Healthcare utilization and costs associated with cancer-related diarrhea.

Abstract

e18623 Background: Diarrhea is a common toxicity of cancer treatments, including radiotherapy, chemotherapy, and/or targeted therapies. Cancer-related diarrhea (CRD) leads to increased healthcare utilization and cost. This study evaluated the all-cause and CRD-specific healthcare utilization and cost of patients with CRD compared to a matched non-CRD cohort. Methods: We conducted a longitudinal observational study among adult patients ( > 18 years) with CRD using diagnosis codes or pharmacy claims compared to matched non-CRD patients using claims data from the IQVIA PharMetrics Plus database (October 2015 to March 2020). The index date was the date of the first cancer claim, and we re-indexed patients based on their CRD claim. Each patient had a 6-month pre-index period and a minimum 3-month post-index period. Patients were also required to have ≥12 months of continuous enrollment following the CRD index date. We directly matched patients 1:1 from the CRD cohort to the non-CRD cohort to adjust for selection bias and baseline differences. Our aim was to compare all-cause healthcare costs over a fixed 12-month post-index period, converting all costs to 2020 USD using the Consumer Price Index's medical component. We analyzed healthcare utilization for CRD-treated, CRD-inadequately treated, and CRD-untreated sub-cohorts (per Buono et al., J Econ 2017). Secondary endpoints included healthcare cost (proportion of patients, per-patient mean and median) and healthcare utilization (prescription fills and visits to the emergency department [ED], physician office, lab/pathology and outpatient ancillary services). We built one generalized estimating equation model with log link and gamma distribution adjusted for type of cancer, therapy and Charlson Comorbidity Index (CCI) to estimate the difference in total healthcare cost between CRD and non-CRD cohorts. Results: We evaluated a total of 104,135 matched pairs of CRD and non-CRD adult patients with solid or hematologic cancer receiving either targeted or chemotherapy, with 12-month continuous enrollment. Patients with CRD incurred significantly higher mean ($104,880 vs $39,664, p < 0.0001) and median ($59,969 vs $8,914, p < 0.0001) all-cause healthcare cost compared to patients without CRD over the 12-month post-index period. Inadequately treated CRD patients had the mean highest cost ($129,531) vs adequately CRD-treated ($107,050) or untreated CRD patients ($56,350) (all p < 0.0001). Mean pharmacy cost for CRD and non-CRD patients were ($35,190 vs $15,883); visits to the ED ($1,107 vs $431), physician office ($3,457 vs $2,058), lab/pathology ($4,074 vs $1,404), and outpatient ancillary services ($15,805 vs $4,940) (all p-values < 0.0001). Conclusions: Our findings show that patients with CRD had nearly 2.9 times higher all-cause total cost than patients without CRD after adjusting for covariates. Prevention of CRD may result in a significant reduction in cancer-treatment cost.