Fear in the Age of COVID-19.

“Should we still start my cancer treatment with all of this going on?”

When variations of this question first started popping up in our radiation oncology clinic, severe acute respiratory syndrome coronavirus-2 was probably already spreading rapidly across New York City. However, its grasp had not yet tightened. Buses and subways were still running, restaurants were overflowing, and schools were open. Our public health authorities had not yet realized the magnitude of the threat it posed.

In those early days, with the city seemingly thriving, it was easy to provide patients with an answer to this question. Although a local viral outbreak was certainly possible, surely the greater risks to our patients were their cancers. After all, we are taught that the treatment of cancer almost always takes precedent over other diseases, right?

But as the weeks passed and coronavirus disease 2019 (COVID-19) began ravaging New York City, our recommendations have changed. We have increasingly been forced to wrestle with the possibility that we are exposing our patients to situations that elevate their risk of acquiring a potentially fatal viral infection by asking them to repeatedly return to the clinic to receive fractionated radiation therapy. In the space of a mere month, this has become an overriding concern in our evaluation of every patient.

It has been an uncomfortable transition. As radiation oncologists, our treatment recommendations are based on solid algorithms and data. Yet, as COVID-19 has spread across our great city, we have been forced to consider whether to delay or modify care without data to guide us.

Some cases, of course, are easier to manage than others. For elderly patients with less aggressive malignancies, like low-risk prostate cancer or basal cell carcinomas of the skin, deferring definitive treatment while the pandemic crescendos does not seem unreasonable. In these situations, there is a clear benefit to keeping high-risk patients away from medical facilities, which are hot spots for the virus. Additionally, with potentially long windows for definitive treatment, these patients would most likely not suffer any negative consequences from delays in care.

Other cases, however, present more difficult quandaries. How, for example, should one manage an elderly patient with a newly diagnosed stage IIIA non-small cell lung cancer? Or a patient with a locally advanced squamous cell carcinoma of the head and neck in the setting of multiple comorbidities? These cancers cannot wait. For patients whose window for cure is quickly closing urgency in starting treatment is generally required. The risk of poor outcomes is likely only amplified in patients receiving immunocompromising chemotherapy.

As disappointing as treatment failures are, before COVID-19, they could at least be rationalized through the lens of a study. Even when the risk was low, they always occurred in a certain percentage of cases. Thus, the rules of probability dictated that they would inevitably happen again. Statistics, by this logic, had let our patients down. We had not.

Now, however, in the midst of a pandemic and with little data to aid our decisions to delay or modify care, we have been forced to embrace an unsettling new reality. Data have been partially replaced by judgment. And our anxiety has morphed into fear.

In these conditions, in which the principle of nonmaleficence looms large, we have found ourselves relying more than ever on shared decision making to individualize treatment plans. Patient preferences and values have proven keys to respecting patient autonomy. Yet, at the end of the day, patients rely heavily on our medical expertise in formulating these plans. They place great trust in our judgment and experience.

It is impossible to know whether the plans we have jointly created are the right ones. But even if our guidance is ultimately proven correct in the vast majority of cases, it is likely that at least some of our patients will suffer adverse outcomes, whether it be via viral infection for those who begin or disease progression for those who delay. Certainly, we will then question our initial recommendations.

Without the veneer of data to potentially rationalize these outcomes, we are left fearing what used to be routine recommendations. Fearing the prospect that we may be wrong. Fearing the possibility that we may be letting our patients down. And fearing the guilt that is almost certain to arrive.

That, for us, is one of the hardest parts of all of this.