The Oncology Care Model (OCM) is a voluntary, episode-based oncology-focused payment model from the Centers for Medicare and Medicaid Services (CMS) that aims to improve care coordination and reduce costs for Medicare Fee-for-Service (FFS) beneficiaries. The model evaluates the total cost of care for 6-month episodes initiated by chemotherapy treatments that are attributed to oncology practices participating in the model. Starting in July 2016, in each performance period (PP), the CMS has measured practice performance based on total episode expenditures relative to a benchmark price. The OCM prediction model is used to calculate the episode benchmark price based on historical spending that is projected forward using a common trend factor. Practice performance and practice elected risk track determine whether the practice earns a performance-based payment or owes a recoupment for the period.

Breast, prostate, and lung cancers are among the most common cancer types in the OCM model, accounting for over half of OCM episodes. Avalere replicated the OCM program methodology and prediction model and calculated episode expenditures and benchmark prices. To compare performance across cancer types, Avalere calculated the ratio of average episode expenditures to average benchmark prices in PP3–PP8 among high/low-risk breast, high/low-intensity prostate, and lung cancer episodes, separately. Ratio values can be interpreted as:

• Less than 1.0: Episode expenditures were less than the benchmark price
• Equal to 1.0: Episode expenditures were equal to the benchmark price
• Greater than 1.0: Episode expenditures were greater than the benchmark price

As shown in Figure 1, the average episode expenditures for high- and low- intensity prostate cancers, and low-risk breast cancer were consistently below the benchmark price (ratio <1.0) and improved relative to the benchmark over time. In contrast, average episode expenditures for high-risk breast cancer and lung cancer were consistently above the benchmark price across the performance periods.

Avalere also examined episode cost components, based on service type, to quantify differences in cost drivers by cancer type. We analyzed differences in cost components in low- and high-risk breast cancer, which were cancer types with expenditures consistently below and above their benchmark prices, respectively (Figures 2–3).

Inpatient payments, as a percentage of total expenditures were relatively stable among low- and high-risk breast cancer episodes in PP3–PP7. The COVID-19 PHE overlapped with episodes triggered in the last half of PP7, and PP8 was the first full PP to overlap entirely with the PHE. In PP8, inpatient payments decreased in both low- and high-risk breast cancer, likely due to avoidance of the hospital setting during this time.

Among low-risk breast cancer episodes, inpatient payment, as a percentage of episode expenditures, decreased 3.3 percentage points from PP7 to PP8. During the same period, the share of hospital outpatient department and carrier (includes physician) payments increased by 2 percentage points and 3.3 percentage points, respectively.

While the share of inpatient payments also decreased in high-risk breast cancer, it was a smaller relative decrease of 1.6 percentage points compared to 3.3 percentage points for low-risk breast cancer. There was not a corresponding increase in carrier and outpatient payments, as seen in low-risk breast cancer.

Implications

During the PHE, healthcare utilization patterns changed as some patients avoided the hospital due to stay-at-home orders or were turned away due to capacity constraints.¹ A shift in the setting of care away from hospital inpatient was observed in low-risk breast cancer OCM episodes but not in high-risk breast cancer, suggesting a differential impact of the PHE by cancer type. Patients with low-risk breast cancer typically have different clinical acuity and treatments, and may have been able to more successfully avoid hospital admissions than high-risk breast cancer patients.

Additionally, OCM performance for low-risk breast cancer episodes, along with low- and high-intensity prostate cancer, while consistently under the benchmark in previous periods, improved even further during periods overlapping with the PHE (i.e., PP7 and PP8). It remains to be seen whether OCM performance for these cancer types continued to improve once additional periods overlapping with the PHE are analyzed. With the OCM ending this month (i.e., June 2022) and uncertainty around future CMS oncology payment models, stakeholders will need to consider how to distinguish broader effects of innovation and practice transformation on OCM practice performance from temporal impacts of the PHE. Understanding these dynamics and how they may vary by cancer type will be of critical importance in preparing for ongoing participation in Oncology Alternative Payment Models (APMs)

Methodology

Avalere performed this analysis using Medicare Part A/B FFS claims and Part D prescription drug event data under a research-focused data use agreement. Avalere replicated the OCM program methodology developed by the CMS, including the attribution of episodes to practices, the assignment of cancer type, and the calculation of benchmark prices. Benchmark prices were calculated based on the CMS prediction model, which estimates episode costs based on key factors such as tumor type, beneficiary demographics and comorbidities, and utilization of certain services. Participants were identified in the claims data through tax identification numbers (TIN) reported on Part B claims. Avalere identified 124 practices that were ever enrolled in the OCM through TIN mapping and general research of oncology practices. This analysis includes practices active in the OCM as of December 2021. The baseline period includes episodes initiated between January 1, 2012, and December 31, 2014; and the intervention period used for this analysis includes episodes initiated between July 2, 2017, and July 1, 2020 (PP3–PP8). Note that the OCM reconciliation methodology excludes episodes with a diagnosis for COVID-19 in PP7 and PP8.

Note

1. Smulowitz, et al. “National Trends in ED Visits, Hospital Admissions, and Mortality for Medicare Patients During the COVID-19 Pandemic,” Health Affairs 40.9.