Why Aren't We Choosing Wisely?

Audrey S. Wallace, MD; Gabrielle B. Rocque, MD, MSPH


J Oncol Pract. 2020;16(8):443-445. 

The increasing emphasis on value in health care is shifting treatment paradigms from those that focus almost exclusively on improving survival outcomes to those that consider opportunities to improve value by finding a balance between appropriate treatment and minimizing unwanted toxicities at the level of the patient, provider, and health care system. The most widely recognized effort to reduce low-value care is the Choosing Wisely guidelines, which engage specialty societies to identify and discourage practices that are either not evidence-based or have the potential to cause unwanted outcomes. In 2013, the American Society for Radiation Oncology (ASTRO) published recommendations for avoidance of prolonged fractionation for patients receiving palliative-intent radiation for bony metastasis who had limited prognosis or transportation difficulties.[1] However, almost 9 years later, we have had several publications that report low adoption of shorter fractionation regimens despite Level I evidence from multiple randomized trials.[2,3] These studies draw attention to the fact that stopping well-established practices is challenging and unlikely to occur without thoughtful consideration of how decisions are truly made.

Efforts to reduce ineffective cancer treatments, practices, and interventions fall under the umbrella of de-implementation of care. The de-implementation framework by Norton et al[4] from the National Cancer Institute identifies that barriers and facilitators exist at the level of the setting, patient, provider, and society. In the article that accompanies this editorial, Gupta et al[5] should be applauded for their assessment of physician factors contributing to this trend, finding that greater number of years since medical school training, practice location in the South, and high-volume practices (treating more than 20 patients during the study period) were associated with higher use of extended-fraction radiation. The observation that radiation oncologists farther from completion of their training had lower concordance rates is important. This may be secondary to a lack of awareness of the ASTRO and Choosing Wisely guidelines. Since publication of these recommendations, we have increasing data showing that a single-fraction regimen is an option, even in the setting of spinal cord compression. The question becomes whether radiation oncologists are aware of the ICORG 05–03 (NCT00968643) study that assessed a dose of 10 Gy in 1 fraction versus 20 Gy in 5 fractions (presented as a plenary session at the ASTRO 2014 Annual Meeting), or the SCORAD III study that assessed a dose of 8 Gy in 1 fraction versus 20 Gy in 5 fractions (presented at the ASCO 2017 Annual Meeting as an oral abstract).[6,7] The ability to learn new information and integrate it into practice may be challenging for providers who treat diverse cancers in the setting of constant updates to patient management.

Although these findings are important, we believe that additional nuanced intangible factors associated with treatment decisions are critically important and are largely unavailable in analyses of large databases. As highlighted in the de-implementation framework, one must consider that situational or patient factors may lead to hesitation to treat with shorter fractionations. Are radiation oncologists worried about previous radiation therapy? Is there concern for cord compression? Is there large-volume metastatic disease, raising concern for treating a large field with a single fraction? The data required to answer these questions are not available in analyses of large databases; thus, there are patients for whom longer fractionation is likely appropriate. So conclusions should be interpreted with the understanding that complete guideline concordance is not a realistic goal in real-world settings.

We must also be cognizant of the fact that patients and physicians often differ in their perception of what is important in their choice of treatment. Are providers worried about the re-treatment rate with a single fraction? Could it be that a radiation oncologist is simply more comfortable re-treating with 1 fraction rather than with 10 fractions? In our previous qualitative work in metastatic breast cancer, we learned that patients and medical oncologists differ in their perspectives of decision making in medical therapy. Patients considered 10 different domains, including treatment logistics and convenience, personal and family responsibilities, and impact on daily activities. Oncologists focused primarily on treatment efficacy, physical adverse effects, and cost.[8] This type of granular qualitative work is needed to better understand physician behavior as it relates to use of extended-fraction radiation and other practices deemed to be of low value.

Choosing Wisely guidelines that focus on the end-of-life setting bring a host of other complex considerations, such as treatment recommendations in the context of predicted longevity. Patients' perceptions of prognosis and goals of treatment are often incorrect.[9] Physicians have historically struggled with estimating prognosis at the end of life.[10,11] It is exciting to practice in an era in which patients with a prognosis of only a few months now may be among the select few who have a prolonged response in the era of immunotherapy. Tools such as the Dutch or Rades scales are becoming available for prognostication not just with systemic therapy, but also in the setting of patients who have bone metastasis,[12,13] but these tools are not commonly used. Finally, there are concerns that patients want physicians to maintain hope, and that patients will view their oncologists in a negative light if they provide bad news about prognosis. These fears coupled with societal reluctance to discuss end-of-life care can be a barrier to open conversations about the best treatment options for patients. Additional work is needed to support oncologists in communicating the value of specific treatment options in the setting of limited prognosis. This can provide benefit not just at the system level, but at the patient level by minimizing the time and cost burden associated with treatment at the end of life.

Although caregivers are often underappreciated as key stakeholders, they play a large role in end-of-life care and decision-making. The informal caregiver, often a spouse or a child, may need to take time off from work to take their loved one to various oncologic appointments. The time, travel, and financial burden are the elephant in the consultation room. Most patients treated for bone metastasis are older and have difficulty with repeated travel. In one study of decision making about treatment options for curative breast cancer, women often chose mastectomy over breast conservation therapy and radiation therapy because of the travel burden of daily radiation treatments.[14] Given the impact on caregivers, should we do more to engage them? Imagine this conversation: "I know your mom has decided to let you help make decisions for her, so let's talk about treatment options. I expect her to live about a month. Radiation will involve 1 day for consultation, 1 day for the radiation planning scan, and 10 days of treatment. She will likely spend half of her remaining days traveling back and forth for radiation therapy." Shared decision making is the cornerstone of the recommendation from the Institute of Medicine that emphasizes the importance of conversations such as this one.[15]

In the era of big data, we strongly believe that studies such as that conducted by Gupta et al[5] are an essential starting point for identifying large-scale patterns of care. We also strongly advocate for the use of additional methodologies (eg, surveys or qualitative interviews) to conduct a deeper dive into granular decision-making drivers. This critical review will be essential not only for reducing use of extended fractionation in radiation therapy, but for all types of de-implementation efforts across cancer care. Without a full understanding of the true driving factors that hinder de-escalation of treatment, we are unlikely to be able to truly change treatment patterns.