Like all survey-based research, the MHRS relies on self-reported information.
Given this limitation, we draw on the findings from studies using evaluation methods that offer the potential for controlling for such confounding factors as a check on the findings from the MHRS.
These pre/post findings on insurance coverage from the MHRS are supported by other studies using national survey data and stronger quasi-experimental designs.
Other studies of the impacts on health care access and affordability using national survey data also support the gains found in the MHRS, although the work generally has been constrained by short follow-up periods for tracking changes in access to care and a limited set of measures.
Those patterns are consistent with the ED reductions for non-emergency care reported for Massachusetts in the MHRS.
While these questions were not included in the MHRS prior to health reform, the NHIS shows some increases in the delays in obtaining needed care because of difficulty getting an appointment under health reform in Massachusetts (Long and Stockley 2011).
Since the MHRS obtained family income relative to the FPL based on a limited number of categories in each year (e.
Beyond that, we know that there are differences in the income measures and the survey years that will introduce measurement error in our estimates, as well differences in our ability to define family in the two surveys (based on immediate family in the MHRS versus all related persons in the household in the NHIS).
As noted above, combining the NHIS and MHRS introduces some limitations.
The 2010 MHRS expanded the sample frame of the survey to include both landline and cell-phone households.
That income break was only added to the MHRS in fall 2010.
Low-income is defined as family income below 100% of the federal poverty level (FPL) for the MHRS sample and at or below 138% of FPL for the NHIS sample.