Adults with diabetes enrolled in Medicaid had higher drug utilization and lower drug costs compared with adults in subsidized Marketplace plans. These findings, from a sample cross-sectional study, were published in JAMA Network Open.
Insurance claim data were sourced from the Colorado all payer claims database (APCD) and income information from the state’s Medicaid and Marketplace programs. Adults (N=22,788) with diabetes aged 19-64 years with incomes 75%-200% of the federal poverty level (FPL) were assessed for filling 8 categories of prescription medications (dipeptidyl peptidase 4 [DPP-4] inhibitors, glucagon-like peptide 1 [GLP-1] agonists, sodium-glucose transport protein 2 [SGLT-2] inhibitors, sulfonylureas, metformin, basal insulin, meal-time insulin, and combination insulins) and total and out-of-pocket costs in 2014-2015.
People eligible for Medicaid (n=20,245) or subsidized Marketplace plans (n=2543) were aged mean 47.70 (standard deviation [SD], 11.33) and 52.12 (SD, 10.60) years, 61.4% and 55.6% were women, income was 106% (SD, 18.50%) and 168% (17.49%) of the FPL. The 2 groups had been covered by insurance for 10.74 (SD, 2.45) and 9.73 (SD, 2.98) years, and their Elixhauser comorbidity index was 0.71 (SD, 1.39) and 0.99 (SD, 1.56), respectively.
For the 5 noninsulin antidiabetic medications, there was no difference between the 2 groups for overall usage (adjusted difference [AD], 0.8; 95% CI, -1.8 to 3.5; P =.54). However, the Marketplace cohort filled fewer brand name prescription medications (AD, -4.7; 95% CI, -6.6 to -2.7; P <.001) as well as DPP-4 inhibitors (AD, -3.7; 95% CI, -5.3 to -2.1; P <.001) and sulfonylureas (AD, -6.6; 95% CI, -8.9 to -4.3; P <.001).
A similar trend was observed for the 3 insulin medications, in which there was no group difference for overall medication usage (AD, -2.3; 95% CI, -5.1 to 0.5; P =.11). Fewer of the Marketplace cohort filled Levemir (AD, -12.9; 95% CI, -15.2 to -10.6; P <.001) and NovoLog (AD, -5.2; 95% CI, -6.9 to -3.6; P <.001) prescriptions, but filled more Humulin N (AD, 5.0; 95% CI, 3.8-6.2; P <.001) and Humulin R (AD, 2.1; 95% CI, 1.3-2.9; P <.001) prescriptions than the Medicaid group.
Total monthly costs were higher for patients insured through the Marketplace for all drug classes, except sulfonylureas. Adjusted differences ranged from $2.77 (95% CI, $1.70-$3.83) per prescription for metformin to $361.19 (95% CI, $251.90-$470.48) per prescription for glucagon-like peptide-1 (GLP-1) receptor agonists.
Average total pharmacy out-of-pocket costs per year were $224.14 (95% CI, $213-69-$236.59; P <.001) higher among individuals insured through the subsidized Marketplace.
Medication adherence, as quantified by the number of active months divided by months of coverage, was similar among both cohorts, except for sulfonylureas, which was 5.3% (P =.04) higher among the Marketplace group than the Medicaid group.
The researchers did not have access to clinical information such as glycemic control, making it infeasible to assess adequacy of care to people in the study groups.
Study authors conclude that these data indicate that, among adults with low incomes in the United States, prescription drugs for diabetes were substantially less costly for individuals enrolled in Medicaid compared with those covered by subsidized Marketplace plans. This trend was particularly relevant for comparatively more expensive brand-name prescription drugs. The authors said these findings suggested “Medicaid drug pricing rules led to better control of pharmaceutical costs for these treatments than did Marketplace drug pricing rules.”
Disclosures: This research was supported by the Robert Wood Johnson Foundation, and several authors received fees from other organizations. Please see the original reference for a full list of authors’ disclosures.
Reference
Khorrami P, Sinha MS, Bhanja A, Allen HL, Kesselheim AS, Sommers BD. Differences in Diabetic Prescription Drug Utilization and Costs Among Patients With Diabetes Enrolled in Colorado Marketplace and Medicaid Plans, 2014-2015. JAMA Netw Open. 2022;5(1):e2140371. doi:10.1001/jamanetworkopen.2021.40371