The Black Box of Out-of-Pocket Cost Communication. A Path Toward Illumination

Sexuality, drug use, domestic violence, end-of-life planning—recently, sensitive subjects such as these went from taboo to openly discussed topics between physicians and patients. But what about finances? Are patient out-of-pocket costs also being openly discussed in America today? Or are the financial aspects of health care relegated to the dark and left unexplored in the patient–physician relationship?

Several decades ago, it might have made sense for doctors and patients to avoid discussing costs. People with insurance were shielded from the vast majority of their health care expenses, and those without insurance often received charity care, with the cost of that care being shifted to third-party payers (1). Today, many people with insurance face high out-of-pocket costs as a result of the growth of consumer-driven health plans (2). In addition, health care providers are often unable to shift costs to third-party payers.

The result is an epidemic of financial toxicity among people receiving medical care. In 2012, more than one in four American families experienced financial burden because of their medical care, and one in five families were paying medical bills late (3). This financial burden reduces quality of life, increases medication nonadherence, and leads to poor health outcomes (4, 5). As physicians, we are in the business of improving patient health and well being, so it is time for us to start discussing costs for the interventions we prescribe (6).

Physicians do not need to become patients’ financial counselors. Instead, they need to provide patients with cost-specific information for the interventions being considered. This task could be as simple as informing patients that Advair typically runs $30–45 monthly, that Xopenex is the pricier cousin to Albuterol, or that Pulmicort has a low-cost generic form available. In addition to providing key information about the financial implications of treatment options, these latter statements open up space in the decision-making process for patients to voice personal budget concerns or ask questions about costs. In the fight against financial toxicity and its sequelae, providing this information and actively weighing medical and financial pros and cons of treatments are powerful weapons.

How often are such conversations occurring in American medicine today? The unfortunate truth is that we have a poor understanding of the status of cost communication between physicians and patients. The literature is scant, and what has been reported reveals marked variability. For example, in a survey done by Shrank and colleagues, 42% of medical oncologists reported discussing cost all or most of the time, and another 32% reported sometimes discussing cost with patients (7). In Bestvina and colleagues, however, only 19% of oncology patients reported discussing costs with their physician (8). Very few studies collected paired physician–patient data, but when they did, information from the two sources was often contradictory: physicians said they discussed cost with 35% of patients, and only 15% of patients reported discussing cost with the physicians (9).

In this issue of AnnalsATS, Patel and Wheeler (pp. 1538–1544) provide a point of clarity in this enigmatic literature by studying cost communication in an important, yet understudied, population (10). In their article, they present findings from a cross-sectional survey of cost communication attitudes and behaviors in a vulnerable population of 422 African-American women. With low average income, a high prevalence of self-reported financial burden, and a chronic medical condition, this population was also at particularly high risk for cost-related nonadherence. Patel and Wheeler observed that three-quarters of the women reported a preference to discuss cost with their health care provider, but only 39% reported having a conversation with their physician about cost during a clinical encounter (10). Although this study featured the highest patient-reported cost conversation frequency to date, it presented a similar, large discrepancy between patients’ desire to discuss cost and the reported frequency of cost discussions. This discrepancy should cause us to pause: How should we explain or interpret this?

One point to consider is the power of recall bias. Studies have shown that patients immediately forget approximately half of the information they are given verbally during clinic visits about their diagnosis, prognosis, treatment, or medications (11, 12). In addition to this, nearly half of the information they do recall about these critical subjects is incorrect (11). A physician could provide key information about the copay of a drug or the pricing of its alternative in one or two spoken sentences lasting less than 10 seconds, but if this were the only mention of cost during a clinic visit, it is plausible the patients would fail to recall this cost discussion when later surveyed about the encounter. Even if the patient did remember that one- or two-sentence exchange, would they describe that as “discussing cost” with their provider? Or do they typically only report rich, two-way dialogues? What if that two-sentence exchange was about securing insurance coverage: “I also want to add on Singular, but first I need to make sure that it’s covered by your insurance.” Such dialogues definitely address patient costs, but are they likely to be counted by patients as cost discussions? What effect do these definitional ambiguities have on Patel and Wheeler’s findings? It is very hard to tell.

The omission of strict definitions for “cost conversation” and “out of pocket costs” and the lack of clarity with regard to how discussion of insurance coverage fits into it all is a problem that has plagued nearly every study in this literature, casting doubt on the precise estimates of cost conversation. However, the lack of definition in no way diminishes the importance of the work of many fine investigators in grappling with this slippery subject.

We propose that it is time to rethink how we address the issue of cost communication in medicine. Instead of more survey studies, what we need is careful analysis of real recorded clinic visits. This would free us from recall bias and would allow for clear and consistent application of a more strict definition of cost conversation. We could begin by characterizing the content, stumbling points, and successful tactics already in use today. This would finally illuminate the black box of cost communication in America and guide our thinking on best practices for discussing cost with our patients as they continue to take on greater shares of their medical costs in the years to come.