Comparing Prescribing Habits in Academic and Nonacademic Oncology Settings

A Conversation With Aaron P. Mitchell, MD

Get Permission

The art of oncology practice is tailored to the individual patient with cancer, and with the advent of highly personalized targeted therapies, patient outcomes have improved markedly over the past several decades. Although much of oncology practice is guideline- or protocol-driven, chemotherapy prescribing habits vary between settings. A recent study looked at this issue by examining and comparing prescribing habits in academic and nonacademic settings.1 The ASCO Post spoke with the study’s lead author, Aaron P. Mitchell, MD, Assistant Attending in the Department of Epidemiology & Biostatistics and the Genitourinary Oncology Service, Memorial Sloan Kettering Cancer Center, New York. His research interests are physician practice patterns, payment policy, and conflicts of interest in oncology.

Aaron P. Mitchell, MD

Aaron P. Mitchell, MD

Financial Incentives in Choice of Treatment

You were the lead author on a study looking at the difference in prescribing habits of cancer therapies between academic institutions and nonacademic institutions. Please tell us what initiated your interest in this work and a bit about the study’s design.

I became interested in studying oncologists’ use of different cancer drugs after I learned about how cancer drugs are reimbursed. Intravenously administered drugs, which include a large number of chemotherapies and biologic agents, are reimbursed on the “buy and bill” model, wherein the provider purchases the drug from the manufacturer and essentially sells it to the patient (or the insurer) at a markup, which runs in the 4% to 6% range for Medicare and significantly more for commercial insurers. The result of this system is that, unlike when doctors write prescriptions for patients to fill at the pharmacy, we oncologists have a financial stake in whether our patients undergo treatment, and in which drugs they are treated with. I find this problematic because I believe that the only considerations that should go into a treatment decision are the medical evidence for a treatment and the patients’ preferences—not how much money the physician might make.

Therefore, in this study, we wanted to explore how these incentives might affect physicians’ choice of cancer drugs. We decided to look specifically at the use of several biologic drugs that were very expensive and yielded a significant financial return to the provider but provided little clinical benefit to patients. We used the National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guidelines®) to identify these drugs. We found cancer types for which the NCCN Guidelines recommended a range of options: a regimen of cytotoxic chemotherapy drugs (which are relatively inexpensive) or the same cytotoxic drugs with the addition of a high-cost biologic agent.

We then measured the chemotherapy treatments real-world patients were receiving in our data set, which included patients with cancer in North Carolina. We decided to compare patients treated within the state’s several large, academic, National Cancer Institute (NCI)-Designated Comprehensive Cancer Centers (NCI Centers) with those treated in non-NCI settings, which largely included private community practices. We made this choice because we hypothesized that the financial incentives in chemotherapy billing might be felt differently in an academic setting than in a community practice. In particular, community oncology has been under increasing financial pressure in recent years, and many community practices may be grappling with financial sustainability. Consequently, oncologists operating in smaller practices might have to pay more attention to the financial return on different drugs, leading to the potential for an increased preference for higher-priced drugs.

Key Study Findings

Please describe the findings of this study and their possible implications across the broader cancer care spectrum.

After controlling for differences in how sick patients were at baseline, we found that the patients in our sample were more likely to receive the high-cost biologic drug than a less-expensive, NCCN-recommended alternative regimen, when treated outside of NCI Centers vs inside NCI Centers. Of note, this difference was not apparent when looking at patients with colon cancer, who received about the same treatment everywhere. This difference was primarily seen in patients with lung cancer and head/neck cancers, who were more likely to receive biologic agents when treated in non-NCI–designated settings.

We must be cautious in generalizing these results, since we studied only three cancer types, and each cancer type is its own “world” in which treatment decisions occur based on a different set of factors. But, in principle, I believe we’ve found differences in the preference for high-cost biologic drugs and would not be surprised if that carried over to other cancer types as well.

Real-World Effectiveness

Do we know if it’s feasible to generalize practice prescribing habits, perhaps with guidelines?

I think we are awaiting ongoing research looking at the influence of guidelines on practice. It is important to note that all of the regimens we studied here were recommended in NCCN Guidelines, so all of these patients received guideline-concordant care.

“Oncologists operating in smaller practices might have to pay more attention to the financial return on different drugs, leading to the potential for an increased preference for higher-priced drugs.”
— Aaron P. Mitchell, MD

Tweet this quote

However, that begs the question of what goes into our guidelines. Not every drug that receives U.S. Food and Drug Administration (FDA) approval offers meaningful (or, in some cases, any) clinical benefit to patients. Our guidelines and pathways need more freedom to recommend against drugs that offer no benefit or offer marginal benefit at an unreasonable financial cost. For example, in 2013, the Dana-Farber Cancer Institute implemented a pathway for non–small cell lung cancer which recommended against the use of pemetrexed-plus-bevacizumab combination regimens in the first-line setting. Patients did just as well after routine use of these drugs was stopped. I think we need guidelines that better reflect the serious questions raised about the real-world effectiveness of many recently approved cancer drugs.

More Than ‘Hanging a Bag of Chemo’

Although not within the scope of this particular study, we know that community practices have been historically under-reimbursed for multiple cognitive services. Do we know what part our complex billing process plays in this issue?

It is difficult to know how much of the association we saw in this study was due to responsiveness to billing incentives and how much may have been due to nonfinancial factors, such as differences in beliefs about the efficacy of these drugs. But yes, I would strongly be in favor of a reimbursement system that placed much more emphasis on the true value we provide to our patients, which is the cognitive work of understanding our patients’ medical concerns and values, educating them on their disease, and deciding on treatments that best meet their needs and preferences. The value we provide is not hanging a bag of chemo—but that is what the “buy and bill” model says our value is. Oncologists should be empowered to recommend whatever is best for each patient, without having to worry about the financial impact of which drugs we do or do not bill for.

Closing Thoughts

Please share some final thoughts on this important issue as it pertains to the delivery of oncology services nationwide.

Our study suggests differences between academic and nonacademic practices with respect to specific high-cost biologic drugs. To the extent that this reflects differing views about these drugs’ clinical effectiveness, we need a broader conversation that may result in the de-implementation of some low-value treatments—including some which may have received FDA approval. To the extent that this reflects oncologists’ responsiveness to financial incentives, we need a reimbursement model that eliminates drug margins and prioritizes other higher-value aspects of cancer care delivery. 

DISCLOSURE: Dr. Mitchell was the recipient of the 2019 ASCO/Conquer Cancer Foundation Abstract Merit Award, which was funded in part by Merck.


1. Mitchell AP, Kinlaw AC, Peacock-Hinton S, et al: Use of high-cost cancer treatments in academic and nonacademic practice. Oncologist 25:46-54, 2020.