Institute for Health Metrics and Evaluation

In the Global Burden of Diseases, Injuries, and Risk Factors (GBD) Study 2010, a worldwide consortium of 500 researchers, coordinated by IHME measured the impact of more than 290 health conditions and 67 health risk factors worldwide.[30] The GBD enterprise produced estimates in 21 regions around the world for disability-adjusted life years by age and sex for the past two decades. Part of this research has involved conducting in-person surveys in several countries and gathering health information through a website survey. The team created a cause of death database that includes 60 years of data, or almost 800 million deaths. To illustrate the findings, IHME released a suite of interactive data visualizations, which are available to the public.

The aim is to allow policymakers and other decision-makers to "compare the effects of different diseases – such as cancer versus depression – that kill people prematurely and cause ill health," to show disease trends over time, and to inform policy, IHME states on its website.

It has been labeled "the most comprehensive review of the state of humanity's health ever undertaken."[31]

The UK and China are among the countries working with IHME to generate subnational burden of disease estimates at the county and province levels.

In January 2014, IHME began releasing updates to the work, called the Global Burden of Diseases, Injuries, and Risk Factors (GBD) Study 2013. The work generated over 1 billion estimates of health outcomes.[32] GBD estimates are now being updated annually.

Some of the findings have been controversial. In 2012, IHME researchers estimated 1.2 million people died from malaria in 2010, double the WHO's estimate.[33][34]

The 2010 Global Burden of Disease report was criticized for its lack of transparency of methods and data as well as its use of complex statistical methods to fill in data gaps when reliable statistics were unavailable. Thomas Bollyky, a senior fellow at the Council on Foreign Relations, called for more transparency. The report was also criticized for its use of verbal autopsies because many diseases have similar symptoms, leading to potential mistakes.[31][35]

Igor Rudan and Kit Yee Chan[36] note that IHME "struggled to generate support, legitimacy, and acceptance for their findings" after publishing the 2010 Global Burden of Disease Study, due to concerns of lack of transparency, as well as the existence of parallel estimates of disease burden from WHO; but argue that the emergence of IHME introduced competition to the field of global health metrics, which was previously a field where WHO maintained a monopoly: "The GBD initiative has emerged as a well-organised and rapidly growing collaboration that is now seriously challenging WHO's role in generating global health estimates".

The World Health Organization did not acknowledge the GBD 2010 estimates.[15][36]

The Global Burden of Disease Study 2013 expanded collaboration and increased transparency but "[fell] short of allowing full independent replication of all results".[36]

In May 2018, WHO and IHME signed a memorandum of understanding, agreeing to strengthen collaboration on the GBD and enhance policy use of GBD findings.[37]

The Disease Control Priorities Network project generates cost-effectiveness estimates for a range of health interventions. A team of demographers, statisticians, economists, and other experts are studying how to improve the allocation of resources among interventions, technologies, hospitals, and other service delivery platforms. They are working in multiple countries, including the US, India, and South Africa. Their cost-effectiveness work has revealed some hidden connections. For example, a 2010 report showed improving girls' education in poor countries is the most effective way to reduce child mortality,[38] which was a surprise for some people.[39]

An effort launched in 2011 called Access, Bottlenecks, Costs, and Equity (ABCE) involves collecting evidence and analyzing data to improve the cost-effectiveness and equity of health systems in Colombia, Ghana, India, Kenya, Lebanon, Uganda, and Zambia. The project examines four components that affect health care delivery: access (to health facilities), bottlenecks (limitations on the supply side), costs, and equity (across populations). The work includes in-depth facility surveys and inventories across a range of service delivery platforms; additional data is collected in countries with high HIV/AIDS burdens through exit interviews and chart extraction.

A 2013 report concluded that neonatal and maternal mortality in Ghana could be best improved not with sophisticated maternity care but with better transportation infrastructure.[40]

The Population Health Metrics Research Consortium created new methods for tracking health intervention coverage in low-resource settings. The methods have been used to measure mortality, causes of death, and incidence of major illnesses where data are incomplete.[41] Researchers collected data in India, Mexico, the Philippines, and Tanzania.

IHME's work on estimating mortality has been at times controversial. In 2009, IHME published a study on maternal mortality[42] which some advocacy groups tried to suppress – worried that results showing a decline in mortality would make it harder to fund-raise. The WHO also initially disagreed with the new results, but later revised their estimates in agreement with those of IHME.[43]

The Malaria Control Policy Assessment project evaluates the effectiveness of malaria-control interventions in Uganda and Zambia by analyzing their effect on child mortality and producing estimates at the national and local levels.

IHME conducts US research, including estimates of mortality, life expectancy, risk factors, health disparities, and disease prevalence. IHME has compiled national and local health trends and integrated multiple data sources to monitor disparities in chronic diseases.

A 2013 report, the State of US Health,[44] looked at trends in premature deaths due to injury or disease, and demonstrated the major health threat stemming from behavioral risk factors such as poor diet and sedentary lifestyles. It concluded that dietary factors cause more deaths each year than cancer or smoking.[45] The results included life expectancy trends broken down by state and county.

First Lady Michelle Obama cited the research in her campaign to improve Americans' diets and increase their level of physical activity.[46]

A paper published in the journal Population Health Metrics in June 2011 showed that life expectancy was rising in some poorer US counties – especially in the South[47] – a surprising result which was widely discussed.[48]

Other research projects include the Salud Mesoamérica 2015 Initiative, which focuses on inequalities in health outcomes and access in southern Mexico and Central America; Gavi Full Country Evaluations, which evaluates immunization programs in Bangladesh, India, Mozambique, Uganda and Zambia; and HealthRise, a partnership with Medtronic Philanthropy to evaluate programs targeting diabetes and heart disease, and sponsor small grants to make improvements.

In 2014, IHME announced the establishment of the University of Washington Center for Demography and Economics of Aging, funded by the National Institute on Aging (NIA). It is one of 14 NIA Demography Centers at leading universities and policy organizations around the United States.

In March 2011, IHME launched the Global Health Data Exchange (or GHDx), which indexes and hosts information about microdata, aggregated data, and research results with a focus on health-related and demographic datasets. At launch, the site listed about 1,000 datasets; as of 2015, there are more than 30,000. As part of a partnership with the Centers for Disease Control and Prevention (CDC), GHDx includes 35 years of CDC data on child and maternal health. GHDx uses the Drupal 7 open source content management system and Apache SOLR for search. The site includes visualization and GIS tools, and has been noted by the health and global health communities.[49]

In 2020, IHME published its model projecting deaths from the COVID-19 pandemic in the US[28][29] which was described as widely influential,[50][51][52] and informed guidelines developed by the Trump administration.[53][3] The model predicted results significantly different from other models,[51][52] and some epidemiologists urged caution in interpreting their model,[50] which is described by some as "optimistic".[53] The model received heavy criticism from some members of the epidemiological community for being flawed and misleading.[3][54][55]

IHME released two major updates to its model in April 2020 and June 2020 to improve its accuracy and reflect different scenarios of social distancing and mask usage. IHME described its updated model as a "hybrid" of the statistical (curve fit) model the institute initially released and a mechanistic model grounded in epidemiological understand of the virus.[56] The hybrid model predicted dramatically more deaths than the curve-fit model and had much better out-of-sample prediction.[57] On June 7, 2020, IHME published its first projections of COVID-19 deaths out to November 1 and forecasted 208,255 deaths (with a range of 186,087 to 244,541) due to the COVID-19 pandemic in the United States. Those numbers drop to 162,808 (157,217 to 171,193), if at least 95% of people wear masks in public.[58]