Burden of insomnia among cancer patients undergoing treatment: A retrospective claims database analysis.

e18816 Background: Almost half of all cancer patients have trouble sleeping, usually insomnia or other sleep-wake disorders. Cancer treatment can exacerbate sleep problems. This study evaluated all-cause healthcare resource utilization (HCRU) and costs in cancer patients undergoing treatment who also have insomnia relative to a matched cohort of cancer patients undergoing treatment who do not have insomnia. Methods: This retrospective study used the IBM MarketScan Commercial and Medicare Supplemental Databases (Jan 2014-Dec 2019) to identify patients: (1) age ≥18 years; (2) with ≥1 ICD-9/ICD-10 codes for insomnia OR with ≥2 prescriptions for zolpidem immediate release (IR) OR trazodone ≤150mg OR benzodiazepine; (3) with ≥12 months of continuous enrollment pre-/post-index date (earliest diagnosis or medication fill date); (4) with ≥2 ICD-9/ICD-10 codes for any cancer; and (5) ≥1 claim for oncology treatment. Patients with sleep disorders other than insomnia were excluded. The resulting patients (“insomnia cohort”) were 1:1 matched on age, sex and Elixhauser Comorbidity Index (ECI) score to patients with cancer undergoing treatment without insomnia or other sleep disorders (“control cohort”). HCRU and costs were reported per patient per month (PPPM). Generalized linear models were used to compare adjusted outcomes between cohorts. Results: For each cohort, 7,132 patients (mean age 60.8 years, 69.5% female, 13.6 ECI score) were identified. In adjusted analyses, significantly more patients in the insomnia cohort had at least one inpatient visit per month, inpatient stays were significantly longer, and mean outpatient visits were significantly higher (all p < 0.001) relative to the control cohort. The insomnia cohort had higher mean number of emergency department (ED) visits (p < 0.05); however, mean number of inpatient visits was not significantly different between cohorts. Adjusted analyses also showed significantly higher mean total PPPM costs for the insomnia cohort than the control cohort ($4,864 and $3,302, respectively; mean ratio = 1.47; p < 0.0001). Mean PPPM costs for inpatient visits, ED visits, and outpatient costs were also higher for the insomnia cohort (p < 0.001). Conclusions: Cancer patients undergoing treatment who have insomnia had more intensive HCRU and higher costs relative to those without insomnia. These results suggest that better managing insomnia in this population may have a beneficial impact on both patients and payers.