CORR Insights®: Small Social Incentives Did Not Improve the Survey Response Rate of Patients Who Underwent Orthopaedic Surgery: A Randomized Trial

Where Are We Now?

Patient-reported outcome measures (PROMs) have increasingly been recognized as an important indicator of patients’ health and recovery, particularly after a major event such as surgery or initiation of chemotherapy [1]. A central challenge to better utilizing these types of data is obtaining information reliably. Patients may have varying engagement levels with this type of outreach and so potential interventions could try to address increasing motivation for providing PROMs.

In the current study, Warwick and colleagues [11] conducted a randomized trial to evaluate whether small social incentives could improve collection of PROMs from patients who underwent orthopaedic surgery. Three thousand patients were randomly assigned to either a control group, a group that received an explanation as to why PROMs were requested, or one of two financial incentive groups with USD 5 allocated to either medical supplies for pediatric orthopaedic patients (patient donation) or a procedure-specific research program (research donation). They found no differences in response rates between the control and any of the intervention arms.

There are several possible explanations for why the financial incentives allocated to donations were not successful. First, the magnitude of the incentive may have been too small to motivate patients to respond. This has previously been referred to as the “peanuts effect” in which individuals tend to be more risk-taking for lower-valued incentives [10]. A larger incentive might have led to a higher response rate. Second, the incentive was “gain-framed” in that patients could receive the incentive after responding to the survey. Many individuals tend to be more motivated to avoid losses than to gain the same-sized reward [5]. Prior work has demonstrated that simply framing incentives as being allocated up front with the possibility of being taken away is more effective than framing them as a potential gain [2, 8]. Third, these patients had surgery 1 to 2 years ago. Response rates might be higher immediately after surgery because patients might feel that their input could lead to changes in their own care, as opposed to the care of others. Prior work conducted with patients at a university hospital found that individuals were more likely to donate to charity when asked closer to their interaction with the healthcare system and that reciprocity decayed rapidly over time [3]. Fourth, the donation to charity might have been more effective if patients were allowed to select the charity rather than be told it was for another patient or research. For example, in a prior clinical trial [4], patients increased physical activity when they were offered an incentive and able to donate it to a charity of their choice.

Where Do We Need to Go?

PROMs provide valuable insights into a patient’s progress and health. However, the best way to embed them within current healthcare delivery settings to engage patients has not been well-explored and further research is needed to identify addressable barriers. A key question is when to use incentives and how to best design them to increase response rates. This includes finding ways to reduce the effort needed from patients to submit PROMs and testing different approaches to motivate patients who may otherwise not respond.

While the incentives used by Warwick and colleagues [11] were not successful, this should not be taken as an indication that incentives do not work in this context. Instead, it is important to understand how even subtle elements of design can have an impact on patient behavior. Further research is needed to help improve patient engagement with PROMs.

How Do We Get There?

Several things could be done to help us better understand the best ways to engage patients in completing PROM questionnaires. We might first engage patients who are in the hospital for a procedure about their preferences and motivations for providing PROMs after surgery. This could be done through survey assessments that all patients must complete before discharge but also through one-on-one interviews to allow for more enriching conversations around specific barriers. This will help provide valuable insights to improve the design of interventions. Second, we need to test different approaches to address the identified barriers within pragmatic settings such as within the first 90 days after surgery. Importantly, these approaches should be tested in well-designed randomized trials to compare effectiveness and evaluate unintended consequences Third, insights from behavioral economics have demonstrated how individuals tend to be predictably irrational and that could be used to frame information and design incentives to address barriers to behavior change [3, 6, 7]. For example, we tend to be more motivated by immediate rather than delayed rewards, take greater action to avoid loses than obtain an equivalent gain, and our decisions are heavily impacted by our emotions and social influences. These insights could lead to better designed financial incentives, such as rewards that are allocated upfront at the time of discharge and taken away if PROMs are not submitted on a regular basis. It could also help to frame information in a way that is more motivating, such as the use of peer comparisons which informs patients how their engagement compares to other similar patients [9].