Expenditure and Health System Productivity by Disease and Injury in the US from 1996 to 2010

In the US, approximately 18 percent of the GDP is spent on health care, about 2.6 trillion dollars in year 2010. The National Health Expenditure Accounts- produced by the Centers for Medicare and Medicaid Services - track the overall health care spending in the US and its growth over time. However, these accounts do not break total health expenditure down by diseases and injuries. Nor is much known about how changes in health expenditure translate into changes in health outcomes at the level of specific diseases and injuries. Understanding health expenditure and trends in health expenditure at disease level is important as health outcomes are naturally measured at disease level. Information on expenditure and outcomes by disease or injury would facilitate evidence-based health policy formulation through more sharply focused policy design; enhance health system stewardship through sounder planning and budgeting; and improve health system performance management through greater transparency and accountability. In addition, by relating trends in disease expenditures to corresponding trends in health outcomes, health system productivity growth could be monitored and drivers of productivity improvement quantified. Such information could also pave the way for identifying the most cost-effective health care interventions.

The goal of this paper is twofold. First, we describe trends in US national health expenditures and corresponding health outcomes (measured in disability-adjusted life-years [DALYs]) for the past 15 years (1996-2010). Both expenditures and health outcomes are classified according to the diseases and injuries defined in the Global Burden of Disease 2010. Second, we estimate US health system productivity growth at the national level by analyzing the relationships between the trends in disease expenditures and the corresponding trends (suitably lagged) in health outcomes, using econometric methods to adjust for time-varying confounders.

We utilize (i) micro data based on encounter to estimate disease volumes and unit prices of services consumed, and (ii) macro data based on allocation of national health expenditure accounts (NHEA) to derive the envelopes for scaling the price and volume micro data. Specifically, data on inpatient bed days and average price per bed-day by disease were obtained from the National Hospital Discharge Survey (NHDS), and National Inpatient Survey (NIS), respectively. Total number of encounters for outpatient care were obtained from the National Hospital Ambulatory Medical Care Survey (NHAMCS) and the National Ambulatory Care Survey (NAMCS).  Average prices per outpatient visit by disease were obtained from the Medical Expenditure Panel Survey (MEPS). Data on other health care functions (services) were obtained from the National Nursing Home Survey (NNHS) and National Home and Hospice Care Survey (NHHCS) from CDC for long-term care; MEPS for pharmaceuticals and dental care; Mental Health Services and Substance Abuse Treatment Survey for specialized  psychiatric and substance abuse clinics; and appropriations and budget documents by program from the Health Resources and Services Administration (HRSA), CDC and a representative sample of state and county health departments for public health and disease prevention services.