Overview
The Oncology Care Model (OCM) was a Medicare payment and care-delivery program designed for patients with cancer. Instead of paying only for each individual service, the model added monthly care-management payments to participating oncology practices and tied financial accountability to the total cost and quality of care during a cancer episode. A cancer episode was triggered when a patient began systemic therapy, such as chemotherapy, immunotherapy, targeted therapy, or hormone therapy, for cancer.
A key policy concern about episode-based payment models is whether they may unintentionally encourage more episodes to be started. In oncology, that could mean more patients being placed on systemic therapy, even when the clinical benefit is uncertain. This study examined whether the OCM was associated with an increased likelihood of initiating systemic therapy for cancer.
Why this question matters
Systemic therapy is a major component of modern cancer treatment, but it is also costly and can cause significant side effects. In some cases, treatment can improve survival, relieve symptoms, or help control disease. In others, especially for advanced cancer with poor prognosis, the benefit may be limited and the risks may outweigh the gains.
If a payment model triggers reimbursement when treatment starts, one might worry that practices could have an incentive to begin therapy more often. That would increase episode volume and potentially raise overall spending. On the other hand, a care model that supports better planning, symptom management, and shared decision-making might reduce unnecessary treatment.
Study design
This was a quasi-experimental study using a matched difference-in-differences approach. The investigators compared changes over time in practices participating in the OCM with changes in similar practices not participating in the model. The data came from Medicare beneficiaries with an index cancer visit between January 2010 and December 2019, and each patient was followed for one year after the index visit.
The analysis focused on two main populations:
1. Patients with newly diagnosed, or incident, cancer.
2. Patients with poor-prognosis cancer, meaning cancers associated with limited life expectancy or advanced disease.
The comparison period looked at outcomes before and after the OCM began in July 2016. The researchers analyzed data from October 2021 to November 2025.
What outcomes were measured
The primary outcome was initiation of systemic therapy in the year after the index cancer visit. In practical terms, the study asked whether patients were more likely to start chemotherapy or another systemic treatment after the OCM was introduced.
A secondary outcome was total Medicare spending in the year after the index visit. This provided a broader view of whether the model affected overall costs, not just treatment initiation.
Main findings
The study included 754,182 patient episodes in the incident-cancer population, representing 750,483 patients, and 517,858 patients in the poor-prognosis cohort. The intervention and comparison groups each included 197 practices.
In the newly diagnosed cancer population, the OCM was not associated with a statistically significant increase in the likelihood of starting systemic therapy. The differential change was -0.9 percentage points, with a 95% confidence interval of -2.2 to 0.3 percentage points, and the result was not statistically significant.
In the poor-prognosis cohort, there was a statistically significant differential decrease in the likelihood of starting systemic therapy, by 1.5 percentage points, with a 95% confidence interval of -2.8 to -0.2 percentage points. In other words, after the OCM began, patients with poorer-prognosis cancers were somewhat less likely to start systemic treatment in participating practices compared with matched controls.
The spending results followed a similar pattern. For the incident cancer population, the reduction in spending after the OCM was not statistically significant: -$898.26 in the year after diagnosis, with a 95% confidence interval of -$1,890.31 to $93.80. In the poor-prognosis cohort, spending decreased significantly by $2,192.15, with a 95% confidence interval of -$3,559.66 to -$833.63.
Interpretation
These findings do not support the concern that the Oncology Care Model increased the initiation of systemic therapy for patients with newly diagnosed cancer. In that major patient group, episode-based payment did not appear to drive more treatment starts.
For patients with poor-prognosis cancers, however, the model was associated with less initiation of systemic therapy and lower Medicare spending. That could reflect more selective use of treatment, better alignment of therapy with patient goals, more palliative care discussions, or avoidance of low-value treatment near the end of life. The study was not designed to determine which of these mechanisms explained the effect.
An important implication is that evaluations of the OCM that focused only on chemotherapy initiation may have underestimated the model’s savings. If fewer patients began systemic therapy, especially in advanced disease, the financial impact would extend beyond direct drug costs to include administration, monitoring, and related care.
Clinical and policy implications
The results suggest that payment reform in oncology does not necessarily lead to overtreatment. A model that combines financial accountability with care-management support may encourage more thoughtful treatment decisions rather than more aggressive treatment use.
For clinicians, the findings reinforce the importance of individualized treatment planning. Decisions about systemic therapy should continue to be guided by cancer type, stage, expected benefit, symptom burden, functional status, and patient preferences. For patients with advanced disease, discussions about goals of care, quality of life, and palliative options remain essential.
For policymakers, the study highlights that episode payment models can reduce spending without clearly increasing treatment volume. This is relevant as Medicare and other payers consider alternative payment arrangements for specialty care, especially in oncology where treatment costs are high and outcomes vary widely.
Limitations
As with any observational study, this research cannot prove causation with the certainty of a randomized trial. Although the investigators used careful matching and difference-in-differences methods, unmeasured factors may still have influenced the results.
The study also reflects Medicare beneficiaries treated in participating practices, so the findings may not apply fully to younger patients, non-Medicare populations, or healthcare systems outside the United States. In addition, the analysis focused on whether systemic therapy was initiated, not on the exact choice of regimen, treatment intensity, toxicity, symptom relief, or survival outcomes.
Conclusion
In this large Medicare-based study, the Oncology Care Model was not associated with an increase in the likelihood of starting systemic therapy among patients with newly diagnosed cancer. Among patients with poor-prognosis cancers, it was associated with less treatment initiation and lower spending. Overall, the findings suggest that the model did not drive more cancer therapy use under one-sided risk and may have promoted more conservative, potentially more value-based treatment decisions in advanced disease.
Citation
Keating NL, Lam MB, Landrum MB, McWilliams JM, Wright AA, Brooks GA, Zubizarreta JR, Buzzee B, Landon BE. The Oncology Care Model and Initiation of Systemic Therapy for Cancer. JAMA Internal Medicine. 2026;186(6):732-741. PMID: 42043828.