Background and Purpose
In this special report, NIH has plotted funding for a variety of diseases and conditions compared to disease burden measurements for many of those conditions. Burden of disease is the impact of a health problem, as measured by prevalence, incidence, mortality, morbidity, extent of disability, financial cost, or other indicators, each of which can be measured in different ways. Understanding the alignment of NIH funding to disease burden is one of many important considerations for setting research funding priorities at NIH (see the NIH-Wide Strategic Plan). NIH believes that it is useful to examine multiple sources of disease burden data which reflect a broad set of measures and can allow for comparison between U.S. and global burdens. Understanding the global health landscape allows for better planning for future global and domestic health needs, and can aid in predicting and analyzing disease outbreaks before they reach the United States.
NIH reports its research investments using the Research, Disease, and Condition Categorization (RCDC) system. NIH currently posts Federally-sourced disease statistics from the National Center for Health Statistics at the Centers for Disease Control & Prevention (CDC) alongside select budgeting categories in the “Estimates of Funding for Various Research, Condition, and Disease Categories (RCDC)” table. NIH posted these data to provide the public and policymakers with information that is helpful for understanding the NIH research portfolio and its relationship to U.S. public health needs.
To supplement NIH’s posting of selected domestic health statistics collected by the CDC, this report examines additional measures and includes data from countries around the world. NIH used burden data from the Global Burden of Disease (GBD) study, which measured the burdens of 282 causes of death and disability across 195 countries in the year 2017.1,2 Two measures of burden captured by the GBD study are deaths and disability-adjusted life years (DALYs), a measure of years of healthy life lost due to disability and premature death. See “additional information” below for more explanation of the measurements chosen for this analysis.
Results
The four graphs below illustrate 1) Categorized NIH Spending vs. U.S. DALYs, 2) Categorized NIH Spending vs. Global DALYs, 3) Categorized NIH Spending vs. U.S. deaths, and 4) Categorized NIH Spending vs. Global deaths. Since some diseases do not cause death, they are not included in figure 3 and figure 4. Note that not every RCDC or GBD category of disease is included in these graphs – see the “Method for Matching Research Funding Categories to Disease Burden Categories” section below for more information on which categories are included in this analysis and why. Also note that the mortality figures presented here were gathered via different sources and methods compared to the figures currently reported on the Categorical Spending page (from the National Center for Health Statistics), and thus it is unlikely that mortality figures for the U.S. will match between the two analyses due to differences in methodology.
Figure 1: Categorized NIH Spending vs. US DALYs
A plot of alignment between the number of 2017 Disability Adjusted Life Years (DALYS) lost to a disease or condition category in the United States, and the amount of NIH funding for research on that disease or condition category in 2017. Mouse over a point to see the name of the category. Each point represents the closest possible match between Global Burden of Disease categories and NIH's Research, Condition, and Disease Categories. Note: some categories may overlap (ex. Lung and Pneumonia).
Figure 2: Categorized NIH Spending vs. Global DALYs
A plot of alignment between the number of 2017 Disability Adjusted Life Years (DALYS) lost to a disease or condition category in 195 countries across the world, and the amount of NIH funding for research on that disease or condition category in 2017. Mouse over a point to see the name of the category. Each point represents the closest possible match between Global Burden of Disease categories and NIH's Research, Condition, and Disease Categories. Note: some categories may overlap (ex. Lung and Pneumonia).
Figure 3: Categorized NIH Spending vs. US Deaths
A plot of alignment between the number of 2017 deaths attributed to a disease or condition category in the United States, and the amount of NIH funding for research on that disease or condition category in 2017. Mouse over a point to see the name of the category. Each point represents the closest possible match between Global Burden of Disease categories and NIH's Research, Condition, and Disease Categories. Note: some categories may overlap (ex. Lung and Pneumonia). Among categories for which DALYs are reported, no deaths were attributed to the following categories: Anxiety Disorders, Attention Deficit Disorder (ADD), Autism, Dental/Oral and Craniofacial Disease, Depression, Headaches, Infertility, Macular Degeneration, Malaria, Migraines, Psoriasis, and Schizophrenia.
Figure 4: Categorized NIH Spending vs. Global Deaths
A plot of alignment between the number of 2017 deaths attributed to a disease or condition category in 195 countries across the world, and the amount of NIH funding for research on that disease or condition category in 2017. Mouse over a point to see the name of the category. Each point represents the closest possible match between Global Burden of Disease categories and NIH's Research, Condition, and Disease Categories. Note: some categories may overlap (ex. Lung and Pneumonia). Among categories for which DALYs are reported, no deaths were attributed to the following categories: Anxiety Disorders, Attention Deficit Disorder (ADD), Autism, Dental/Oral and Craniofacial Disease, Depression, Headaches, Infertility, Macular Degeneration, Malaria, Migraines, Psoriasis, and Schizophrenia.
Conclusion
This analysis provides a limited snapshot of the alignment between NIH funding and disease burden domestically and internationally, while still providing some choice of the appropriate burden measurement for a given condition. NIH believes that there is no comprehensive, standard approach for measuring burden across diseases. For the most part, measures of public health burden are designed to detect changes in one disease or condition (or a small subset of related conditions) over time or between populations, with the choices of measurements and methods appropriately tailored for that disease. Very few public health studies are designed to compare burden across a large number of diseases. Different diseases may impose different kinds of burdens on society, requiring different measurements of burden. Some diseases may cause premature death, while other chronic conditions may cause long-term disability and impose a great monetary burden on family members and society. Many diseases vary widely in the severity of symptoms, treatment strategies, and health outcomes. For example, the cost of treating a thousand people with influenza is not equivalent to the cost of treating a thousand people with tuberculosis. Because of the many nuances of measuring public health burden across different studies, populations, and regions, NIH believes that it is not possible to have a justifiable 'one size fits all' approach for reporting burden across diseases. Rather, careful consideration of multiple data types and sources on a case-by-case basis provides the best strategy for understanding disease burden and public health need.
Choice of Data Source
For this analysis, NIH chose The World Health Organization's Global Burden of Disease (GBD) study as its source of burden data, using data provided from 2017. The GBD study is one of the few sources of public health burden data that is explicitly designed for comparison across conditions and across regions.i,ii Using an estimation method integrating data ranging from national vital statistics registries to semi-quantitative surveys, the GBD study attempts to make the subjective judgements of disability and severity as consistent as possible. The modeling estimates from the GBD study also provide estimates of under- and over-reporting associated with different nations and different data sources, with the goal of creating a large repository of burden data that is comparable both across conditions and across more than 100 nations.
Measurements Used In this Study
Since lengthening life and reducing illness and disability are central tenets of the NIH mission, for this analysis NIH chose to compare our portfolio investments as measured by RCDC to two different measurements in the GBD dataset: DALYs and Deaths. As noted above, the mortality figures here are gathered via different sources and methods compared to the figures currently reported on the Categorical Spending page (from the National Center for Health Statistics), and thus it is unlikely that mortality figures for the U.S. will match between the two analyses due to differences in methodology. Disability Adjusted Life Years (DALYs)—a frequently used metric for disease burden—are calculated as a measure of years of healthy life lost due to morbidity and premature mortality. DALYs combine mortality and morbidity estimates into one numerical measurement, and include subjective judgements of the severity of disability of each condition, which can vary between individual studies. CDC, as a federal source, generally provides case-by-case choices of measurements like prevalence, incidence, and mortality. CDC does not provide data that involves subjective weighting of disability or life quality. However, when DALYs are used to compare vastly different diseases that impose a variety of types of burden (financial, disability, mortality, U.S. vs global), they can provide an incomplete picture of the differences between diseases. Given these concerns, NIH believes that this data should be taken into consideration as one of several measurements, in order to form the most comprehensive picture of disease impact. NIH believes that these data should be considered as a complementary source to those posted on the Categorical Spending page.
Method for Matching Research Funding Categories to Disease Burden Categories
The categories represented in RCDC are different from those in the GBD study. As implied by its name, many RCDC categories are not disease categories, but rather, research areas, like genetics or neuroscience. Other categories capture NIH investments in research on specific populations, such as pediatrics or minority health. Still others involve rare diseases, such as Pick's Disease, that are below the threshold of measurement for the GBD study. In addition, category definitions and delineations used in RCDC do not always perfectly match those used by the GBD study, given RCDC’s focus on categorizing areas of research rather than specific causes of death and disability. The differing characteristics of the two data sets mean that there are caveats to interpreting the matched data, and that there are judgement calls required in some cases to determine the best fits between the two sources.
In order to align GBD data to RCDC categories, NIH staff made subjective judgements for determining which GBD categories could serve as appropriate proxies for RCDC categories. Straightforward matches in the GBD dataset did not exist for every RCDC category. For some categories, the definitions used to capture research related to a given condition may be broader than the definition used to capture the burden of that condition. For example, while opioids make up a large enough fraction of the Prescription Drug Abuse category that it could be reasonably matched to GBD data on Opioid Use Disorders, the RCDC category contains research on other, non-opioid prescription drug abuse as well. In addition, some RCDC categories required adding and/or subtracting components of the GBD categories in order to make them a reasonable match. For example, to match the RCDC category of 'Lung', staff combined GBD categories for lung cancer, lower respiratory infections, and chronic pulmonary diseases such as COPD.
In total, NIH was able to match 74 RCDC categories to GBD burden data. Using methods similar to those used in previous independent academic analyses of the same or similar questions 3,4,5 NIH plotted Categorized NIH Spending levels for these 74 categories alongside measurements of U.S. Deaths, U.S. DALYs, Global Deaths, and Global DALYs. This analysis provides a limited snapshot of the alignment between NIH funding and disease burden domestically and internationally, while still providing some choice of the appropriate burden measurement for a given condition.
1 GBD 2017 Causes of Death Collaborators. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the Global Burden of Disease Study 2017.The Lancet. 2018 Nov 10;392(10159):1736-1788. doi: 10.1016/S0140-6736(18)32203-7. Epub 2018 Nov 8.
2 GBD 2017 DALYs and HALE Collaborators. 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. The Lancet. 2018 Nov 10;392(10159):1859-1922. doi: 10.1016/S0140-6736(18)32335-3.
3 Moses H, 3rd, Matheson DH, Cairns-Smith S, George BP, Palisch C, Dorsey ER. The anatomy of medical research: US and international comparisons. JAMA. 2015;313(2):174-189.
4 Sampat BN, Buterbaugh K, Perl M. New evidence on the allocation of NIH funds across diseases. The Milbank quarterly. 2013;91(1):163-185.
5 Gillum LA, Gouveia C, Dorsey ER, et al. NIH disease funding levels and burden of disease. PloS one. 2011;6(2):e16837.