Reducing diagnostic errors is the next big challenge for patient safety.1 The diagnostic process is complex and involves making decisions under uncertain conditions and limited time.2 The National Academies of Sciences, Engineering, and Medicine’s (NASEM’s) 2015 report Improving Diagnosis in Health Care3 recommends that accrediting organizations and Medicare “require that health care organizations have programs in place to monitor the diagnostic process and identify, learn from, and reduce diagnostic errors and near misses in a timely fashion.”3 In 2019, however, external incentives and policies to implement NASEM’s recommendations are still absent and practical guidance on how to develop these programs is limited. Improving diagnosis remains a rare organizational priority. Yet diagnostic safety improvement efforts should be a priority for all stakeholders who value patient safety and quality improvement. Beyond health care organizations, these include payers, because diagnostic errors lead to increased costs and utilization, and accrediting bodies, which aim to reduce patient harm and protect consumer interests by maintaining standards and safety.
The pursuit of diagnostic excellence involves (1) reducing errors; (2) improving diagnostic processes to address harm from missed, delayed, or incorrect diagnoses, overdiagnosis, and overtesting; and (3) communicating diagnosis-related explanations to patients effectively. Accurate and timely diagnosis depends on improving not only clinician cognition and behavior, for which there are few successful interventions,4 but also the robustness of underlying health care delivery and microsystem factors, many of which are potential targets for intervention.5
The NASEM report3 highlights that most health care organizations do not have processes in place to identify diagnostic errors in clinical practice. Nevertheless, gathering this information, learning from transpired events, and implementing improvements as a response are all essential steps for progress. The report calls for “adopt[ing] policies and practices that promote a nonpunitive culture that values open discussion and feedback on diagnostic performance.” To help achieve NASEM’s vision for progress, there is a compelling need to create health care organizations that value learning and exploration of diagnostic excellence (which we henceforth refer to as LEDE organizations). The characteristics of LEDE organizations ideally would include (1) use of diagnostic quality and safety surveillance methods to create a continuous feedback cycle that engages frontline clinicians and (2) leaders who act upon outcomes data to reduce variations in diagnostic care and prevent diagnostic harm. Based upon current scientific understanding as well as our recent experiences in developing such a learning organization at Geisinger, an integrated health care system in Pennsylvania, we propose a 5-point action plan (Figure 1) and corresponding policy levers to support development of LEDE organizations. While our recommendations are directly applicable to larger academic medical centers and health care systems, most can be readily adapted by any type of health care organization.
;)
Figure 1: Five-point action plan for developing health care organizations for learning and exploration of diagnostic excellence (LEDE organizations). Note: The phrase “tyranny of metrics” is taken from the title of the 2018 book by Jerry Z. Muller.
14A 5-Point Action Plan to Develop a LEDE Organization
1. Implement a virtual hub to coordinate organizational activities for improving diagnosis
Diagnostic events are challenging to address because they typically involve teams, evolve across multiple settings, and overlap with other patient care processes (e.g., screening, prevention, management). Diagnostic accuracy and safety are often ignored in initiatives focused on accreditation- and regulation-driven safety concerns, such as readmissions and poor patient experience scores, despite being inherently related to some of these priorities. The coordination of activities for improving diagnosis requires first establishing a centralized or virtual hub, such as a multidisciplinary advisory council or committee of diverse stakeholders. Members of this virtual hub could include chief medical officers; clinical department leaders; frontline clinicians; information technology (IT) analysts; and institutional quality, safety, and risk management professionals. Representatives of continuing medical education (CME)/continuing professional development offices and others that help individuals and teams learn should be included to help create initiatives to involve clinicians in diagnostic learning and improvement activities. This virtual hub can be tasked to identify risks among competing priorities and to prioritize interventions that cross intra-institutional silos.
At Geisinger, the Committee to Improve Clinical Diagnosis (CICD) has garnered institutional leadership support and created a formal charter to develop innovative approaches to identify and analyze diagnostic missed opportunities and provide recommendations for high-volume, high-risk events. The CICD aims to pursue diagnostic excellence through a program that identifies and seeks to reduce diagnostic errors while promoting learning and a culture of safety. This virtual hub began in 2014 as an exploratory work group and, by 2017, developed into the full committee, which includes several of the aforementioned clinical and operational stakeholders. The CICD works with academic officers, residency and fellowship program directors, and representatives from the Geisinger Commonwealth School of Medicine, and it partners with Geisinger’s Center for Professionalism and Well-Being. To mitigate confidentiality concerns that might arise while expanding lessons learned to all clinicians, not just those involved in the case being analyzed, the CICD was instituted as a subcommittee of the health system’s peer-review process. Table 1 suggests approaches for developing a virtual hub based on the CICD’s functions, many of which could be adapted by other health care organizations.
Table 1: Suggested Approaches for LEDE Organizations to Develop a Virtual Hub Based on Functions of Geisinger’s Committee to Improve Clinical Diagnosis (CICD)
Over the past 2 years, the CICD has analyzed diagnostic events reported through risk management and other clinical sources, reviewing missed opportunities in diagnosis in more than 220 cases across a variety of clinical settings. Over the past year, the CICD has initiated partnerships with clinical leaders and laid the foundation for creating mechanisms to provide supportive and nonpunitive feedback to frontline clinicians by sharing more than 100 cases of missed opportunities with different clinical departments. To understand how best to provide such feedback, the CICD has partnered with the emergency medicine, hospital medicine, and primary care departments and a team of researchers to pilot a feedback toolkit. This toolkit, developed by a multidisciplinary team, provides recommendations on providing feedback on missed opportunities in a sensitive, nonpunitive, and constructive manner to improve the effectiveness of the intended message. It provides guidance for helping set the stage for a learning environment and letting clinicians know what to expect in such feedback sessions.6 In addition to case reviews, feedback sessions include discussion and debrief of the decision-making process, clinical context, and system barriers, all of which can inform learning.
The CICD’s tasks are constantly evolving in an effort to create a learning health system, but this virtual hub is slowly changing the local culture to encourage acknowledging, reporting, and learning from missed diagnostic opportunities. Such a hub can also support and implement educational initiatives and cognitive training for clinicians. For example, the CICD recently facilitated the development of a diagnostic error reduction curriculum for interns.
Creation of a committee or advisory council that stimulates exploration of diagnostic events should be the norm in all health care organizations and should be the first step in the LEDE journey. Leadership could be engaged by emphasizing that timely and accurate diagnosis is closely tied to ongoing quality improvement initiatives linked to payment or public reporting, such as Hospital Compare.7 Improving diagnostic safety also mitigates risk and could potentially reduce malpractice-related payments.
However, absent any other incentives, a virtual hub could be considered as a condition for accreditation or a requirement for Medicare conditions of participation to ensure that health care organizations devote time and attention to this core patient safety issue. Basic activities could include developing an institutional framework to gather intelligence related to diagnostic near misses and adverse events with the ultimate vision of influencing and promoting a culture that values diagnostic excellence and transparency.
2. Participate in novel scientific initiatives to generate and translate evidence
The “basic science” of diagnostic excellence is rapidly evolving, and a recent $85 million investment from the Gordon and Betty Moore Foundation will generate additional evidence for practice improvement.8 LEDE organizations will need to expand their capacity and infrastructure to keep pace with emerging research and implement newly developed best practices. At present, effective and meaningful partnerships between safety researchers, health systems leaders, and clinicians are limited to a few health systems.9 To help LEDE organizations gather and apply evidence to improve diagnosis, a researcher-in-residence (RIR) model,10 similar to an embedded researcher model,11 could be useful. The RIR model “positions the researcher as a core member of a delivery team, actively negotiating a body of expertise which is different from, but complementary to, the expertise of managers and clinicians.”10 At Geisinger, the RIR serves as the scientific liaison between the organization’s internal operations and quality improvement teams and an external research team (described below), helping create actionable evidence for organizational decision makers.
The RIR at Geisinger is a central component of a unique collaboration between diagnosis researchers and clinical operations stakeholders, funded by the Moore Foundation since 2017. The collaboration, titled the Safer Dx Learning Lab, is helping develop and expand the learning health system model to measure and improve diagnostic safety. The lab funds the full-time Safer Dx RIR at Geisinger, whose tasks include helping build partnerships to align the priorities of the CICD and clinical operations teams with those of the research team, coordinating efforts to ensure that all operational activities are informed by evidence, and promoting continuous evaluation and learning of what works and what does not. To create a shared mental model to address the complexity of diagnostic error, the RIR helped develop a local protocol to review cases shared with the CICD for missed diagnostic opportunities using standardized instruments and validated review procedures.
New models of embedded and partnership research are evolving9 and need federal support for adoption.12 Broader evaluation and adoption of such models could become a reality through the Consolidated Appropriations Act of 2018 (Pub L No. 115-141), which requested the Agency for Healthcare Research and Quality (AHRQ) convene a federal interagency workgroup (including NIH representatives) focused on enhancing scientific research to improve diagnosis. The Department of Veterans Affairs has already stimulated partnerships between researchers and clinical operations leaders by funding RIRs and national centers of excellence that focus on research and implementation of practical tools to improve patient safety.13 Such partnerships might be more feasible for academic medical centers to implement but are relevant to all types of health care organizations in the long term. Multidisciplinary partnerships between clinicians, health system leaders, and researchers are essential to solve complex problems and should leverage embedded researcher models already in use.11
In the Safer Dx Learning Lab, the research team is based at Baylor College of Medicine in Houston, Texas, and works remotely with the clinical operations and leadership team based at Geisinger in Danville, Pennsylvania, with the help of the RIR. Nonacademic health systems could seek to partner with external or university-based patient safety researchers, who could provide methodology to study the magnitude of the problem and help implement evidence-based strategies to address diagnostic error. As one example of an RIR serving as a liaison, the Safer Dx RIR has catalyzed project engagement with multiple stakeholders in diverse roles related to contracts, legal, IT, informatics, data analytics, risk management, quality and safety, finance, and clinical and operational leadership.
3. Avoid the “tyranny of metrics”14 by focusing on measurement for improvement
Although measurement has become foundational to quality and safety improvement, no reliable and valid measures for diagnostic excellence currently exist.15 However, LEDE organizations have the opportunity to explore health IT capabilities to use their own ever-increasing stores of electronic data for learning, research, and local quality improvement.16 Machine learning and data science techniques could help inform development of new algorithms to identify or prevent diagnostic harm. Rather than using measures to reward or punish, LEDE organizations could focus on creating actionable insights from safety data and channeling them to the right people in the right organizational roles.17 This will support users in understanding and responding to data and implementing solutions based on the gathered intelligence. Thus, measurement for improvement can proceed while robust quality measures for diagnosis are still being developed.
We suggest LEDE organizations begin using electronic trigger (e-trigger) tools to identify missed opportunities in diagnosis.16 These tools mine vast amounts of clinical and administrative data to identify signals for likely adverse events and help users select a manageable number of records requiring additional human review to identify and confirm the presence of diagnostic error. We previously developed e-trigger algorithms to identify patterns of care suggestive of missed or delayed diagnoses in both primary care and inpatient settings18–20; for example, analysis of primary care or emergency department (ED) visits followed by unplanned hospitalizations within a week could be indicative of something missed at the first visit18 or other potential problems with the diagnostic process. Analyzing such cases can provide information on diagnostic process breakdowns and contributory factors,21 thus generating learning and feedback for improvement purposes. These records could be reviewed using the Revised Safer Dx Instrument,22 a structured data collection tool to help reviewers identify potential diagnostic errors from medical records in a standardized way. This information can help users in analysis and safety improvement efforts and provide clinicians actionable data for feedback and reflection.
The Safer Dx Learning Lab is taking a systematic measurement approach to learning how health care systems can enhance the safety and accuracy of the diagnostic process. The lab is using data from multiple sources—the health care system (risk management, e-triggers), clinicians, and patients—to identify missed opportunities in diagnosis. For instance, the lab is developing an e-trigger to identify when a patient visits primary care or the ED and then dies within 30 days as well as another e-trigger to identify unexpected care escalations (or death) within a defined time period among patients admitted to Geisinger’s hospitals. After an initial rigorous review by the lab team, missed opportunities are reviewed by the CICD to further understand how to uncover contributory factors and provide feedback to individuals, teams, and divisions. Several aspects of the diagnostic process—including patient–provider interactions, diagnostic test performance and interpretation, the referral process, appropriate follow-up of test results, and patient factors—are considered by the lab and the CICD to generate lessons based on a given event.21 One of the goals of the Safer Dx Learning Lab is to help develop a repository of best practices that can be used to create a future learning collaborative to share tools and strategies on what works and what challenges to expect while tackling diagnostic errors.
Accrediting organizations and payers could stimulate similar measurement efforts at other health care organizations by requiring them to demonstrate credible activities related to learning from measurement. Health care organizations could demonstrate how clinicians, patient safety teams, and data scientists collaborate to create actionable insights for diagnostic safety or process improvement. Policy efforts could support simplified reporting from clinicians, who have substantial knowledge of diagnostic concerns but find reporting onerous. For example, AHRQ’s Common Formats program23 could champion the development of streamlined standardized reporting protocols to collect minimal essential information about diagnostic concerns from clinicians for subsequent analysis by institutional safety teams.
4. Engage clinicians in improving diagnosis activities and frame missed opportunities as learning opportunities
Clinicians should be involved in analyzing and learning from missed opportunities in diagnosis and should help lead efforts to change processes and systems determined to be unsafe. In addition, diagnostic missed opportunities should be framed positively as learning opportunities rather than negatively as errors. The Safer Dx Learning Lab, in collaboration with the CICD, has implemented a hotline and created methods within the electronic health record (EHR) and paging system for clinicians to easily report diagnostic learning opportunities. The chief medical officer and the chief legal officer at Geisinger also developed a newsletter, entitled Teachable Moments, to share deidentified cases and lessons learned broadly with the staff. Additionally, clinicians should participate in self-assessment; for example, they could use the Revised Safer Dx Instrument22 to review a small sample of their own flagged records to assess how they approached diagnosis in certain situations. This approach would serve both educational and quality improvement needs and could be tied to CME and maintenance of certification (MOC) activities.
In addition to initiatives to strengthen clinical reasoning skills,24,25 newly defined competencies should be considered for inclusion in educational programs to improve diagnosis.26 Skills that need to be promoted include managing and communicating uncertainty, listening to and communicating effectively with patients and caregivers, tracking patient outcomes to improve feedback and learning, seeking assistance in cases of diagnostic uncertainty, and reconsidering the diagnosis when the patient does not improve as expected. These skills are essential hallmarks of humble and accountable clinicians. Training strategies should include simulation, with exposure to a variety of case presentations, and reflective practice (i.e., critically reflecting on one’s own reasoning and decisions) to reactivate and restructure knowledge.27 In addition to promoting these skills, CME and MOC should engage clinicians in quality improvement activities focused on implementing tools and strategies related to the diagnostic process. For instance, CME and MOC could emphasize implementing measurement activities (such as those described above), managing diagnostic data to make sure they are acted upon,28 conducting a self-assessment of reliability of communication and reporting of test results using the Office of the National Coordinator for Health Information Technology’s SAFER Guides,29 closing the referral loop,30 and enhancing teamwork and communication with patients.31
However, clinicians have limited time for participation in improvement activities. To avoid the perception of diagnostic improvement initiatives as another “checkbox,” LEDE organizations should implement targeted strategies to engage clinicians with minimal disruption to their routine workflow. Reduction of administrative burdens,32 for example, would enable clinicians to engage more meaningfully in self-assessments as well as in the implementation of solutions such as decision support tools and huddles to discuss challenging cases. One health care system reduced administrative burdens through its Getting Rid of Stupid Stuff program,33 which included crowd-sourcing suggestions for eliminating wasteful work related to the EHR and improving efficiency for necessary documentation. One outcome of this program was eliminating 10 of the 12 most frequent EHR alerts for physicians because most of them were being ignored. We suggest other health care systems create similar programs.
Becoming a LEDE organization may be intrinsically motivating34 to clinicians and organizations alike because diagnostic excellence is both meaningful and necessary for achieving positive outcomes related to patient experience and safety. Pursuing diagnostic excellence requires “mastery” of certain exemplary competencies35 and capabilities, including giving and learning from constructive feedback, as well as meaningful and actionable measurement of diagnostic performance, providing frontline staff with needed resources, and supporting opportunities for professional development. Learning organizations that engage frontline clinicians and teams in decisions and processes that affect them facilitate collaborative opportunities to improve care.36 Collaborative initiatives, such as the Centers for Medicare and Medicaid Services’ Partnership for Patients, can be further developed by LEDE organizations to harness the intrinsic motivation of clinicians to do the best for their patients.36
Payment innovations involving both clinicians and organizations are also needed to support cognitive work and engagement in safety improvement activities.37 Diagnostic performance should be prioritized in value-based payment models. Changes in both traditional fee-for-service Medicare and the recently proposed Alternative Payment Model (APM) structure38 could allow for the cognitive processes and reflection needed for accurate and timely diagnosis while accounting for diagnostic uncertainty. Performing quality improvement activities related to improving diagnosis could be a part of the accountability of APM recipients. Documentation guidelines could be improved and made less burdensome, and payment structures could better support second opinions, collaborative care models, and team-based diagnosis. To better engage clinicians in diagnosis improvement activities, policy efforts could employ strategies from behavioral sciences to harness clinicians’ intrinsic motivation using some of the determinants they value,39 such as relationships with patients and superior diagnostic decision-making skills.
5. Develop an accountable culture of engaging and learning from patients
LEDE organizations should develop a culture of engaging and learning from patients,40 who are often underexplored sources of information. For example, promoting a culture of “no news is not good news” to encourage patients to expect timely receipt of test results may help reduce the high frequency of test results without follow-up. LEDE organizations could also develop and implement methods to collect patient reports of diagnostic concerns.41 Although most organizations conduct patient surveys and review patient complaints, the linkage to safety is limited. However, patient data often identify safety issues that are hard to pinpoint through traditional data sources.41 The Safer Dx Learning Lab researchers are evaluating patient complaint data and patient experience survey scores for potentially missed diagnostic opportunities.
Recently, the AHRQ funded the development of the Diagnostic Safety Supplemental Item Set to be administered with the Medical Office Survey on Patient Safety Culture, a provider and staff survey that evaluates institutional safety culture.42 Similar items focusing on diagnostic excellence culture could be developed to supplement widely implemented patient surveys (e.g., Hospital Consumer Assessment of Healthcare Providers and Systems, Press Ganey), making these tools more useful for assessing and raising institutional accountability for diagnosis.
Conclusion
Reducing diagnostic errors, a high-priority patient safety challenge, involves a diverse set of stakeholders and requires a collaborative shared responsibility approach. Stakeholders include not just clinicians and clinical educators but also policymakers, health system leaders, payers, and accrediting bodies. Next steps involve implementing practical recommendations and innovations to improve diagnosis at the level of the health care organization. The 5-point action plan outlined here can stimulate progress in scientific, practice, and policy initiatives for achieving diagnostic excellence and reducing preventable patient harm.
Acknowledgments:
The authors would like to thank the members of Geisinger’s Committee to Improve Clinical Diagnosis for their support, their work in learning from diagnostic opportunities, and making diagnosis an organizational priority.
References
1. Singh H, Graber ML. Improving diagnosis in health care—The next imperative for patient safety. N Engl J Med. 2015;373:2493–2495.
2. Graber M, Gordon R, Franklin N. Reducing diagnostic errors in medicine: What’s the goal? Acad Med. 2002;77:981–992.
3. National Academies of Sciences, Engineering, and Medicine. Improving Diagnosis in Health Care. 2015.Washington, DC: National Academies Press;
4. Ely JW, Graber ML, Croskerry P. Checklists to reduce diagnostic errors. Acad Med. 2011;86:307–313.
5. Graber ML, Trowbridge R, Myers JS, Umscheid CA, Strull W, Kanter MH. The next organizational challenge: Finding and addressing diagnostic error. Jt Comm J Qual Patient Saf. 2014;40:102–110.
6. Meyer AND, Singh H. The path to diagnostic excellence includes feedback to calibrate how clinicians think. JAMA. 2019;321:737–738.
7. Medicare.gov. Hospital Compare: Hospital Compare overall hospital rating.
https://www.medicare.gov/hospitalcompare/Data/Hospital-overall-ratings-calculation.html. Accessed October 8, 2019.
8. The Gordon and Betty Moore Foundation. The Moore Foundation invests $85 million to improve diagnostic performance. November 1, 2018.
https://www.moore.org/article-detail?newsUrlName=the-moore-foundation-invests-$85-million-to-improve-diagnostic-performance. Accessed October 15, 2019.
9. Kupersmith J, Eisen S. A new approach to health services research. Arch Intern Med. 2012;172:1033–1034.
10. Marshall M, Pagel C, French C, et al. Moving improvement research closer to practice: The researcher-in-residence model. BMJ Qual Saf. 2014;23:801–805.
11. Luft HS, Tai-Seale M, Greene SM. Advancing learning health systems through embedded research: The 23
rd Annual Conference of the Health Care Systems Research Network. J Patient Cent Res Rev. 2017;4:139–143.
12. Bindman AB, Pronovost PJ, Asch DA. Funding innovation in a learning health care system. JAMA. 2018;319:119–120.
13. Bates DW, Singh H. Two decades since To Err Is Human: An assessment of progress and emerging priorities in patient safety. Health Aff (Millwood). 2018;37:1736–1743.
14. Muller JZ. The Tyranny of Metrics. 2018.Princeton, NJ: Princeton University Press;
15. Singh H, Graber ML, Hofer TP. Measures to improve diagnostic safety in clinical practice. J Patient Saf. 2019;15:311–316.
16. Murphy DR, Meyer AN, Sittig DF, Meeks DW, Thomas EJ, Singh H. Application of electronic trigger tools to identify targets for improving diagnostic safety. BMJ Qual Saf. 2019;28:151–159.
17. Vincent C, Burnett S, Carthey J. Safety measurement and monitoring in healthcare: A framework to guide clinical teams and healthcare organisations in maintaining safety. BMJ Qual Saf. 2014;23:670–677.
18. Singh H, Thomas EJ, Khan MM, Petersen LA. Identifying diagnostic errors in primary care using an electronic screening algorithm. Arch Intern Med. 2007;167:302–308.
19. Singh H, Giardina TD, Forjuoh SN, et al. Electronic health record-based surveillance of diagnostic errors in primary care. BMJ Qual Saf. 2012;21:93–100.
20. Bhise V, Sittig DF, Vaghani V, Wei L, Baldwin J, Singh H. An electronic trigger based on care escalation to identify preventable adverse events in hospitalised patients. BMJ Qual Saf. 2018;27:241–246.
21. Singh H, Sittig DF. Advancing the science of measurement of diagnostic errors in healthcare: The Safer Dx framework. BMJ Qual Saf. 2015;24:103–110.
22. Singh H, Khanna A, Spitzmueller C, Meyer A. Recommendations for using the Revised Safer Dx Instrument to help measure and improve diagnostic safety. Diagnosis (Berl). 2019;6:315–323.
23. Agency for Healthcare Research and Quality. Patient Safety Organization (PSO) Program: Common Formats.
https://www.pso.ahrq.gov/common. Accessed October 15, 2019.
24. Olson APJ, Singhal G, Dhaliwal G. Diagnosis education—An emerging field. Diagnosis (Berl). 2019;6:75–77.
25. Society to Improve Diagnosis in Medicine. Clinical reasoning toolkit.
https://www.improvediagnosis.org/clinicalreasoning. Accessed October 8, 2019.
26. Graber ML, Rencic J, Rusz D, et al. Improving diagnosis by improving education: A policy brief on education in healthcare professions. Diagnosis (Berl). 2018;5:107–118.
27. Singh H, Zwaan L. Web exclusives. Annals for hospitalists inpatient notes—Reducing diagnostic error—A new horizon of opportunities for hospital medicine. Ann Intern Med. 2016;165:HO2–HO4.
28. Singh H, Graber M. Reducing diagnostic error through medical home-based primary care reform. JAMA. 2010;304:463–464.
29. Sittig DF, Singh H. Toward more proactive approaches to safety in the electronic health record era. Jt Comm J Qual Patient Saf. 2017;43:540–547.
30. Institute for Healthcare Improvement, National Patient Safety Foundation. Closing the Loop: A Guide to Safer Ambulatory Referrals in the EHR Era. 2017.Cambridge, MA: Institute for Healthcare Improvement;
31. Agency for Healthcare Research and Quality. TeamSTEPPS.
https://www.ahrq.gov/teamstepps/index.html. Accessed October 8, 2019.
32. Downing NL, Bates DW, Longhurst CA. Physician burnout in the electronic health record era: Are we ignoring the real cause? Ann Intern Med. 2018;169:50–51.
33. Ashton M. Getting rid of stupid stuff. N Engl J Med. 2018;379:1789–1791.
34. Judson TJ, Volpp KG, Detsky AS. Harnessing the right combination of extrinsic and intrinsic motivation to change physician behavior. JAMA. 2015;314:2233–2234.
35. Kao AC. Driven to care: Aligning external motivators with intrinsic motivation. Health Serv Res. 2015;50(suppl 2):2216–2222.
36. Berenson RA, Rice T. Beyond measurement and reward: Methods of motivating quality improvement and accountability. Health Serv Res. 2015;50(suppl 2):2155–2186.
37. Berenson R, Singh H. Payment innovations to improve diagnostic accuracy and reduce diagnostic error. Health Aff (Millwood). 2018;37:1828–1835.
38. Centers for Medicare and Medicaid Services. Quality payment program.
https://www.cms.gov/Medicare/Quality-Payment-Program/Quality-Payment-Program.html. Accessed October 15, 2019.
39. Wynia MK. The risks of rewards in health care: How pay-for-performance could threaten, or bolster, medical professionalism. J Gen Intern Med. 2009;24:884–887.
40. McDonald KM, Bryce C, Graber M. The patient is in: Patient involvement strategies for diagnostic error mitigation. BMJ Qual Saf. 2013;22(suppl 2):ii33–ii39.
41. Giardina TD, Haskell H, Menon S, et al. Learning from patients’ experiences related to diagnostic errors is essential for progress in patient safety. Health Aff (Millwood). 2018;37:1821–1827.
42. Agency for Healthcare Research and Quality. Under development: New diagnostic safety supplemental item set for the medical office SOPS. March 2019.
https://www.ahrq.gov/sops/events/news/dxsafety.html. Accessed October 15, 2019.
