Understanding health care data

Managing uncertainty in health care has created a demand for actionable data to improve quality outcomes and manage growing expenditure. The term “big data” is often used to describe the overwhelming amount of data that is produced and available to large corporations. In an increasingly digitized medical world, the constant stream of big data is growing. Raw data can be unclear, disorganized and can add undo stress to physicians. Even once it has been organized, some numbers and costs mean nothing until comparisons can be made and conclusions are drawn. How can this data be translated into helpful insights?

In 2020, the U.S. health care spend as a percent of GDP climbed to an all-time high of 19.7 percent. Although this steep increase was, in part, driven by the pandemic and subsequent government-issued relief packages and public health initiatives, health care costs have been increasing steadily at an average annual growth rate of 4.2 percent from 2010-2019.

The largest share of the total national health expenditures by type of service in 2020 was attributed to hospital care at 31 percent. Hospitals consume resources based on case mix, which refers to the “mix” or burden of illness of patients being treated at a health care facility. A sicker population, in most cases, uses more of a hospital’s resources and is thus more costly, while a healthier population is less costly. Consider the population of a major medical center such as Mayo Clinic compared to a rural hospital setting. For this reason, it is important to evaluate the sickness or risk of a population before predicting potential expenses.

Expected rates are used in calculating utilization rates for a population based on burden of illness. These rates are used in multiple ways to provide relativity of utilization. Expected rates may be based on facility type, client specific needs or other factors, but the same ideals remain true in each case. Calculating the risk of a population and establishing expected utilization serves as a benchmark to measure how a facility or physician group is performing in quality and expenditure compared to a larger population.

To maximize efficiency and costs in health care, the target spend should be at or below the expected amount. The expected amount, however, will not be the same for each hospital. Analysts and consultants use formulas to show which hospitals’ expected amount should be higher and which should be lower based on the case mix. A simple percent change calculation is then used to demonstrate what percent above or below expected a facility is, based on the expected amount.

The scenario below can help illustrate the importance of expected amounts:

Hospital A
Actual spend: $100
Expected rate: $80

Hospital B
Actual spend: $100
Expected rate: $120

Although, in this example each of these hospitals spent $100, since Hospital A has a healthier population, its expected amount is only $80. This means that Hospital A, given its case mix, could have potentially been more efficient with its resources. Hospital B, on the other hand, spent less than expected given its sicker case mix.

Expected rates give health care administrators an ability to look at risk-adjusted performance of facilities and provider practices in comparison to peers. Payers can use this information in pay for performance models. Data analysts and consultants use health care data to show where facilities may have room for improvement based on the population’s case mix. Health care facilities that can keep costs at or below the expected amount seem to be using resources efficiently to care for its population. Those hospitals whose actual costs are over expected costs may have procedures or patients that it could be treating just as well with fewer resources

The ability to categorize, organize and analyze the health care data available to us can provide actionable solutions and valuable insights on patient health, utilization and cost. Analysts and consultants work together to help make sense of health care data so that conclusions can easily be drawn to promote effective resource management and positive changes in the health care industry.

Jessie Murphy, lead methodology consultant for the regulatory and payer solutions team at 3M Health Information Systems.

Jessica Houle, health care data analyst for the regulatory and payer solutions team at 3M Health Information Systems.