“Practices participating in the OCM are expected to vary in performance against their benchmarks,” said Richard Kane, associate principal at Avalere. “But the variation we’re seeing between actual and predicted costs across all patients being treated by OCM practices suggests the program’s benchmarks and performance-based payments are too low.”

The OCM is a voluntary 5-year episode-based payment program developed by the Center for Medicare & Medicaid Innovation that started in 2016. The OCM encourages practices to improve care coordination and lower costs through episode-based cost performance and quality measures. Practices achieve a performance-based payment if their per-episode costs are below their spending target during a 6-month performance period.

As part of its analysis, Avalere compared OCM participants and non-participants across a variety of metrics. Avalere estimates that OCM participants’ actual Medicare costs per episode were 2–4% higher than for non-participants. The prediction model estimates that participants’ per-episode costs would be only 0.5–1% higher than for non-participants. The remaining 2–3% difference could be due to key risk factors missing from the prediction model.

“If the OCM’s prediction model underestimates actual costs, then success for practice participants becomes more difficult,” said Lance Grady, managing director at Avalere. “We expect CMS to continue to improve the model, as it recently said it would incorporate stage-of-tumor information for breast, lung, and prostate cancers into the prediction model.”

CMS created the OCM prediction model to risk-adjust Medicare costs for performance measurement. The prediction model estimates the dollar effect of numerous risk factors (e.g. type of cancer, patient comorbidities) on each episode’s total Medicare costs, based on a regression analysis of non-participating practices’ Medicare costs.

OCM Practices’ spending targets are shaped by their benchmarks set by the CMS to measure each OCM practice’s Medicare cost growth against oncology practices not participating in OCM. Each participating practice’s benchmark reflects the predicted cost growth calculated by the prediction model for its episodes.

Funding for this research was provided by Amgen. Avalere retained full editorial control.

Methodology

Avalere performed this analysis using Medicare Part A/B fee-for-service (FFS) claims and Part D prescription drug event data under a CMS research data use agreement. A cohort of patients was selected that included all OCM-eligible Medicare FFS cancer patients receiving cancer treatment and represented less than 20% of total Medicare beneficiaries. Performance periods 1 and 2 include 6-month episodes beginning between July 1, 2016 and June 30, 2017.

Avalere replicated the OCM payment method developed by CMS, including the attribution of episodes to practices, the calculation of their benchmarks (including the novel therapy adjustment), and the estimation of their performance-based payments. Avalere’s analysis uses a version of the OCM prediction model first implemented in performance period 3; it includes changes made by CMS in the way breast, prostate, and bladder cancers are separated into high-risk and low-risk categories.

To receive Avalere updates, connect with us.