity of the mortality to incidence ratio as a proxy for site-specific cancer survival Eur J Public Health 2011 21 5 573 577 10.1093/eurpub/ckq120 20813895 Asadzadeh Vostakolaei F, Karim-Kos HE, Janssen-Heijnen MLG, Visser O, Verbeek ALM, Kiemeney LALM. The validity of the mortality to incidence ratio as a proxy for site-specific cancer survival. Eur J Public Health. 2011;21(5):573–7. 10.1093/eurpub/ckq120. 20813895 10.1093/eurpub/ckq120

4. Kocarnik JM Cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life years for 29 cancer groups from 2010 to 2019 a systematic analysis for the global burden of disease study 2019 JAMA Oncol. 2022 8 3 420 444 10.1001/jamaoncol.2021.6987 34967848 PMC8719276 Kocarnik JM, et al. Cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life years for 29 cancer groups from 2010 to 2019 a systematic analysis for the global burden of disease study 2019. JAMA Oncol. 2022;8(3):420–44. 10.1001/jamaoncol.2021.6987. 34967848 10.1001/jamaoncol.2021.6987 PMC8719276

5. Ferlay J Cancer incidence and mortality patterns in Europe: estimates for 40 countries and 25 major cancers in 2018 Eur. J. Cancer 2018 103 356 387 10.1016/j.ejca.2018.07.005 30100160 Ferlay J, et al. Cancer incidence and mortality patterns in Europe: estimates for 40 countries and 25 major cancers in 2018. Eur J Cancer. 2018;103:356–87. 10.1016/j.ejca.2018.07.005. 30100160 10.1016/j.ejca.2018.07.005

6. Ostrom QT Patil N Cioffi G Waite K Kruchko C Barnholtz-Sloan JS CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2013–2017 Neuro-Oncol 2020 22 Supplement_1 iv1 iv96 10.1093/neuonc/noaa200 33123732 PMC7596247 Ostrom QT, Patil N, Cioffi G, Waite K, Kruchko C, Barnholtz-Sloan JS. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2013–2017. Neuro-Oncol. 2020;22(Supplement_1):iv1–96. 10.1093/neuonc/noaa200. 33123732 10.1093/neuonc/noaa200 PMC7596247

7. Oliveira CC A systematic review of studies that estimated the burden of chronic non-communicable rare diseases using disability-adjusted life years Orphanet J. Rare Dis. 2024 19 1 333 10.1186/s13023-024-03342-3 39252105 PMC11384705 Oliveira CC, et al. A systematic review of studies that estimated the burden of chronic non-communicable rare diseases using disability-adjusted life years. Orphanet J Rare Dis. 2024;19(1):333. 10.1186/s13023-024-03342-3. 39252105 10.1186/s13023-024-03342-3 PMC11384705

8. Smith MP Estimating total morbidity burden of COVID-19: relative importance of death and disability J Clin Epidemiol 2022 142 54 59 10.1016/j.jclinepi.2021.10.018 34715312 PMC8547965 Smith MP. Estimating total morbidity burden of COVID-19: relative importance of death and disability. J Clin Epidemiol. 2022;142:54–9. 10.1016/j.jclinepi.2021.10.018. 34715312 10.1016/j.jclinepi.2021.10.018 PMC8547965

9. Murray CJL GBD 2010: design, definitions, and metrics Lancet 2012 380 9859 2063 2066 10.1016/S0140-6736(12)61899-6 23245602 Murray CJL, et al. GBD 2010: design, definitions, and metrics. Lancet. 2012;380(9859):2063–6. 10.1016/S0140-6736(12)61899-6. 23245602 10.1016/S0140-6736(12)61899-6

10. Kyu HH Global, regional, and national disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017 Lancet 2018 392 10159 1859 1922 10.1016/S0140-6736(18)32335-3 30415748 PMC6252083 Kyu HH, et al. Global, regional, and national disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392(10159):1859–922. 10.1016/S0140-6736(18)32335-3. 30415748 10.1016/S0140-6736(18)32335-3 PMC6252083

11. Vos T Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet 2020 396 10258 1204 1222 10.1016/S0140-6736(20)30925-9 33069326 PMC7567026 Vos T, et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):1204–22. 10.1016/S0140-6736(20)30925-9. 33069326 10.1016/S0140-6736(20)30925-9 PMC7567026

12. Murray CJL Findings from the Global Burden of Disease Study 2021 Lancet 2024 403 10440 2259 2262 10.1016/S0140-6736(24)00769-4 38762327 Murray CJL. Findings from the Global Burden of Disease Study 2021. Lancet. 2024;403(10440):2259–62. 10.1016/S0140-6736(24)00769-4. 38762327 10.1016/S0140-6736(24)00769-4

13. Murray CJ Lopez AD Jamison DT The global burden of disease in 1990: summary results, sensitivity analysis and future directions Bull World Health Organ 1994 72 3 495 8062404 PMC2486716 Murray CJ, Lopez AD, Jamison DT. The global burden of disease in 1990: summary results, sensitivity analysis and future directions. Bull World Health Organ. 1994;72(3):495. 8062404 PMC2486716

14. Salomon JA Disability weights for the Global Burden of Disease 2013 study Lancet Glob Health 2015 3 11 e712 e723 10.1016/S2214-109X(15)00069-8 26475018 Salomon JA, et al. Disability weights for the Global Burden of Disease 2013 study. Lancet Glob Health. 2015;3(11):e712–23. 10.1016/S2214-109X(15)00069-8. 26475018 10.1016/S2214-109X(15)00069-8

15. Arnesen T Nord E The value of DALY life: problems with ethics and validity of disability adjusted life years BMJ 1999 319 7222 1423 10.1136/bmj.319.7222.1423 10574867 PMC1117148 Arnesen T, Nord E. The value of DALY life: problems with ethics and validity of disability adjusted life years. BMJ. 1999;319(7222):1423. 10.1136/bmj.319.7222.1423. 10574867 10.1136/bmj.319.7222.1423 PMC1117148

16. Murray CJL Findings from the global burden of disease study 2021 Lancet 2024 403 10440 2259 62 10.1016/S0140-6736(24)00769-4 38762327 Murray CJL. Findings from the global burden of disease study 2021. Lancet. 2024;403(10440):2259–62. 38762327 10.1016/S0140-6736(24)00769-4

17. Havelaar AH World Health Organization Global estimates and regional comparisons of the burden of foodborne disease in 2010 PLoS Med 2015 12 12 e1001923- 10.1371/journal.pmed.1001923 26633896 PMC4668832 Havelaar AH, et al. World Health Organization Global estimates and regional comparisons of the burden of foodborne disease in 2010. PLoS Med. 2015;12(12):e1001923-. 10.1371/journal.pmed.1001923. 26633896 10.1371/journal.pmed.1001923 PMC4668832

18. Haagsma JA Polinder S Stein CE Havelaar AH Systematic review of foodborne burden of disease studies: quality assessment of data and methodology Int J Food Microbiol 2013 166 1 34 47 10.1016/j.ijfoodmicro.2013.05.029 23827806 Haagsma JA, Polinder S, Stein CE, Havelaar AH. Systematic review of foodborne burden of disease studies: quality assessment of data and methodology. Int J Food Microbiol. 2013;166(1):34–47. 10.1016/j.ijfoodmicro.2013.05.029. 23827806 10.1016/j.ijfoodmicro.2013.05.029

19. Rawla P Sunkara T Gaduputi V Epidemiology of pancreatic cancer: global trends, etiology and risk factors World J Oncol 2019 10 1 10 27 10.14740/wjon1166 30834048 PMC6396775 Rawla P, Sunkara T, Gaduputi V. Epidemiology of pancreatic cancer: global trends, etiology and risk factors. World J Oncol. 2019;10(1):10–27. 10.14740/wjon1166. 30834048 10.14740/wjon1166 PMC6396775

20. Arnold M Progress in cancer survival, mortality, and incidence in seven high-income countries 1995–2014 (ICBP SURVMARK-2): a population-based study Lancet Oncol 2019 20 11 1493 1505 10.1016/S1470-2045(19)30456-5 31521509 PMC6838671 Arnold M, et al. Progress in cancer survival, mortality, and incidence in seven high-income countries 1995–2014 (ICBP SURVMARK-2): a population-based study. Lancet Oncol. 2019;20(11):1493–505. 10.1016/S1470-2045(19)30456-5. 31521509 10.1016/S1470-2045(19)30456-5 PMC6838671

21. Ferrari AJ Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021 Lancet 2024 403 10440 2133 2161 10.1016/S0140-6736(24)00757-8 38642570 PMC11122111 Ferrari AJ, et al. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet. 2024;403(10440):2133–61. 10.1016/S0140-6736(24)00757-8. 38642570 10.1016/S0140-6736(24)00757-8 PMC11122111

22. Arem H Mama SK Duan X Rowland JH Bellizzi KM Ehlers DK Prevalence of healthy behaviors among cancer survivors in the United States: how far have we come? Cancer Epidemiol Biomarkers Prev 2020 29 6 1179 1187 10.1158/1055-9965.EPI-19-1318 32409489 PMC7778877 Arem H, Mama SK, Duan X, Rowland JH, Bellizzi KM, Ehlers DK. Prevalence of healthy behaviors among cancer survivors in the United States: how far have we come? Cancer Epidemiol Biomarkers Prev. 2020;29(6):1179–87. 10.1158/1055-9965.EPI-19-1318. 32409489 10.1158/1055-9965.EPI-19-1318 PMC7778877

23. Tran KB The global burden of cancer attributable to risk factors, 2010–19: a systematic analysis for the Global Burden of Disease Study 2019 Lancet 2022 400 10352 563 591 10.1016/S0140-6736(22)01438-6 35988567 PMC9395583 Tran KB, et al. The global burden of cancer attributable to risk factors, 2010–19: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2022;400(10352):563–91. 10.1016/S0140-6736(22)01438-6. 35988567 10.1016/S0140-6736(22)01438-6 PMC9395583

24. Surveillance Research Program NCI. SEER* Explorer: an interactive website for SEER cancer statistics. Natl Cancer Inst, 2021.

25. Allemani C Global surveillance of trends in cancer survival 2000–14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries Lancet 2018 391 10125 1023 1075 10.1016/S0140-6736(17)33326-3 29395269 PMC5879496 Allemani C, et al. Global surveillance of trends in cancer survival 2000–14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. Lancet. 2018;391(10125):1023–75. 10.1016/S0140-6736(17)33326-3. 29395269 10.1016/S0140-6736(17)33326-3 PMC5879496

26. Harrell FE. Regression Modeling Strategies With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis Second Edition Springer Series in Statistics. [Online]. Available: http://www.springer.com/series/692

27. Altman DG Royston P The cost of dichotomising continuous variables BMJ 2006 332 7549 1080 10.1136/bmj.332.7549.1080 16675816 PMC1458573 Altman DG, Royston P. The cost of dichotomising continuous variables. BMJ. 2006;332(7549):1080. 10.1136/bmj.332.7549.1080. 16675816 10.1136/bmj.332.7549.1080 PMC1458573

28. Force LM The global, regional, and national burden of cancer, 1990–2023, with forecasts to 2050: a systematic analysis for the Global Burden of Disease Study 2023 Lancet 2025 10.1016/S0140-6736(25)01635-6 PMC12687902 41015051 Force LM, et al. The global, regional, and national burden of cancer, 1990–2023, with forecasts to 2050: a systematic analysis for the Global Burden of Disease Study 2023. Lancet. 2025. 10.1016/S0140-6736(25)01635-6. 10.1016/S0140-6736(25)01635-6 PMC12687902 41015051

29. Percy C, van Holten V, Muir CS, Organization WH. International classification of diseases for oncology. World Health Organization; 1990.

30. Torgbenu E Luckett T Buhagiar MA Chang S Phillips JL Prevalence and incidence of cancer related lymphedema in low and middle-income countries: a systematic review and meta-analysis BMC Cancer 2020 20 1 604 10.1186/s12885-020-07079-7 32600278 PMC7325022 Torgbenu E, Luckett T, Buhagiar MA, Chang S, Phillips JL. Prevalence and incidence of cancer related lymphedema in low and middle-income countries: a systematic review and meta-analysis. BMC Cancer. 2020;20(1):604. 10.1186/s12885-020-07079-7. 32600278 10.1186/s12885-020-07079-7 PMC7325022

31. Bray F Ferlay J Soerjomataram I Siegel RL Torre LA Jemal A Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries CA Cancer J Clin 2018 68 6 394 424 30207593 10.3322/caac.21492 Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. 30207593 10.3322/caac.21492

32. Vos T et al. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019, The lancet, vol. 396, no. 10258, pp. 1204–1222, 2020. 10.1016/S0140-6736(20)30925-9 PMC7567026 33069326

33. Ellis L, Aurélien Belot B, Rachet, Coleman MP. The Mortality-to-Incidence Ratio Is Not a Valid Proxy for Cancer Survival, 2019. [Online]. 10.1200/JGO.19.00038 PMC6550058 31070980

34. Murray CJ Lopez AD Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet 1997 349 9064 1498 1504 10.1016/S0140-6736(96)07492-2 9167458 Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet. 1997;349(9064):1498–504. 10.1016/S0140-6736(96)07492-2. 9167458 10.1016/S0140-6736(96)07492-2

35. Norheim OF Guidance on priority setting in health care (GPS-Health): the inclusion of equity criteria not captured by cost-effectiveness analysis Cost Eff Resour Alloc 2014 10.1186/1478-7547-12-18 25246855 PMC4171087 Norheim OF, et al. Guidance on priority setting in health care (GPS-Health): the inclusion of equity criteria not captured by cost-effectiveness analysis. Cost Eff Resour Alloc. 2014. 10.1186/1478-7547-12-18. 25246855 10.1186/1478-7547-12-18 PMC4171087

36. Ebrahimi H Global, regional, and national burden of respiratory tract cancers and associated risk factors from 1990 to 2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet Respir Med 2021 9 9 1030 49 10.1016/S2213-2600(21)00164-8 34411511 PMC8410610 Ebrahimi H, et al. Global, regional, and national burden of respiratory tract cancers and associated risk factors from 1990 to 2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Respir Med. 2021;9(9):1030–49. 34411511 10.1016/S2213-2600(21)00164-8 PMC8410610

37. Ebrahimi H Global, regional, and national burden of respiratory tract cancers and associated risk factors from 1990 to 2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet Respir Med 2021 9 9 1030 1049 10.1016/S2213-2600(21)00164-8 34411511 PMC8410610 Ebrahimi H, et al. Global, regional, and national burden of respiratory tract cancers and associated risk factors from 1990 to 2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Respir Med. 2021;9(9):1030–49. 10.1016/S2213-2600(21)00164-8. 34411511 10.1016/S2213-2600(21)00164-8 PMC8410610

38. Salomon JA Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010 Lancet 2012 380 9859 2129 2143 10.1016/S0140-6736(12)61680-8 23245605 PMC10782811 Salomon JA, et al. Common values in assessing health outcomes from disease and injury: disability weights measurement study for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2129–43. 10.1016/S0140-6736(12)61680-8. 23245605 10.1016/S0140-6736(12)61680-8 PMC10782811