This is Greg Scandlen, writing at the State Policy Network blog site here:

This author reviewed all of the 139 studies that comprised [Institute of Medicine's] analysis and found that only seven of them adjusted for income, but 44 identified the results of Medicaid enrollees separately from the uninsured and 26 compared the uninsured only to people with private insurance, omitting the role of Medicaid and Medicare. In 31 of the 44 studies that separated out the Medicaid experience, people on Medicaid did worse than the uninsured on a range of health treatments and outcomes. In a few cases, the uninsured and Medicaid patients both did better than the privately insured, such as mortality in the hospital.