Cost of Acute and Follow-Up Care in Patients With Pre-Existing Psychiatric Diagnoses Undergoing Radiation Therapy.

PURPOSE: The impact of psychiatric comorbidities on the cost of cancer care in radiation oncology practices is not well studied. We assessed the acute and 24-month follow-up costs for patients with and without pre-existing psychiatric comorbidities undergoing radiation therapy.

METHODS AND MATERIALS: Patients with cancer undergoing radiation therapy at our institution from 2009 to 2014 were denoted as having pre-existing psychiatric conditions (Psych group) if they had associated billing codes for any of the 422 International Classification of Diseases, 9th revision psychiatric conditions during the 12 months before their cancer diagnosis. The Elixhauser comorbidity index was calculated, excluding psychiatric categories. Medicare reimbursement was assigned to professional services, and Medicare departmental cost-to-charge ratios were applied to service line hospital charges and adjusted for inflation to create 2017 standardized costs. Acute (0-6 month) and follow-up (6-24 month) costs were subcategorized into clinic, emergency department, hospital inpatient, and outpatient costs.

RESULTS: Among 1275 patients, 126 (9.9%) had at least 1 pre-existing psychiatric diagnosis. On univariate analysis, both acute and long-term costs were higher in the Psych group. The largest significant differences in costs were follow-up hospital inpatient costs ($5861 higher; 95% confidence interval [CI], $687-$11,035; P = .002), follow-up hospital outpatient costs ($2086 higher; 95% CI, -$142 to $4,314; P = .040), and follow-up emergency department costs ($396 higher; 95% CI, $149-$643; P < .001). Age, race, sex, and treatment modalities were comparable, but the Psych group patients had more median comorbidities (2 vs 1) and had more respiratory cancer diagnoses than the nonpsychiatric group (31% vs 17%). On multivariate analysis adjusted for age, sex, cancer diagnosis, and comorbidities, global follow-up costs remained 150% higher in the Psych group (P < .001). Acute costs were similar after adjustment (P = .63).

CONCLUSIONS: Psychiatric comorbidities independently predict elevated healthcare costs in patients treated for cancer. Radiation oncology payment models should consider adjustments to account for psychiatric comorbidities because addressing these may mitigate cost differential.