According to this article, the answer is:
“Both policies decreased medication adherence. The days’ supply policy [decreasing the days supply of each prescription from 100 to 34 days] had a much larger effect on adherence than did the copayment increase. Total Medicaid spending declined from the days’ supply policy, but the copayment policy resulted in a net increase in Medicaid expenditures.”
If someone is very sick, the time/inconvenience cost to refill a prescription seems to be a more important factor than the out-of-pocket cost. This is true even for the Medicaid population, which is of course made up of exclusively low income individuals. Many economists measure the elasticity of demand with respect to price, but health economists may need to start constructing a demand elasticity measure with respect to inconvenience.