Appendixes

Appendix II. Data sources, measures, country groupings, and benchmarking methods used in Chapter 2

Data sources

Chapter 2 estimates are based on data from the Global Burden of Disease (GBD) Study 2023 by the Institute for Health Metrics and Evaluation (IHME) (1). The analysis in the chapter draws on GBD estimates for disability-adjusted life years (DALYs) (referred to in the chapter as “health loss”), years of life lost due to premature mortality (YLLs), years lived with disability (YLDs), the leading causes of health loss as classified by the GBD Level 2 causes (which group related diseases and conditions together), Level 2 risk factors, and risk–cause pairings for selected years between 2000 and 2023. The primary analytic years are 2000, 2015, and 2023, with annual time series used when trend interpretation is required. The analysis includes the 35 Member States of the Region of the Americas, grouped using PAHO’s Sustainable Health Index (SHIx) framework.

Core disease burden measures

DALYs are defined as the sum of YLLs and YLDs. YLLs combine deaths with the standard life expectancy at the age of death, so deaths at younger ages contribute more years of life lost. YLDs combine the prevalence of disease sequelae with disability weights that reflect severity. DALYs are therefore interpreted as years of healthy life lost. While useful for comparing overall health loss, DALYs should not be interpreted as a direct measure of health system performance without considering risks, social determinants, demographic structure, and other contextual factors.

Age-standardized vs. crude DALY rates

Age-standardized DALY rates are used to compare burden intensity across countries, clusters of countries, and WHO regions after accounting for differences in population age structure. Crude rates are used selectively to reflect the population-level service pressure associated with demographic change, especially where population aging and growth increase the demand for health services. Figure 14 uses crude DALY rates for this reason: they show the population-level burden and associated service pressure that may rise even when age-standardized rates suggest partial improvement in underlying burden intensity. Crude estimates are available in the underlying analytic dataset and are reported in the main body of the report only where they add a distinct policy interpretation not captured by age-standardized rates.

Sustainable Health Index country clusters

To describe between-country equity gradients within the Region, the 35 Member States were stratified using PAHO’s 2025 SHIx framework (Table AII.1). The SHIx incorporates six dimensions:

• Health outcomes, measured using health-adjusted life expectancy (HALE)
• Health access, measured using the Universal Health Coverage (UHC) Service Coverage Index
• Inequality, measured as 100 minus the Gini coefficient
• Economic capacity, measured using log10 gross national income (GNI) per capita
• Social conditions, measured using years of educational attainment
• Environmental conditions, measured using access to water, sanitation, and hygiene (WASH) services

Each dimension is standardized to a 0–⁠1 (lowest to highest) scale using goalposts.

The SHIx is calculated as the geometric mean of two components – the Health Status Index and the Health Determinants Index – comprising the six dimensions outlined above. The Health Status Index is calculated as the geometric mean of standardized health outcomes and access. The Health Determinants Index is calculated as the geometric mean of inequality, economic capacity, social conditions, and environmental conditions. Higher SHIx scores indicate more favorable health status, service coverage, and enabling determinants, including lower income inequality.

The SHIx is used in Chapter 2 as a proxy for between-country development-related gradients relevant to health equity. It does not measure inequalities within countries. The interpretation of SHIx-based findings is therefore limited to differences between country groupings while recognizing that within-country inequalities by income, territory, ethnicity, sex, age, or other dimensions may be substantial.

Cause and risk classification; risk-attribution analysis

GBD Level 2 causes are used to identify the leading causes of DALYs, YLLs, and YLDs. Cause rankings are based primarily on age-standardized rates, both sexes combined. Level 2 risk factors are used for the risk-attribution analysis. Risk-attributable DALY rates and risk–cause pairings are interpreted as comparative risk assessment estimates and are not summed across risks because risk exposures may overlap through mediation pathways.

The risk-attribution analysis uses IHME/⁠GBD comparative risk assessment estimates to identify leading Level 2 risk factors and risk–cause pathways contributing to DALY burden in the Region. For Figure 21, attributable age-standardized DALY rates per 100 000 population were extracted for Level 2 risk factors for 2000, 2015, and 2023, both sexes combined. Risks were ranked in descending order of attributable DALY rate for each year, and the leading risks were retained for display. Because GBD comparative risk assessment methods and inputs are periodically revised, attributable risk estimates – especially for some behavioral and dietary risks – may vary across GBD iterations (2). Accordingly, changes over time and small differences in rank should be interpreted cautiously, as they may reflect either true epidemiologic change or updates in underlying models, data inputs, and risk–outcome specifications.

For Figure 22, attributable age-standardized DALY rates per 100 000 population were extracted for Level 2 risk–cause pairings in 2023. Pairings were ranked in descending order of attributable DALY rate to identify the pathways contributing the largest attributable burden. Pairings with negative attributable values were excluded to focus interpretation on positive contributions to health loss.

Joinpoint trend analysis for selected causes

In Chapter 2, joinpoint regression analysis was used to assess whether the underlying rate of change accelerated, slowed, or reversed over time for two selected Level 2 causes – cardiovascular diseases, which remained among the two leading causes of DALYs between 2000 and 2023, and diabetes and kidney diseases, which rose from the ninth- to the fifth-leading cause over the same period.

A joinpoint is the year at which the estimated trend changes direction or slope. The model divides the full time series into distinct periods and estimates the annual percentage change (APC) within each period. The APC represents the average yearly percentage increase or decrease in the age-standardized DALY rate during that segment. Negative APC values indicate declining rates, while positive APC values indicate increasing rates. Confidence intervals (CIs) and p values are used to assess the statistical precision and significance of each estimated trend.

For the analyses presented in Box 1, joinpoint regression was applied to annual age-standardized DALY rates for two cardiometabolic causes – cardiovascular diseases and diabetes and kidney diseases – covering 2000–⁠2023. Before fitting the joinpoint models, first-order autocorrelation in residuals was assessed using the Durbin-Watson test. Where statistically significant autocorrelation was detected, the models were adjusted using generalized least squares with a first-order autoregressive structure. Results are presented as APCs with 95% CIs and p values (Table AII.2). The results showed that the age-standardized DALY rates for cardiovascular diseases declined throughout 2000–⁠2023, but the pace of improvement became progressively weaker over time. In contrast, age-standardized DALY rates increased for diabetes and kidney diseases over the same time period but with no clear pattern in pace.

The analysis should be interpreted as a description of temporal trends, not as evidence of causality. Identified joinpoints indicate when the slope of the observed burden trend changed; they do not explain why the change occurred. The results are therefore used to support interpretation of burden trajectories and priority-setting, while causal explanations require additional epidemiological, policy, and health system analysis.

Potentially avertable DALY benchmark

Potentially avertable DALYs were calculated by comparing each country’s 2023 age-standardized DALY rate for each Level 2 cause of health loss with a global best-performance benchmark for the same cause. The benchmark was defined as the 10th percentile of age-standardized DALY rates across 204 countries and territories in 2023. This benchmark was selected because it reflects comparatively strong performance already observed globally while reducing sensitivity to outliers that would arise from using the single lowest observed rate.

For each country in the Region and Level 2 cause, the gap was calculated as the difference between the country’s observed age-standardized DALY rate and the global 10th percentile benchmark for that cause. When a country’s rate was already at or below the benchmark, the gap was set to zero. The resulting cause-specific gap rate was interpreted as benchmark-defined potentially avertable DALY burden per 100 000 population.

To support aggregation across causes, countries, and SHIx clusters, cause-specific gap rates were converted into approximate DALY numbers by multiplying the gap rate by the country population and dividing by 100 000. These approximate numbers were then summed across causes to estimate country-level potentially avertable DALYs, and across countries to estimate cluster-level and regional totals. Because the gaps are derived from age-standardized rates and subsequently scaled using country populations, the resulting DALY numbers and shares should be interpreted as approximate, population-scaled equivalents rather than observed counts.

The measure should be interpreted as benchmark-defined improvement potential, not as a prediction of the burden that would be averted by any specific intervention package. It does not identify causal drivers of the gap, nor does it attribute differences solely to health system performance. Observed gaps may reflect a combination of risk exposure, social determinants, demographic and epidemiological context, health system access and quality, policy choices, and data/modeling uncertainty.

The equity interpretation is nevertheless central. Because the benchmark reflects levels already observed among better-performing countries globally, the gap provides a policy-relevant measure of avoidable inequity: the portion of health loss that remains above an empirically observed better-performance threshold. Differences in potentially avertable DALYs across countries and SHIx clusters show the technical potential for improvement and the uneven distribution of health loss across the Region compared with the defined benchmark.

The measure is summarized in several ways in Chapter 2. For Figure 23, potentially avertable burden by country is expressed as a share of total DALYs: total potentially avertable DALYs divided by total DALYs, multiplied by 100.

For Figure 24, potentially avertable DALYs were summed across all 22 Level 2 causes and all countries in each SHIx cluster, and then expressed as a share of total DALYs in that cluster. The cluster-level potentially avertable share was calculated as total cluster gap DALYs divided by total cluster DALYs, multiplied by 100. This figure therefore summarizes the proportion of each cluster’s total DALY burden that remains above the global best-performance benchmark. Appendix IV provides the cluster-specific cause composition of potentially avertable DALYs. Within each SHIx cluster, cause-specific gap DALYs were summed across countries to obtain cluster-level gap DALYs by cause. Each cause was then expressed as a percentage share of the cluster’s total potentially avertable DALYs, with causes ranked within each cluster by this share.

In Figure 25, each Level 2 cause’s share of total DALYs is compared with its share of potentially avertable DALYs at the regional level. The total DALY share was calculated as cause-specific total DALYs divided by all-cause total DALYs. The potentially avertable share was calculated as cause-specific gap DALYs divided by all-cause gap DALYs, after country-cause gap DALYs had been summed across the Region.

In Figure 26, potentially avertable DALYs were decomposed into YLL and YLD components. This was done by applying the same 10th percentile benchmarking approach separately to cause-specific, age-standardized YLL rates and cause-specific, age-standardized YLD rates for each year shown, then converting the resulting excess rates to approximate numbers using population. Panel (a) compares the YLL versus YLD composition of potentially avertable DALYs for the Region in 2000, 2015, and 2023 while Panel (b) presents the cause-specific YLL versus YLD composition of potentially avertable DALYs in 2023.

Appendix III. SHIx cluster-specific cause-rank trajectories, 2000–⁠2023

This appendix presents cluster-specific rank trajectories for Level 2 causes of health loss across the six SHIx country clusters over 2000–⁠2023. The figures in this appendix complement Figure 20 in the main body of the report, which summarizes the top five causes by SHIx cluster in 2023. The appendix adds a temporal dimension by showing how leading causes changed within each cluster, where pandemic-era reordering was most visible, and where longer-term shifts differed from the average for the Region of the Americas.

The figures should be interpreted as rank trajectories, not as changes in absolute burden of disease. A cause moving upward in rank may reflect an increase in its own burden, a decline in other causes, or a temporary shock that changes the relative ordering of causes. The appendix is therefore intended to support interpretation of cause prominence and prioritization across country groupings, while the main body of the report retains the regional synthesis.

Cluster 1

Cluster 1 shows the most pronounced lower-SHIx burden profile, with persistently high prominence of both NCDs and historically high burden – albeit diminishing – communicable and maternal and neonatal conditions (Figure AIII.1). Cardiovascular diseases, neoplasms, musculoskeletal disorders, mental disorders, and diabetes and kidney diseases remained important or gained prominence over time, confirming that the NCD transition is well established even in the lowest-SHIx cluster. At the same time, maternal and neonatal disorders and respiratory infections and tuberculosis remained more visible than in higher-SHIx clusters, indicating that the cluster’s priority profile is not defined by NCDs alone.

The pandemic-era reordering was especially pronounced in Cluster 1. Respiratory infections and tuberculosis rose sharply during 2020–⁠2021 and remained relatively prominent afterward, suggesting a more sustained infectious-disease shock than in some higher-SHIx clusters. The 2010 peak visible in the regional analysis is also particularly relevant to this cluster because it coincides with the Haiti earthquake.

Core insight: Overall, Cluster 1 illustrates the need for a dual agenda: continued progress on maternal, neonatal, and infectious conditions alongside stronger prevention and long-term management of chronic diseases.

Cluster 2

Cluster 2 stands out as one of the most differentiated profiles. Cardiovascular diseases and neoplasms remain highly prominent, but the cluster also shows a distinctive role for self-harm and interpersonal violence, which ranks more prominently than in most other SHIx groupings (Figure AIII.2). This pattern suggests that injury- and violence-related burden is more central to the cluster’s priority profile than would be apparent from the regional average alone.

Cluster 2 also differed from most other clusters in its long-term trajectory. While several clusters experienced declining age-standardized DALY rates before the pandemic, Cluster 2 did not. Its cause-rank profile helps explain why a uniform regional NCD agenda is insufficient: cardiometabolic and neoplasm priorities remain central, but they need to be complemented by stronger attention to violence, injuries, and context-specific drivers of premature mortality and disability.

Core insight: Pandemic-era reordering was visible, but the more important message for Cluster 2 is the persistence of a differentiated burden profile that diverges from the regional average.

Cluster 3

Cluster 3 broadly reflects the regional pattern of NCD dominance, with cardiovascular diseases, neoplasms, musculoskeletal disorders, mental disorders, and diabetes and kidney diseases remaining central to the burden profile (Figure AIII.3). Diabetes and kidney diseases became more prominent over time, consistent with the broader regional shift toward cardiometabolic conditions. At the same time, the cluster includes a diverse set of countries, and its rank trajectories show that the regional average can obscure variation in how chronic, injury-related, and infectious causes interact within middle-SHIx settings.

The pandemic produced a marked temporary reordering in Cluster 3, with respiratory infections and tuberculosis rising sharply during 2020–⁠2021. By 2023, however, the cluster’s profile again reflected the wider regional pattern in which NCDs dominated the leading causes of health loss.

Core insight: Cluster 3 illustrates both the depth of the COVID-era shock and the durability of the longer-term NCD transition.

Cluster 4

Cluster 4 shows a burden profile strongly shaped by NCDs, with cardiovascular diseases, neoplasms, diabetes and kidney diseases, musculoskeletal disorders, and mental disorders among the leading causes (Figure AIII.4). The prominence of diabetes and kidney diseases is especially important because it indicates that cardiometabolic conditions are not only high burden but also central to the cluster’s evolving priority agenda.

Compared with lower-SHIx clusters, maternal and neonatal disorders and respiratory infections and tuberculosis are less dominant in the longer-term profile, although pandemic-era reordering was still visible during 2020–⁠2021. By 2023 the cluster’s leading causes again pointed toward chronic disease prevention, risk factor management, and continuity of care as the main delivery priorities.

Core insight: Cluster 4 reinforces the importance of moving from broad NCD recognition to more operational cardiometabolic and chronic-care strategies.

Cluster 5

Cluster 5 has a burden profile in which NCDs dominate but with continued variation in the relative prominence of chronic diseases, mental disorders, and residual infectious causes. Cardiovascular diseases, neoplasms, musculoskeletal disorders, and mental disorders remained central (Figure AIII.5). These rankings point to a burden profile increasingly shaped by long-term chronic disease management, disability, and conditions requiring sustained continuity of care.

The COVID-era rise of respiratory infections and tuberculosis was visible in Cluster 5, and the cause remained important in the later pandemic period compared with higher-SHIx settings. This suggests that even higher-SHIx clusters are not insulated from infectious-disease shocks. However, the longer-term profile remains dominated by chronic and noncommunicable causes.

Core insight: Cluster 5 illustrates the need to combine resilience to acute shocks with continued investment in chronic care, rehabilitation, mental disorders, and cardiometabolic prevention.

Cluster 6

Cluster 6 has the most differentiated high-income profile. NCDs dominate the ranking, but the profile is especially marked by prominent burden from mental, musculoskeletal, neurological, and substance use disorders (Figure AIII.6). Cardiovascular diseases and neoplasms remain important, but the cluster’s leading causes also reflect a larger role for chronic disability, behavioral health, pain, and long-term conditions that may not be fully captured by a mortality-focused policy agenda.

Pandemic-era reordering occurred in Cluster 6, but respiratory infections and tuberculosis receded more quickly than in several lower-SHIx clusters. By 2023 the cluster’s leading causes again reflected a high-income chronic morbidity profile, with substance use disorders standing out more clearly than in most other clusters.

Core insight: Cluster 6 illustrates why the regional priority agenda should not be limited to cardiometabolic and neoplasm priorities alone. In higher-SHIx settings, chronic disability, mental disorders, neurological conditions, and substance use also require sustained policy and delivery attention.

Appendix IV. Cause composition of potentially avertable DALYs by SHIx cluster, 2023

Appendix IV presents the cluster-specific cause composition of potentially avertable DALYs in 2023. These figures complement Figure 24 in the main body of the report, which shows the overall magnitude of potentially avertable burden across SHIx clusters. The appendix adds a compositional lens by showing which Level 2 causes of health loss account for the largest shares of each cluster’s benchmark-defined gap.

The figures should be interpreted as the distribution of potentially avertable DALYs within each SHIx cluster, not as total burden levels. A cause with a large share in a cluster does not necessarily mean that the cluster has the highest absolute burden for that cause; rather, it means that the cause accounts for a larger proportion of that cluster’s total benchmark-defined gap. The appendix is therefore intended to support priority-setting and delivery implications, while the main body of the report retains the regional synthesis.

Cluster 1

Cluster 1 shows a mixed potentially avertable burden profile. The top three causes account for 39% of the cluster’s benchmark-defined gap, while the top five account for 52% (Figure AIV.1). This indicates a moderately concentrated profile: improvement potential is shaped by a small group of leading causes but not dominated by one cause alone.

The leading contributors are cardiovascular diseases, respiratory infections and tuberculosis, and diabetes and kidney diseases. This combination reflects the dual agenda already visible in the cluster’s broader burden profile: lower-SHIx countries face substantial improvement potential from chronic and cardiometabolic conditions, while infectious causes remain more prominent than in higher-SHIx clusters.

Core insight: Cluster 1 requires both strengthened chronic disease prevention and long-term management and continued attention to infectious disease control, respiratory health, and health system resilience.

Cluster 2

Cluster 2 has one of the most concentrated potentially avertable burden profiles. The top three causes account for 48% of the cluster’s benchmark-defined gap, and the top five account for 64% (Figure AIV.2). This suggests that a relatively small number of causes explain a large share of the cluster’s remaining improvement potential.

The leading causes are cardiovascular diseases, maternal and neonatal disorders, and self-harm and interpersonal violence. This profile differs from a standard NCD-only interpretation. Cardiovascular diseases remain central, but the large contributions from maternal and neonatal conditions and violence-related burden point to a more differentiated priority agenda.

Core insight: For Cluster 2, reducing potentially avertable DALYs requires a focused but multi-domain strategy: cardiometabolic prevention and care, stronger maternal and newborn services, and more explicit attention to violence, injury prevention, mental health, and crisis response.

Cluster 3

Cluster 3 has the broadest spread of potentially avertable burden across causes. The top three causes account for 32% of the cluster’s benchmark-defined gap, and the top five account for 48% (Figure AIV.3). This indicates that improvement potential is distributed across a wider set of conditions rather than concentrated in a small number of causes.

The leading contributors are diabetes and kidney diseases, self-harm and interpersonal violence, and cardiovascular diseases. This combination suggests that Cluster 3 requires a broad implementation agenda. Cardiometabolic care is clearly important, but it does not fully define the cluster’s improvement potential. Injury, violence, mental health, and infectious causes also contribute meaningfully.

Core insight: For Cluster 3, the policy implication is less about one single, focused package and more about strengthening delivery platforms that can address multiple causes through prevention, primary health care, referral systems, chronic care, and multisectoral responses.

Cluster 4

Cluster 4 shows a relatively concentrated potentially avertable burden profile. The top three causes account for 46% of the benchmark-defined gap, and the top five account for 59% (Figure AIV.4). This indicates that improvement potential is meaningfully concentrated in a small number of causes, making focused priority packages especially relevant.

The leading contributors are cardiovascular diseases, diabetes and kidney diseases, and maternal and neonatal disorders. The prominence of the first two causes points to a strong cardiometabolic agenda, including prevention, early detection, risk management, medicine continuity, and long-term follow-up. At the same time, the continued contribution of maternal and neonatal disorders indicates that Cluster 4’s avertable gap is not limited to chronic disease alone.

Core insight: A credible priority strategy would combine cardiometabolic care with targeted strengthening of maternal and newborn services.

Cluster 5

Cluster 5 presents a more distributed, multi-pillar potentially avertable burden profile. Several causes contribute at similar levels, including neoplasms, cardiovascular diseases, respiratory infections and tuberculosis, and mental disorders (Figure AIV.5). This suggests that the cluster’s benchmark-defined gap is spread across major chronic, infectious, and mental health categories, rather than being dominated by a single leading cause.

This profile has two implications. First, Cluster 5 still requires sustained attention to major NCDs, especially cancer and cardiovascular diseases. Second, the contributions from respiratory infections and tuberculosis and mental disorders show that improvement potential also depends on resilience to infectious disease shocks, continuity of essential services, and stronger mental health care.

Core insight: For Cluster 5, the policy response is best framed as a multi-pillar strategy: chronic disease prevention and management, cancer control, respiratory and infectious disease readiness, and mental health service strengthening.

Cluster 6

Cluster 6 has a distinctive high SHIx profile, with a relatively concentrated benchmark-defined gap. The top three causes account for 43% of potentially avertable DALYs, and the top five account for 58% (Figure AIV.6). Unlike most other clusters, however, the largest contributor is not cardiovascular diseases or diabetes and kidney diseases. Substance use disorders account for the largest single-cause share, followed by musculoskeletal disorders and cardiovascular diseases.

This pattern shows why the regional priority agenda cannot be limited to cardiometabolic and cancer priorities alone. In the highest-SHIx cluster, a large share of remaining improvement potential is linked to behavioral health, chronic disability, pain, functional limitation, and long-duration conditions. Cardiovascular diseases remain important, but reducing the benchmark-defined gap in Cluster 6 also requires stronger substance use prevention and treatment, mental health and behavioral health integration, rehabilitation, and chronic disability support.

Core insight: Cluster 6 illustrates a different type of avoidable inequity – that is, one driven not primarily by infectious or maternal and neonatal conditions but by gaps in the prevention and management of chronic morbidity, behavioral health, and disabling conditions.

Appendix V. Trade-offs, challenges, and lessons from explicit health benefit package experiences in the Americas

Appendix VI. WHO OneHealth Tool modeling approach

Purpose and scope

The modeling exercise presented in Chapter 3 was designed to illustrate how a high-priority cardiometabolic service bundle can be translated into explicit assumptions about intervention coverage, target populations, costs, delivery platforms, and modeled health gains across diverse country contexts. It focused on cardiovascular diseases and diabetes because Chapter 2 identified cardiometabolic conditions as central to the Region’s current burden and benchmark-defined improvement potential. The exercise was not intended to define a universal regional package, produce definitive country investment cases, or replace country-led planning and deliberative priority-setting. Rather, it was used as an illustrative application of the WHO OneHealth Tool (3) to show how evidence on disease burden, risk attribution, and potentially avertable burden can be translated into a costed and modeled service bundle with estimated costs and health gains.

Country selection

Six countries were selected to represent the six SHIx clusters: Haiti (Cluster 1), Guyana (Cluster 2), Paraguay (Cluster 3), Costa Rica (Cluster 4), Uruguay (Cluster 5), and Canada (Cluster 6). These countries were selected because they represented distinct SHIx clusters and had the required OneHealth Tool inputs available for the modeled interventions. The use of one country per cluster was intended to keep the exercise parsimonious while illustrating how the same service bundle can produce different costs and health gains across diverse health system and financing contexts. The results should therefore be interpreted as an illustrative six-country exercise across SHIx clusters, not as a complete regional model of all countries in the Americas.

Model, module, and simulation settings

The exercise used the WHO OneHealth Tool, version 6.43. Simulations were run for the period 2025–⁠2030 using the program areas route under the NCD module, specifically the cardiovascular diseases and diabetes module. The same 10-intervention package and coverage scale-up assumptions were applied across the six countries to support cross-country comparability. The mode of the analysis was set to “program areas.” Human resource needs were calculated based on target-setting rather than policy decision. Tuberculosis and the Lives Saved Tool were excluded from the simulations.

The model used the preloaded OneHealth Tool structure for target populations, populations in need, treatment inputs, delivery channels, drugs, and supplies. Bottleneck analysis was left unchanged, using the default preloaded assumptions in the OneHealth Tool.

Intervention package and coverage assumptions

The modeled package included 10 selected cardiovascular disease and diabetes interventions (Table AVI.1). These interventions were selected to represent a combined package of screening, risk-based prevention, chronic disease management, acute cardiovascular care, and post-acute management. The same intervention package was used across all six countries. Baseline and target coverage values were entered to create an illustrative scale-up scenario from 2025 to 2030. The coverage assumptions were illustrative and were not validated country targets.

The relationship between these interventions and the broad GBD categories used in Chapter 2 is not one-to-one. Within cardiovascular diseases, the modeled interventions address ischemic heart disease, including acute myocardial infarction and established disease, cerebrovascular disease and post-stroke care, and elevated blood pressure and cholesterol as cardiovascular risk factors. The intervention for high blood pressure addresses hypertension as a risk factor and treatment target. Within the GBD category of “diabetes and kidney diseases,” the exercise includes diabetes risk assessment and standard glycemic control. The results should consequently not be interpreted as representing the full cardiovascular diseases or diabetes and kidney diseases categories used in Chapter 2.

For most interventions, the population in need was set at 100% of the OneHealth Tool-defined target population. Two exceptions were applied. First, for follow-up care among those at low absolute risk of cardiovascular disease/diabetes, the lower percentages autopopulated by the OneHealth Tool were retained. Second, for screening for cardiovascular disease/diabetes risk, the population in need was manually revised from 100% to 50% in 2025, rising to 75% by 2030.

Program cost assumptions

A program cost layer was added to avoid modeling the package only through commodities and clinical inputs and to capture selected system support functions required for scale-up, such as supervision, training, monitoring, coordination, and transport. Program cost values were based on WHO-CHOICE values for 2025, extracted from the Generalized Cost-Effectiveness Analysis toolkit (4) within the OneHealth Tool. The results reported in Chapter 3 used the baseline cost scenario, in which program costs were held constant at 2025 assumptions.

Delivery channels

The delivery channel structure that was preloaded in the OneHealth Tool was used for the selected NCD interventions. This structure was kept unchanged across the six country simulations. Intervention-level delivery channels included community, outreach, clinic, and hospital (Table AVI.2).

Cost outputs

The cost outputs used in Chapter 3 included total package costs, intervention costs, program costs, and cumulative additional scale-up costs. Total package costs were reported cumulatively for 2025–⁠2030. Cumulative additional scale-up costs were calculated as the sum of annual incremental costs over 2026–⁠2030 relative to the 2025 baseline. Costs were reported in US dollars as generated by the OneHealth Tool simulation output. The cost outputs reported in the main body of the report are from the baseline scenario, in which program costs were held constant at 2025 assumptions.

Domestic general government health expenditure (GGHE-D) values for 2023 were drawn from the WHO Global Health Expenditure Database (5) and used only to provide public financing context. They were not used as affordability thresholds or cost-effectiveness benchmarks.

Health-impact outputs

Health impacts were taken from the OneHealth Tool’s NCD module outputs for cardiovascular disease and diabetes. The main health-impact metrics were healthy life years gained and deaths averted.

Healthy life years gained were used in the main body of the report because this is the health-impact output reported by the OneHealth Tool. In WHO-CHOICE terminology, healthy life years gained are conceptually related to DALYs averted but expressed in positive terms as health gained rather than burden averted (6). They were therefore not relabeled as DALYs averted in the chapter.

Health impacts were reported separately for cardiovascular disease and diabetes. The model generated substantially larger modeled health gains for cardiovascular disease than for diabetes over 2026–⁠2030. This difference should be interpreted in light of the short modeling horizon, the selected interventions, and the way impacts are captured in the OneHealth Tool module. The exercise should therefore not be interpreted as directly estimating closure of the GBD-based potentially avertable DALY gap used in Chapter 2.

Package-to-impact ratios

Illustrative package-to-impact ratios were calculated to compare cumulative additional scale-up costs with modeled cardiovascular health gains. These ratios were used as allocative efficiency discussion aids, not as formal incremental cost-effectiveness ratios. They were reported for the baseline scenario used in the main body of the report.

Ratios were calculated as cumulative additional scale-up costs over 2026–⁠2030 divided by cumulative modeled cardiovascular health gains over 2026–⁠2030. Two ratios were reported in Chapter 3: cost per cardiovascular healthy life year gained, and cost per cardiovascular death averted.

Ratios were presented only for cardiovascular outcomes because modeled diabetes impacts were much smaller over the 2026–⁠2030 horizon, making diabetes ratios less informative for policy interpretation. Because the package was modeled jointly but health impacts were reported separately for cardiovascular disease and diabetes, the resulting ratios should not be interpreted as intervention-specific cost-effectiveness estimates.

Supporting diabetes impact results not emphasized in the main text

Health impacts were also generated for diabetes (Table AVI.3), but these results were not emphasized in the main body. Notably, 9 of the 10 modeled interventions were related directly to cardiovascular risk detection, risk treatment, acute cardiovascular care, or post-event management, while only standard glycemic control was exclusively specific to diabetes. In addition, the benefits of diabetes care often accrue over a longer time horizon through reductions in complications, disability, kidney disease progression, and cardiovascular risk; these longer-term gains are not fully captured in a five-year scale-up scenario. For this reason, diabetes is retained as part of the integrated cardiometabolic platform and the results are presented in Table AVI.3, but the main body of the report focuses on cardiovascular outcomes.

Limitations

The modeling exercise had several limitations. The selected countries were used as illustrative examples for each SHIx cluster and do not constitute a complete regional analysis. Coverage values and scale-up assumptions were illustrative and not validated country targets; therefore, they should not be interpreted as official national plans. The model relied on preloaded OneHealth Tool target populations, population-in-need assumptions, delivery channels, treatment inputs, drugs, and supplies, except where specific adjustments were made for selected population-in-need assumptions.

Canada’s program cost values in the Generalized Cost-Effectiveness Analysis package drew on the Uruguay cost structure rather than a dedicated Canada template; this does not affect the intervention cost component, which draws on Canada’s own epidemiological and demographic data, but it introduces uncertainty into the program cost estimates for Canada specifically.

The package was modeled jointly, but health impacts were reported separately for cardiovascular disease and diabetes, and the short 2026–⁠2030 modeling horizon is more likely to capture near-term cardiovascular effects than longer-term diabetes-related benefits. Healthy life years gained were used as reported by the OneHealth Tool and were not relabeled as DALYs averted; accordingly, the exercise does not estimate how much of the potentially avertable DALY gap would be closed by the modeled scale-up. Package-to-impact ratios are illustrative and should not be interpreted as formal, incremental cost-effectiveness ratios or used to rank countries or interventions mechanically.

Overall, the results are most useful as a planning and deliberation aid: they show how a priority bundle can be specified, costed, linked to delivery channels, and connected to modeled health gains, but they do not replace country-led priority-setting, equity analysis, budget-impact assessment, or implementation planning.

Appendix references

1. Global Burden of Disease Collaborative Network. Global Burden of Disease Study 2023 Seattle: IHME; 2025 [cited 10 August 2026].

2. Zavalis EA, Pezzullo AM, Ioannidis JPA. Instability of Global Burden of Disease estimates of deaths and DALYs from major risk factors JAMA Health Forum. 2026;7(​3):e260108.

3. World Health Organization. OneHealth Tool Geneva: WHO; 2026.

4. Hutubessy R, Chisholm D, Edejer TT. Generalized cost-effectiveness analysis for national-level priority-setting in the health sector Cost Eff Resour Alloc. 2003;1:8.

5. World Health Organization. Global Health Expenditure Database Geneva: WHO; 2026 [updated December 2025].

6. World Health Organization. Frequently asked questions on the updated Appendix 3 of the WHO Global NCD Action Plan 2013–⁠2030: Responses from WHO in relation to the main questions received from Member States and non-State actors at or after the consultation sessions Geneva: WHO; 2022.