Three states (Maine, Arizona and New York) in the 2000-2005 time frame increased Medicaid eligibility to mostly include childless adults meeting a variety of poverty thresholds. The authors compared changes ("pre-post") in publicly available death rates and health status statistics in these "intervention" states to neighboring states that acted as quasi-experimental "controls (New Hampshire for Maine, Nevada and New Mexico for Arizona and Pennsylvania for New York).
Over time, new Medicaid enrollees were slightly older (40.6 years vs. the average of 40 years), more likely to be male (57% vs. 49% in the general population), nonwhite (27% vs. 20%) and in fair or poor health (20% vs. 11%).
What was interesting was that the authors compared the county-level changes in mortality for the entire state population, not just Medicaid enrollees. Using standard statistical methods to account for baseline differences, the authors found that adjusted all-cause mortality for the intervention states declined by 19.6 per 100,000 versus the control states. Since it takes time for Medicaid enrollment to actually increase after a change in eligibility, the authors also examined the impact over time. They found a strong statistically significant correlation between growing Medicaid enrollment and mortality.
Medicaid expansion was also associated with decreases in rates of patient surveys showing that there was "delayed care" and increases in self-reported excellent or good health status.
The obvious conclusion of the study was that expanding Medicaid eligibility allows persons who are otherwise without insurance to access the health system and receive care for conditions that would otherwise kill them. The DMCB finds the results convincing and should inform the debate in some states about the life-saving merits of expanding Medicaid.
Critics could quibble that unknown factors not captured by the study could have accounted for the observed differences (did pumping more Medicaid money into the system enrich hospitals, enabling them to provide a higher level of care?) and that association does not prove causality (could booming state economies lead to a healthier population, while a generous Medicaid expansion had nothing to do with it?).
DMCB questions 1) This study doesn't compare Medicaid insurance vs. commercial insurance. If there were a way to use the commercial markets (for example, vouchers), would patients far better? There is research that suggests the answer could be yes .
2. A cruel but important question: how much did it cost? Expanding Medicaid did not save money, it cost and it would be interesting to know the cost per person, per person-year or per quality adjusted life year .