Getting Cost Discussions Right: Nudging Patients to Avoid Cognitive Pitfalls

High out-of-pocket medication costs are common across a variety of medical conditions in the United States and can be a barrier to uptake and use of effective treatment options. Recently, high out-of-pocket costs have become particularly salient in the treatment of heart failure.¹ Concerns about out-of-pocket cost can influence patients’ behavior and have been implicated in avoidance of medical care, nonadherence to medications, and the acceleration of healthcare disparities.² On the other hand, patients who receive expensive treatments can face financial toxicity through negative impacts on quality of life or access to other forms of care.³ These effects may be particularly pronounced in patients with low income.⁴ Though there have been efforts to rein in out-of-pocket medication costs, they remain a part of clinical medicine for the foreseeable future.

One approach to address this reality and improve decisions is to integrate cost information into medical decisions and discussions, but patient-clinician cost discussions appear rare and often ineffective. A major reason is that clinicians and patients typically lack accurate cost information. Though logistically complex, this barrier is surmountable, and efforts to improve price transparency are underway.⁵ A more substantive barrier may be the lack of evidence-based approaches to talking with patients about costs and integrating this information into clinical discussions, recommendations, and decisions. Cost-sensitive shared decision-making requires weighing the marginal risk-benefit profile of a particular medication against its out-of-pocket cost and integrating patients’ values on both sides of this equation. Providing cost information is thus necessary but not sufficient; this information may not result in patients’ making choices that align with their values and financial constraints. Existing decision-making literature suggests that multiple cognitive biases may be operative that may skew decisions in different directions.

For instance, sacubitril/valsartan can cost 10 times more than generically-available angiotensin converting enzyme (ACE) inhibitors even for insured patients.⁶ However, sacubitril/valsartan has demonstrated a 3% absolute risk reduction in 2-year mortality compared to ACE-inhibitors for patients with heart failure.⁷ For this reason, it is recommended by guidelines. Unfortunately, existing data suggest that patients may trivialize this benefit, potentially due to unrealistic expectations about benefits of medications and a tendency to discount relatively small differences in absolute risks. A countervailing concern is that patients sometimes conflate cost with value; some patients may overestimate the benefits of sacubitril-valsartan precisely because of its cost. These are two examples of different cognitive biases that may simultaneously undermine effective decision-making within a single encounter and decision, even in the context of price transparency and clear presentation of benefits. These biases are likely exacerbated in the presence of low health numeracy, a challenge that unfortunately correlates with financial need⁸ and complicates the role of shared decision-making in addressing inequity and health disparities.

Cognitive biases in decision-making are well known. We call attention here to their interactions and to their potential to skew decision-making in different directions, as well as to the inadequacy of simple price transparency as a solution. While there has been a common insistence that tools such as decision aids present information in as neutral a fashion as possible, we argue that these problems suggest an important role for using nudge strategies as focused counterweights in order to help patients address out-of-pocket costs and other complex medical decisions. Contrary to the way they are typically described, nudges in these contexts represent an important corrective tool that has potential to promote balance in shared decision-making interactions.

Nudges to counterbalance cost-related biases in shared decision-making

Insights from behavioral economics reveal that nudges involving manipulations in framing and choice architecture can be harnessed to impact decisions.⁹ As defined by Sunstein and Thaler, nudges are manipulations of choice architecture which alter people’s behavior in a predictable way without restricting their options or significantly changing their economic incentives.¹⁰ One example of a nudge would be to default individuals into being organ donors and require people to opt out of if they prefer not to donate their organs after they die. Mandated organ donation, however, would not be a nudge. Clinicians may employ a wide variety of nudges in the context of complex clinical discussions with patients.

Some nudge strategies may be effective, targeted tools for correcting specific cognitive biases in the context of shared decision-making discussions related to out-of-pocket cost (Table 1). For instance, to counter the tendency of patients to discount the value proposition of sacubitril/valsartan due to its modest absolute benefit in the context of expense, clinicians may present the drug as the default choice as compared to ACE-inhibitors. For instance, patients could be told, “Though lisinopril is cheaper, it is less effective than sacubitril/valsartan, which is recommended as the first choice for heart failure.” This approach would use the power of default framing and the tendency toward loss aversion to counter cognitive bias related to dismissal of low-probability absolute risks. To address cost-value conflation, a nudge in the opposite direction could be employed that explicitly describes the absence of relationship between medications’ efficacy and cost. “You should not use the price to indicate how much better one drug is compared to the other.” Importantly, these nudges are entirely compatible with each other and could be implemented simultaneously within the same encounter to counter both types of threats to effective decision-making.

Table 1:

Different types of cognitive biases and potential counteracting, corrective nudges

Decision-making heuristics	Clinical situation	Counterbalancing nudge	Example use of counterbalancing nudge
Cost-benefit conflation	The patient believes a medication is much more effective efficacious because it is more expensive.	Emphasizing the lack of direct connection between price and effectiveness	“You should not use the price to indicate how much better one drug is compared to the other”
Anchoring Bias	The patient’s current medications are only $5, so a $25 medication seems overpriced even if it is associated with a significant benefit	Norming by stating that the cost is consistent with other medications that are very effective	“Many effective medications for heart failure and other conditions have a similar cost”
Loss Aversion	The patient would be willing to pay more for a medication that decreases deaths rather than one described to improve survival.	Altering the framing between mortality and survival to counteract this bias	“Patients who take lisinopril are less likely to be alive in several years compared to sacubitril/valsartan”
Status Quo Bias	The patient is unwilling to switch to more effective medication because there have been no issues with the medication.	Emphasizing that prior experience may not predict future outcomes	“How you have been doing up to this point is not the best way to decide what will work for you moving forward.”

Neither of these strategies is novel on its own. However, the focus on specific components of decisions and the use of nudges in tandem in order to facilitate more balanced decision-making differ from typical invocations of nudging. In addition, two features of nudges bolster the extent to which they are compatible with advancing shared decision-making. Most importantly, nudges are liberty-preserving. They harness decisional heuristics to impact decision-making but do not coerce individuals or undermine decision-making freedom. Second, nudges can be targeted. The nudges we suggest as examples in the context of cost discussion are designed to provide a counterweight against common biases or misapplied heuristics that interfere with effective shared decision-making. Importantly, these are just examples. We are not advocating any particular form of nudge but rather the concept that using multiple, contextualized nudges to optimize specific components of shared decision-making may be helpful to improve complex decisions such as those involving out-of-pocket cost.

There are important barriers or challenges to address. First, nudges are controversial in part because they are paternalistic. The appropriate use of nudges is complex, but this proposed use is not intended to hijack decision-making or to deliver a specific choice. Rather, the goal is to acknowledge ways in which people actually use (and more importantly misuse) information in order to help them achieve a decisional outcome that aligns with their values. Balance and neutrality are emphasized in shared decision-making; however, insistence on neutrality regarding specific components of a decision while patients systematically misapply heuristics or biases does not align with goals of shared decision-making. Second, it can be difficult to determine what types of heuristic-based decision-making count as problematic. In-depth discussions with patients are critical to identify the nature of decision-making, and studies illustrating the frequency and magnitude of errors or biases can help to clarify when nudges may be warranted. Third, rigorous evidence is needed regarding the impact of nudge strategies themselves- both individually and in combination- to ensure that they work and that they do not result in overcorrection of cognitive errors. Fourth, it is critical to identify the goal carefully and to clarify what precise outcome a nudge is intended to achieve. In our example, the goal is countering bias rather than effecting a specific decision. This distinction is consistent with an emphasis in decision science scholarship on measures that reflect adequacy of shared decision-making rather than distribution of decisions within a population.

Nudges used as a counterweight against biases in specific components of a decision may be valuable for enhancing shared decision-making related to out-of-pocket cost and other complex decisions. These decisions have multiple components, and numerous biases may be operative simultaneously. Even when biases in different components of the decision point in different directions, counterbalancing nudges may be combined in order to advance shared decision-making. Testing nudges in this context, in conjunction with a clear definition of the goals of the process, will reveal whether these strategies help patients to avoid cognitive pitfalls and unintended consequences related to price transparency and other complex clinical situations where multiple biases are operative and threaten effective decision-making.