Global health financing and the need for a data revolution

Although knowledge about global health financing has expanded over the past two decades, major gaps remain. We know little, for example, about how much governments spend on major disease areas, how these amounts have evolved over time and how countries compare. A global health financing data revolution is sorely needed.

The suite of papers that make up this special issue underline the importance of reliable information about financial resources for health. Data on health financing provide a foundation for assessing the provision of health care services, global public goods and resource mobilization for improving health. In a world of scarce resources, competing priorities, persistent inequality and increasingly complex health systems, allocating resources for health effectively is more difficult – and essential – than ever.

Granular, comparable and comprehensive health data can inform health system decision-making. These data could be used to better understand health systems, identify gaps and inefficiencies, assess equity and provide estimates of resources needed to reach prospective health goals. The lack of comprehensive, comparable health financing information is a roadblock that prevents robust health policy-making.

What health financing data are currently available?

There are five basic types of data available, each furnishing essential information to policymakers, although each type is also limited in key ways.

The most widely used, internationally-comparable data on health spending is the WHO’s Global Health Expenditure Database (World Health Organization: WHO). This database captures national health spending disaggregated by the source of funds and financing agent, including external, out-of-pocket (OOP) and government spending. Expenditure on curative and rehabilitative care, and prevention and public health services is also reported. However, more detailed data on the type of care, subnational unit and health focus area are lacking. The data set could also benefit from more robust estimation and methodological transparency.

National Health Accounts (NHAs) also comprise an important source of health financing data. While recent investments will likely increase the depth, reliability and frequency of these reports, historically, the framework has not been systematically or comparably applied (Bui et al., 2015). Health accounting is challenging and wrought with important assumptions that are not necessarily transferred from one health accounting team to the next. NHAs continue to require technical expertise and resources that make conducting them on a regular basis difficult for low- and middle-income countries. Additional effort is needed to increase the comparability and usability of NHAs across time and countries.

The third, core source of health financing data captures development assistance for health. Development assistance for health data is produced annually by the Institute for Health Metrics and Evaluation, with international spending on health broken down by approximately 20 program areas, all low- and middle-income countries and more than 25 yr (Institute for Health Metrics and Evaluation (IHME), 2016). The comparability and reliability of these data are strong, but development assistance makes up a relatively small share of health spending in many countries. Development assistance for health is also not disaggregated by type of care or subnational unit.

A fourth set of data consists of the disease-specific financing annually reported to international organizations such as UNAIDS (Joint United Nations Programme on HIV/AIDS), the WHO and others. These data provide a foundation off which we can understand how much is spent domestically on major health focus areas, such HIV/AIDS, vaccinations, maternal and child health, and malaria. However, the rigor of the tracking underpinning these data vary widely. Furthermore, these efforts often operate in silos. Data reporters are not forced to divide funding among different health focus areas. This likely leads to double-counting across areas of spend and makes compiling data in a cross-country time series infeasible.

Finally, an ever-growing set of surveys and country-level resource tracking exercises also capture financial resources for health. Public expenditure reviews, household surveys, compilations of administrative data and other data collection efforts are important sources for estimating OOP spending as well as breaking down spending by different types of care. However, these exercises are conducted intermittently and thus are not available for all countries for all time periods. Furthermore, their ad-hoc nature and the lack of standardization limits comparability.

What is needed?

We believe that simplification and harmonization of health financing data are sorely needed. One, simple process for producing the information local and global policymakers need would greatly reduce the effort and time required of health officials. Procedures can be put in place to harness country-specific accounting systems and prepare administrators to report data on an annual basis. If the same data are required year after year, country-specific methods for addressing the complexities of existing systems can be developed.

The categories of expenditure collected should be simple and justified based on their applicability to a range of local and global health issues. Tying break-downs to a critical but limited set of health outcomes is also fundamental. We propose focusing on four core elements:

1. 1. Health focus areas. Capturing the most important health focus areas in a comparable and comprehensive fashion is vital to connecting dollars with outcomes, and understanding whether spending aligns with burden of disease. Generating data on these health focus areas together will force accountants to decide where each dollar goes.

2. 2. Type of care. Wide categories of types of service could be also easily captured, including: inpatient, ambulatory/outpatient, pharmaceuticals, administration and public health. Characterizing the distribution of funding across these areas can help health officials understand important system-wide characteristics, including inefficiencies, and under- and over-utilization.

3. 3. Payer. The payer is also fundamental, as a representation of the use of pooled vs non-pooled funding can affect utilization and efficiency substantially. We propose dividing funding flows into: government, insurance/pre-paid and OOP spending. We recognize that, critically, OOP may require additional estimation or household surveys which may not be feasible on a regular basis. However, this category of payer is vital to understanding the weight of catastrophic expenditure and medical impoverishment.

4. 4. Subnational unit. Depending on the context, it may be essential to develop state- or province-level expenditure. The benefit of these data are that they can be linked to the administrative level at which decisions are made.

Standardization and timeliness of production will also be important to making these data useful for the global health community. Producing data in a standardized fashion will ensure flows can be compared and contrasted across countries and time. Timeliness allows stakeholders to take action in close-to real time.

This endeavor will require up-front investment and in some cases technical assistance. Generating consensus in the health financing community will be challenging; it will not be possible to capture all categories of spend every year, and more granular, intervention- and disease-specific sub-categories may have to be excluded. Supporting administrators in developing a standardized approach to build off data produced through existing financial systems will also require technical assistance. However, if efforts are pooled across reporting mechanisms, time and resources will be reduced overall.

The lack of comparable and comprehensive health financing estimates limits our ability to make evidence-based decisions in the health sector. These important health financing data will equip decision-makers in low-, middle- and high-income countries to make better policy for health. A foundation rooted in comparable and robust health financing data can provide important evidence for the progressive realization of universal health coverage and essential health gains through greater transparency, accountability and efficiency.