In a recent perspective published in JAMA, Ravi B. Parikh and colleagues revisit the Oncology Care Model (OCM) and question the widely accepted conclusion that it failed. Their analysis suggests a more nuanced reality one shaped as much by timing and interpretation as by the data itself.
A Headline That Missed the Trend
The final evaluation of the OCM produced a seemingly clear result. While the model generated savings at the episode level, it ultimately cost Medicare more than it saved, with a net loss of $639 million over six years once program payments were included.
This figure quickly became the defining narrative. Yet it obscures a critical detail: savings improved steadily over time, reaching more than $1,200 per episode in the final performance period.
Rather than a static outcome, the data suggest a model that was evolving and potentially becoming more effective—just as it was being judged.
Where the Savings Actually Came From
The mechanisms behind these savings tell an important story. The OCM was built around care coordination, patient navigation, and improved access, with the expectation that these changes would reduce hospitalizations and emergency visits. In practice, those reductions did not materialize in a meaningful way.
Instead, most savings came from lower spending on non-chemotherapy drugs, particularly supportive care medications.
This pattern reflects a deeper structural issue. Oncologists have limited control over the cost of modern systemic therapies, which are increasingly driven by biomarker-defined, high-cost regimens. In diseases like non–small cell lung cancer, treatment options have shifted toward expensive immunotherapy combinations, leaving little room for cost reduction through clinical decision-making alone.
The Problem With Averaging Everyone Together
One of the most consequential limitations of the evaluation lies in how the data were analyzed. All participating practices were treated as a single group, despite major differences in behavior and engagement.
Some practices invested in care redesign and accepted financial risk. Others participated under more favorable conditions and exited once accountability increased.
When combined, these groups dilute each other’s impact. In fact, a relatively small subset of practices accounted for most of the observed savings.
This raises a fundamental question about interpretation: was the model ineffective, or were its results obscured by uneven participation?
The Value That Wasn’t Measured
The OCM’s impact likely extended beyond what the evaluation captured. Many of the care improvements introduced under the model—clinical pathways, triage systems, integration of supportive care were applied across entire practices.
As a result, patients outside the Medicare fee-for-service population may also have benefited. These spillover effects, however, are not reflected in the official accounting, which focuses narrowly on Medicare spending.
A Timing Problem With Real Consequences
Perhaps the most important issue is timing. The OCM ended in 2022, its successor launched in 2023, and the final evaluation was only published in 2026.
This delay meant that key policy decisions were made using incomplete evidence. The most relevant finding—that savings improved significantly in later years—was not available when the next model was designed.
The result is a successor, the Enhancing Oncology Model, that reflects earlier concerns more than the full trajectory of the data. It includes lower payments, stricter requirements, and mandatory financial risk from the outset—features that have likely contributed to limited participation.
An Incomplete Experiment, Not a Failure
Parikh and colleagues ultimately frame the OCM as an unfinished experiment rather than a failed one. It demonstrated that savings are possible, but also highlighted the limits of physician-driven cost control in a landscape dominated by expensive therapies.
More importantly, it showed that performance can improve over time—if models are given the opportunity to mature.
The lesson is not that value-based oncology care does not work. It is that how we evaluate, design, and time these models matters just as much as the results themselves.
The Question Going Forward
The OCM leaves behind an open and unresolved question. Not whether value-based care can succeed in oncology, but whether the system is willing to support it long enough and thoughtfully enough for that success to emerge.
Written By Aren Karapetyan, MD