State-Level Medicaid Reimbursement and Imaging Utilization by Medicaid and Children's Health Insurance Program Patients.

Purpose: To determine if relative Medicaid-to-Medicare reimbursement rates are associated with patient imaging utilization.

Methods: This cross-sectional study estimated the association of diagnostic imaging utilization with the state-level Medicaid-to-Medicare reimbursement ratio (MMRR) of professional payments. State-specific reimbursement ratios were computed for each imaging modality. Logistic regression was used to estimate the likelihood of having imaging, and gamma regression was used to estimate the average number of imaging studies for those with imaging. These models were performed for each gender-modality combination controlling for patient characteristics.

Results: Among 48,835,765 Medicaid patients, 54.3% were women. The median MMRR was 0.82 (interquartile range [IQR]: 0.73-0.94) for CT, 0.87 (IQR: 0.76-1.01) for MR, 0.76 (IQR: 0.69-0.99) for nuclear medicine (NM), 0.85 (IQR: 0.73-1.09) for ultrasound, and 0.82 (IQR: 0.74-0.97) for radiography or fluoroscopy (XR). The probability of having imaging was 25.9% for CT, 25.9% for MR, 21.4% for ultrasound, and 31.8% for XR higher at 75th percentile of the MMRR distribution compared with the 25th percentile (P < .001). For those with imaging, the mean number of imaging studies received was associated with 5.7% fewer studies for NM at the 75th percentile compared with the 25th percentile (P < .001), although there was no difference for other modalities.

Conclusions: Medicaid payments are related to imaging utilization. A higher MMRR is associated with a substantially increased likelihood of Medicaid patients receiving CT, MR, ultrasound, and XR imaging but no difference in the amount of imaging studies received for those with imaging for these modalities.