highly significantP < 0.001
Adding the health variables (Model 2) reduced the difference in fills between the non-enrollees and LIS recipients by just over a quarter to −16.9, a difference that remained highly significant ( P < 0.001).
Adding the health variables (Model 2) reduced the difference in fills between the non-enrollees and LIS recipients by just over a quarter to −16.9, a difference that remained highly significant ( P < 0.001).