Radiation Oncology Alternative Payment Model (RO-APM) Monte Carlo Simulations to Predict Financial Impact and Viability for Common Practice Types

Abstract

The RO-APM will trial an episode-based payment schema over a five-year period starting in 2022. Practices were randomly selected for mandatory inclusion in the RO-APM. Participating departments have different case mixes, patients, and clinical practices that will dictate the relative impact of the RO-APM. It is our aim to create a highly versatile Monte Carlo model to simulate RO-APM effects on reimbursement and practice financial stability. In this study, we specifically focus on suburban and rural settings.Monte Carlo simulations were developed in Python to model the effects of the RO-APM model over the 5-year trial period. The model requires a patient population, distribution of delivery techniques, and fractionation patterns per delivery technique for each of the 16 defined disease sites. This information was gathered from two non-associated regional clinics: one that serves primarily a suburban population (SP) while the other serves a mixed SP and rural population (MP). All fixed and variable financial expenses were collected and included. The inclusion of extra administrative work for the RO-APM submission was also included using time-driven activity-based costing methods. Considering each clinic's Medicare population, statistical analyses of how the patient population, delivery technique, and fractionation patterns impact reimbursement for this subset population was performed. The RO-APM model was also compared to the fee-for-service model (FFS-M).Disease site distributions were within 7.5% absolute difference between the two clinics. The four most common disease sites treated averaged between the two clinics were lung (20.1%), breast (16.9%), prostate (12.6%), and head & neck (H&N) (9.9%). Fractionation patterns were within 3 fractions (fx) across all disease site modalities except for bladder (MP +19fx) and lung (MP +13fx) IMRT, and 3D H&N (MP +3.4fx). Medicare population was 30% and 33% for the SP and MP, respectively. The SP had a 10.4% ± 0.5% higher reimbursement rate in the RO-APM when compared to the FFS-M for its set of patients seen. The MP had a 5.0% ± 0.3% higher reimbursement rate in the RO-APM when compared to the FFS-M for its patient population. Investigating fractionation impact further, the correlation between the number of fractions across all sites/modalities and reimbursement rate increase for the RO-APM compared to the FFS-M was r (498) = -.45, P < 0.01, meaning that a lower number of fractions leads to a higher return in the RO-APM across multiple disease sites. Finally, additional administrative work in the RO-APM is expected to cost approximately 0.8% of the total reimbursement rate.In our analysis using extracted clinical patterns from two separate clinics, lower fractionation patterns can lead to a higher return on investment in the RO-APM compared to the FFS-M, demonstrating viability of the proposed RO-APM.