Helseanalyse
What are disease burden calculations used for?
Article
|Published
Disease burden analyses provide an overview of how diseases, injuries and risk factors contribute to premature death and reduced health in a population. This is very useful for comparing and following the development of health in a population. However, there are several limitations to such calculations.
Why do we need disease burden analyses?
The Norwegian Institute of Public Health shall produce, summarize and disseminate knowledge that will contribute to good public health work and good health and care services. This includes knowledge about the population’s health status, and factors that affect it [1]. Traditionally, knowledge about the health status of the population has been largely based on knowledge about mortality, often presented as life expectancy or the number of deaths from various causes of death. Mortality is an important indicator of the state of public health, but with clear limitations. Depression, back pain, migraines, and psoriasis for example do not necessarily lead to a shorter life expectancy, but the conditions undoubtedly cause suffering and loss of health as well as significant use of public resources [2]. A large proportion of sick leave cases in Norway are due to mental disorders and musculoskeletal disorders [3] – conditions that cause death to only a small extent.
A complete overview of a population's health status should therefore include information on mortality and causes of death, as well as information on loss of health due to diseases or injuries that are not necessarily associated with increased mortality. Disease burden calculations provide such a complete overview.
The Global Burden of Disease Study
The Global Burden of Disease Study (GBD) has developed a composite measure of the burden of disease called the Disability-Adjusted Life Year (DALY). The term DALY was first described in the World Bank’s Investing in Health report in 1993 [4]. Since then, the GBD project has been continuously expanded. Currently, it estimates the burden of disease from over 350 diseases and injuries by age, sex and over time (GBD 2023) [5-7]. This includes estimating the incidence and severity of over 3000 sequelae. Calculations are made for 204 countries and territories, and for 660 subnational units, including all Norwegian counties. It is also estimated how much of the disease burden can be attributed to 88 known and modifiable (groups of) risk factors.
The GBD project is currently coordinated by the Institute for Health Metrics and Evaluation (IHME) at the University of Washington, USA, with the Gates Foundation as the main source of funding. However, the project is carried out in collaboration with over 16,600 researchers and other contributors (“GBD collaborators”) from more than 160 countries and is the world's largest scientific study for estimating the burden of various diseases, injuries and risk factors across age and gender, country and over time. The Center for Burden of Disease at the Norwegian Institute of Public Health is the Norwegian contact point for the global project.
Benefits of burden of disease calculations
The purpose of the GBD project is to provide regularly updated and knowledge-based overviews of levels and trends in the disease burden so that health systems can be improved and disparities reduced. The results highlight health inequalities between population groups (for example by sex, age, and geography) and changes over time. Comparisons across geographical areas are possible since the same methodology is used regardless of geography. Such comparisons can reveal challenges that are common to several countries or regions or are specific to individual countries. The GBD project also calculates potential gains that can be achieved by changing the level of exposure to risk factors in the population. These risk factors are related to behavior (for example, an unhealthy diet, tobacco and drug use), metabolic risk (such as high blood pressure, high cholesterol or high body mass index), the environment (such as air pollution and radon), or the work environment.
Results and insights from disease burden analyses are used in planning and prioritizing health services, preventive measures, and research. Among others, the results from the disease burden project are included in the Public Health Report for 2022-2023 and in the Perspective Report 2021. Globally, disease burden analyses contribute to identifying important public health challenges and are used to evaluate the status of political goals such as the achievement of the UN Sustainable Development Goals (SGD) [8]. Burden of disease analyses can also be used as a basis for discussions of priorities among various preventive measures or when allocating research funds [9]. Burden of disease calculations cannot, however, be used as the sole criterion, as the calculations do not say anything about the effect and costs of an intervention or research project [10].
Challenges and uncertainty
There are also important aspects of health and disease that are not included in GBD analyses. For example, the GBD project does not include calculations of socioeconomic differences in disease burden, or differences between majority and minority populations. Furthermore, disease burden calculations do not give any impression of the potential disease burden from diseases that are kept under control, for example through good vaccination coverage, good hygiene conditions and other infection prevention measures. These important measures are thus “invisible” in the public health picture that the GBD project gives. The current disease burden should therefore not be a guide for downgrading effective measures that help keeping the disease burden low.
Using a single aggregate measure, such as DALY, to compare the burden of death and health loss from disease and injuries over time, between sexes, age groups and geographical areas, is both methodologically and theoretically challenging. It is also associated with considerable uncertainty. Data can be incomplete or even absent, or difficult to access. There are also different ways in which data can be collected in various health surveys and health registries, as well as over time. This gives rise to uncertainty in the calculation of the number of deaths, years of life lost (YLL), health loss (YLD), and DALYs. The uncertainty is expressed using uncertainty intervals and is generally wider for health loss than for mortality. In tables listing the most important causes of DALYs and other disease burden measures, uncertainty intervals are often overlapping. The ordering should therefore not be read as a strict ranking. However, it is obvious that some diseases stand out as larger contributors to the disease burden than others.
Can we trust the results from the GBD project?
The quality of the disease burden estimates from the GBD is entirely dependent on the quality of the underlying data. The results for mortality and causes of death are the most reliable. This applies to both Norway and other countries with well-functioning cause-of-death registration. However, there are only few such countries in the world. In countries without reliable cause-of-death registration, data from surveys and questionnaires filled by relatives of the deceased (verbal autopsy) are used.
Data on the incidence of non-fatal diseases are more uncertain. For Norway, there is a higher uncertainty related to for example the incidence of musculoskeletal diseases and mental disorders. Norway has no long-standing tradition of reporting and updating such data. Where the GBD project lacks updated Norwegian data, the calculations are partly based on data from the Nordic countries and other comparable countries in Western Europe. This must be considered when interpreting the disease burden estimates for Norway.
The way the severity of various diseases and injuries in GBD is calculated has been criticized. The GBD project measures severity using a disability weight which expresses the extent of the health loss associated with the disease or injury from 0 (“completely healthy”) to 1 (“one year of life lost”). One of the criticisms is that the disability weight attributed to blindness, deafness, amputations and paralysis of the legs may be based on the fact that these conditions are not necessarily understood as health conditions. Furthermore, it has been argued that one cannot simply compare very different health conditions with each other, and that what is measured by the disability weights is not health loss per se, but the values that are attributed to different types of health conditions [11]. Despite their weaknesses, however, the disability weights used by GBD are useful because they allow for comparisons both between diseases and between countries.
As with non-fatal diseases and injuries, Norway currently also lacks a system for regularly updating data on the occurrence of important risk factors and how they affect the risk of disease and death. For some topics, such as smoking, estimates are solid and based on many participants from many surveys collected over a long period of time. For other risk factors, such as several dietary factors, the evidence is more controversial and based on a weaker foundation.
The ranking of the contribution of each risk factor to the burden of disease also depends on whether one looks at aggregate categories (e.g. diet) or more detailed, individual factors (e.g. low fruit and vegetable intake), how reference values for the lowest possible risk are defined, and how risk factors can influence each other are accounted for. GBD results can also only be used for comparison between risk factors that are currently included in the GBD. Although 88 (groups of) risk factors are currently included, they do not cover all potentially modifiable health risk factors that exist.
Continuous quality assurance
An important task for the Centre for Disease Burden is to ensure high data quality for data used in GBD for Norway. This is done, among other things, by identifying areas where the Norwegian data base is weak and where improvements are needed, actively searching for existing relevant Norwegian data in these areas, and by supplementing GBD with new country-specific analyses.
The diseases, injuries and risk factors included in the GBD project are specified at a more detailed level than is usually found in general health statistics, and the lists are constantly being revised and expanded. Likewise, the methods and analysis models used in GBD are subject to continuous revision. The data base is also expanded with newly identified data sources from previous years. All analyses for all years are therefore re-run for each update of the GBD project, which means that calculations for years back in time may also change after adjustments to the analysis model or the underlying data. The GBD results must therefore not be considered final or complete. All results from GBD are freely available through online visualization tools on the website www.healthdata.org. Through collaboration with the Center for Burden of Disease at the Norwegian Institute of Public Health, some GBD tool are also available in Norwegian.