OBJECTIVES: Investigate the utilization of cycled antineoplastic therapies (where dosages vary per monthly supply) for various cancers among Medicare-eligible patients, and medical costs and outcomes associated with discontinuation gaps in monthly adherence.

METHODS: This retrospective study used the MarketScan Supplemental and Coordination of Benefits Design databases from 2019-2020. Discontinuations were calculated for 2020 utilizing cycled medications (8 generic products) from 6 therapeutic categories, where discontinuation was defined as the first gap in medication supply during the last 6 months of the year at intervals of ≥1.5*prior supply. Continuous enrollment was required, as were ICD oncology diagnosis codes for those at least 65 years of age. New to therapy/diagnosis was inferred from 2019 metric files. Patients were excluded if hospice care and inpatient transplant services were indicated. Models examined adherence level (discontinued or not) and 10 covariates (including COVID-19 indications), with interaction terms, for significant associations with medical costs (inpatient and outpatient), hospitalizations, and length of hospital stay (LOS).

RESULTS: Patients who met metric criteria (n=1,357; Mean age 73.9) had a discontinuation rate of 16.7%. A total of 64.0% were new to therapy/diagnosis and 5.2% were identified having COVID-19 in 2020. However, COVID-19 was not significantly correlated with a discontinuation. Hospitalization rate was at 30.3% with mean LOS=9.8. Models found significant discontinuation effects with increased total medical costs ($11,977, p<.03), odds of hospitalization (2.7, p<.0001), and LOS (5.4, p<.03). COVID-19 patients had significantly higher total medical costs ($18,145, p<.007), odds of hospitalization (6.3, p<.0001), and non-oncology LOS (4.6, p<.009). Significant covariate interactions are discussed in the poster.

CONCLUSIONS: Medication adherence (fewer gaps in medication coverage) on cycled oral antineoplastics can lead to lower medical costs and fewer hospitalization events or LOS as compared to lower adherence levels. This was independent of COVID-19 indications that were associated with increased costs, hospitalization, and non-cancer LOS.