Author: Mat Reidhead
Last week, I hit the data-nerd power-ball. I received a notice in the mail informing me that by law, I needed to complete the Census Bureau’s American Community Survey. The ACS is an annual random sampling of American homes that has been called the most important government function you’ve never heard of. This survey enables policy makers and community planners to monitor the population in terms of income, education, employment and other societal trends at an interval more frequent than the decennial census allows. The ACS randomly selects around 300,000 of the nation’s 134 million households every month to complete a brief questionnaire. This means my odds of being selected were around 2 in 1,000. So maybe I hit the data-nerd scratch-off $100 winner, but nonetheless, I was unbound by the shackles of my assumed existence by the government. The size and social conditions of the other 998 households will be estimated with the statistically generalizable properties of the ACS’ underlying sampling frame. Their existence will be assumed, based in part on my answers.
The hospital community in Missouri is very fortunate to have a strong data program that enables us to present facts in place of assumptions in support of the issues that are important to the health of hospitals and the people they care for. Also last week, we used these data to present some interesting findings on Medicaid managed care to the MO HealthNet Oversight Committee. While we were discussing the chart below, Dr. Tim McBride, committee chair, leaned over and said “it’s the 80-20 rule,” which reminded me of some fundamental principles of microeconomic theory that he bestowed on me 10 or so years ago in his Econ-251 class.
Vilfredo Pareto was an Italian economist at the turn of the 20th century who bears the distinction of being the grandfather of inequality research. Pareto observed inequalities in everything — from individual plant production in his garden to land ownership in 19th century Italy. These observations gave way to the Pareto Principle, or the 80-20 rule — 80 percent of effects come from 20 percent of causes.
Like Pareto’s vegetable garden, health care is not granted exception by the 80-20 rule. In general, 80 percent of health care resources are consumed by 20 percent of patients. But, there are patient populations where you’d expect exceptions to the rule. For example, you might expect a more even distribution of health care resource consumption among a relatively young and healthy patient population consisting primarily of mothers and children who have the benefit of a well-resourced organization being paid to manage their care. One would at least expect a more even distribution for this group compared to another population with large numbers of high-risk aged, blind and disabled patients — i.e. you would expect the curve for the healthier group to be much closer to the line of equality in the chart above. But, you’d be wrong. The 80-20 rule is steadfast. It is at least for the Medicaid managed care population in Missouri. The Lorenz curve and Gini coefficient (developed by two other economists inspired by Pareto’s work on inequalities) for Medicaid managed care hospital utilization in Missouri are nearly identical to Medicaid fee-for-service. This is despite large known differences in the medical complexity and availability of resources dedicated to care management for the two patient populations.
Like the ACS, proponents of Medicaid managed care in Missouri are resource-constrained in a way that forces them to rely on some fairly heavy assumptions. Thanks to the data submitted by Missouri hospitals to MHA’s Hospital Industry Data Institute, we don’t need to assume. Data points, such as “over the last ten years, Medicaid managed care ED visits for mental health and substance abuse have more than doubled in Missouri,” are facts, not estimates. That actually happened and it raises powerful questions for policy makers to consider when judging how well Medicaid MCOs are doing what they’re paid to do — manage care.