Patient Safety Incentives for Residents: A Slippery Slope or Reinforcement of Desirable Behavior?

In a study of error reporting in the care of Medicare beneficiaries, the US Department of Health and Human Services Office of Inspector General found that only 14% of patient safety events were reported.¹ A multicenter study found that 44.6% of residents were uncomfortable reporting diagnostic errors, possibly related to fear of repercussions and issues with the reporting system, among other factors.² A key question at the intersection of graduate medical education and patient safety is how to encourage resident and fellow error reporting.

The intervention described in the article, “Improving Resident and Fellow Engagement in Patient Safety through a Graduate Medical Education Incentive Program,”³ in this issue of the Journal of Graduate Medical Education produced results: nearly 4000 patient safety events were reported, a significant increase over the data from the preintervention period. Among those 4000 reports could be 1 or more producing measures that have a significant impact on patient safety, potentially saving lives. The intervention—paying residents for reporting errors—is simple. From a health systems perspective, the costs incurred are a concern and need to be weighed against their formal or informal returns on investment (ROIs). The following are considerations for that analysis.

Increased Reporting as an Outcome

Health professionals' engagement in error reporting is a desired outcome in a safety culture, and is a key part of a patient safety program. The more reports, the better. Knowledge of events, along with subsequent actions such as disclosure, has been shown to reduce liability-related costs for institutions and resolution time for patients.⁴

From a quality improvement standpoint, error reporting is a fairly blunt measure. A given error may have many causes.⁵ For example, the category “Provision of Care,” discussed by Turner et al, included “a wide range of themes,” such as response to patient condition, transitions in care, and deviations from protocol and order completion.³ While it would be difficult to show improvement in a such a broad category, it can highlight areas of needed improvement.

Most importantly, the reported events in this study are from a resident's or fellow's point of view. In most settings, nurses and pharmacists provide the bulk of event reporting, and it follows that reported errors are from the perspective of these disciplines.¹ Trainee reporting adds a key piece to the patient safety puzzle and allows us to understand how this key stakeholder group see their work regarding efforts to improve safety.

Patient Safety Incentives as the Return

Herzer and Provonost⁶ noted 2 basic means of physician motivation: extrinsic motivators (based on reward and punishment) and intrinsic motivators (where physicians are motivated intellectually and professionally to solve complex and challenging health care issues).

Evidence supporting the efficacy of pay for performance (P4P) programs, 1 type of extrinsic motivator, is weak. There are several reasons why P4P may not work. First, there may be a perception that clinician autonomy is reduced—the clinician cannot determine his or her own method to improve outcomes. Second, incentive programs may, as the authors of this study point out,³ simply result in “check the box” activities to achieve the measure, which creates less useful data, and may ignore other worthy improvement initiatives. Finally, clinicians may not know how to address or be unable to effect improvements.^3,6

Despite the lack of evidence supporting P4P, incentive programs are common in health care organizations and are offered in 4 primary ways. First, “Good Catch” programs recognize individual event reporters and often include a small award and recognition. Second, event reporting goals may be linked to individual goals as a part of a periodic evaluation process. Additionally, increased reporting is found in organizational goals, tracked by dashboards, and sometimes linked to incentives. Finally, P4P measures are a part of payer, Medicare, and other agreements. The first 2 types of incentives are individual and developmental, and do not involve significant compensation, although they may include merit raises. Incentives based on organizational event reporting are usually team-based, involving a unit, hospital, or clinical site.

The incentive offered in this study differs from typical organizational incentives in that the incentive is paid to individual residents. Applying individual incentives may shift the motivation from being an intrinsic, team-based, driven motivation where problems are collaboratively solved, to an extrinsic, “check the box” type of motivation.⁷ The worry is that, once the bonus is removed, the average resident reporting of 2 reports per month may stop.

Incentivization of physicians does not happen in a silo. Other health care team members will at some point become aware that residents are being paid to report. Other professional groups may raise the question as to why medical trainees are being paid for safety event reporting, when other team members are not. Other professionals may reduce their reporting, as they may perceive the organization does not value their reporting: if one is not going to be paid, why report anything? Team-based initiatives remove this issue.

In a study to increase near-miss reporting in physician practices,⁷ each practice received $1,500 per month when 10 near miss events were reported and 1 event was addressed. Individual staff members were not paid to report, but practices did offer small team-based awards. The authors found that reporting continued even when the study and the cash bonuses ended. In this approach, the incentive rewarded the team, rather than the individual, and created a sustainable infrastructure. Additionally, part of the incentive was to address at least 1 event per month, which promoted team-based problem-solving.

Costs

An important question in any health care setting is how this type of initiative will be funded. In the described study,³ for 1 year of the intervention, 516 residents received a $200 bonus, adding up to $103,200. In applying this model to other settings, the incentive amount and the potential number of eligible residents needs to be considered.

The authors of the study correctly state that in calculating the cost impact of a reporting incentive, the cost offset of an increased number of reports and resulting improvement activities should be considered and estimated. At minimum, every event filed requires review and follow-up. Some events may require root cause analysis, and others quality improvement initiatives. The hypothesis generated in this study, that residents report different and new patient safety events, needs further testing.

This program would also compete with other patient safety initiatives funded by the organization. An ROI analysis, which takes into account other competing priorities, is important to fully understand the organizational impact of this type of intervention. Leadership support is essential for any patient safety program. In addition, the question as to who would fund the initiative—the graduate medical education enterprise, health care system, hospital, or clinical practice—must be answered.

Conclusions

This study describes a program that promotes a culture of safety by eliciting reports from residents that would not have been previously reported. The increased reporting appears to provide a clearer picture of patient care errors through the eyes of residents, and gives a more complete picture of how health care looks on the frontlines. Additionally, the beneficial effects on trainee event reporting have been sustained for 3 years.

To make this initiative cost-efficient and promote long-term sustainability, a consideration may be to ensure that the incentive is team based, yet requires active individual participation. This would reduce potential discord within the health care team regarding pay. Funding the program requires a robust analysis that includes the ROI and how the program will be funded.