Must-Read: Benjamin D. Sommers, Atul A. Gawande, and Katherine Baicker: Health Insurance Coverage and Health

Must-Read: Benjamin D. Sommers, Atul A. Gawande, and Katherine Baicker: Health Insurance Coverage and Health: “Perhaps no research question better encapsulates this policy debate than, ‘Does coverage save lives?… http://www.nejm.org/doi/full/10.1056/NEJMsb1706645#.WUriDI_R7Z4

…Beginning with the Institute of Medicine’s 2002 report Care without Coverage, some analyses have suggested that lack of insurance causes tens of thousands of deaths each year in the United States…. The Oregon study… was limited by the infrequency of deaths in the sample. The estimated 1-year mortality change… confidence interval of −82% to +50%…. Several quasi-experimental studies using population-level data and longer follow-up offer more precise estimates…. Compar[ing] three states implementing large Medicaid expansions in the early 2000s to neighboring states that didn’t expand Medicaid… a significant 6% decrease in mortality over 5 years of follow-up… [with] the largest decreases were for deaths from “health-care–amenable” conditions such as heart disease, infections, and cancer…. Massachusetts’ 2006 reform found significant reductions in all-cause mortality and health-care–amenable mortality… particularly [among] those with lower pre-expansion rates of insurance coverage… a “number needed to treat” of 830 adults gaining coverage to prevent one death a year. The comparable estimate in a more recent analysis of Medicaid’s mortality effects was one life saved for every 239 to 316 adults gaining coverage….

[Why] the nonsignificant cardiovascular and diabetes findings in the Oregon study?… First… hypertension, dyslipidemia, and elevated glycated hemoglobin levels are important clinical measures but do not capture numerous other causes of increased risk of death. Second… hundreds of thousands of people gaining coverage over 4 to 5 years of follow-up, as compared with roughly 10,000 Oregonians gaining coverage and being assessed after less than 2 years. It may take years for important effects… to manifest in reduced mortality…. Third… changes in self-reported health—so clearly seen in the Oregon study and other research—are themselves predictive of reduced mortality over a 5- to 10-year period… a 25% reduction in self-reported poor health could plausibly cut mortality rates in half (or further) for the sickest members of society, who have disproportionately high rates of death. Finally, the links among mental health, financial stress, and physical health are numerous,45 suggesting additional pathways for coverage to produce long-term health effects…