Regarding "Mass. patients face cost barriers to care after reform: study
The article reporting the study of access to care in Massachusetts should point out the obvious bias in the sample. If you select only patients visiting an ER department, you are much more likely to get those having trouble accessing care than if you draw a random sample.
The fact that the sampling technique is seriously flawed makes the results unreliable, which should be noted. Researchers use this sort of approach because it is more convenient than valid sampling.