Improving Education—A Key to Better Diagnostic Outcomes

Improving diagnosis education will require good answers to each of these questions: Whom to teach, what to teach, how to teach it, and how to assess it? Using a competency-based framework is an important major recommendation for improving diagnosis-related education. We must define the goals we want to achieve in diagnosis education in order to design effective programs. Competency-based medical education aims to set forth the outcomes programs should aim to achieve in their learners, thereby allowing educators to meet this need.^40-47

Whom To Teach

Diagnosis has always been a team effort; that is, many individuals, including health professionals, patients, and their families, are involved in making a diagnosis. However, the widespread, formal and informal cultural recognition of the shared nature of the diagnostic process and its outcomes is more recent.

Various aspects of educational culture and other forces have led to a greater emphasis on diagnosis in some professions (e.g., physicians) and less in others, such as pharmacists. However, it is clear that all health professionals have a role on the diagnostic team, so all health professions education programs should have a substantial, role-appropriate curricular focus on diagnosis.^11,12,31,33 Increasing this focus will require culture change and buy-in at a national and institutional level and support in clinical training environments.

What To Teach

The traditional curriculum for medical students is designed to convey a large body of knowledge relevant to diagnosis, with a strong basis in anatomy, biochemistry, physiology, and pathophysiology. There is widespread agreement that building this foundation of knowledge is important and the most critical element in enabling diagnosis. Through a consensus-building process, a set of 12 diagnosis-related competencies in three domains (individual, team, and systems) that incorporate missing elements has been developed and endorsed.^48,49 The competencies elucidate knowledge, skills, and attitudes that are directly relevant to achieving diagnostic excellence.

Domain 1: Individual competencies

These individual competencies emphasize knowledge, skills, and attitudes that clinicians must have and perform (within their own professional role) to be an effective member of the diagnostic team. These largely relate to clinical reasoning and ways to mitigate the harmful effects of clinicians' cognitive fallibility. These competencies begin with hypothesis-driven data collection, followed by synthesis and differential diagnosis development, followed by use of decision-support resources. Finally, the importance of reflection and cognitive humility is emphasized.

In addition, competence must be demonstrated with regard to:

• Using decision-support resources for differential diagnosis. Many excellent programs are available that have the potential to help catch unusual conditions and common conditions presenting atypically.⁵⁰ Students rarely have exposure to these in current training programs. Instead, use of these programs must become the norm and a habitual component of clinical reasoning.
• Engaging with second opinions and consults.⁵¹ Fresh eyes catch mistakes, and input from experts is invaluable. Ultimately, crowd-sourcing may be a beneficial option, although this resource is in its infancy.
• Avoiding common cognitive pitfalls.⁵² Many cognitive errors arise from unconscious tendencies that can be avoided or at least recognized in time to avoid harm. Just taking the time to reflect on the differential diagnosis may help avoid too quickly settling on the first one that comes to mind.^53,54 Formal structured reflection is one of the most effective strategies for improving diagnostic performance in research studies, and students should be equipped to engage in this process to improve diagnosis.^54-57

Domain 2: System-related competencies

System-based breakdowns are common contributors to diagnostic error,⁵⁸ and using a sociotechnical perspective to understand diagnosis and diagnostic error has provided unique insights into how diagnosis can be improved.^59,60 Important competencies in this domain include:

• Learning to engage effectively with the EHR and available digital resources to maximum advantage. The EHR is a defining and important tool for essentially all modern healthcare; thus, formal training regarding its use, pitfalls, and opportunities is fundamental; 
• Using human factors principles to improve diagnosis, including addressing interruptions and distractions, time pressures, and clinical chaos; and
• Encouraging feedback and learning from errors and near-misses.^61,62

Domain 3: Teamwork competencies

Improving teamwork was the number 1 recommendation from the report on diagnostic error, citing the impact of teamwork on safety in aviation and other high-reliability organizations.¹⁶ Through interprofessional education, teamwork could be ingrained from the start and used to great advantage in diagnostic practice.⁶³

Unfortunately, few schools offer educational programs aimed at interprofessional practice, and some of these are brief or superficial. Further, learners may have little opportunity to observe or participate in deeply interprofessional shared diagnostic reasoning.⁶⁴ Teamwork in diagnosis can be optimized by:

• Encouraging patient engagement in diagnosis.^65,66 Diagnosis is, ideally, coproduced with clinicians and patients in a trusting partnership. To allow this partnering to occur, patient engagement must go beyond simply being a worthwhile goal for healthcare professions. Instead, training in this area should translate into specific skills and practices regarding effective listening, teach-back, and strategies to build therapeutic relationships, as well as ways to mitigate bias in healthcare encounters.
  Engaged patients have better health outcomes, and tools to promote patient engagement exist, including comprehensive recently released resources.⁶⁷ Sharing visit notes is an important way to develop a collaborative relationship. Therefore, learners should be equipped to develop notes and engage in other activities that promote patient and family understanding and ability to raise concerns, express values, and participate in shared decision making.⁶⁸ 
• Encouraging a collaborative approach to care. Input from everyone who has any role with a patient must be encouraged and valued; their unique approaches, knowledge base, and perspectives can only serve to enhance diagnostic outcomes. Engaging directly with laboratory professionals and radiologists is another practice that has substantial potential to improve the diagnostic process.^69,70 They can and should provide guidance and assistance on selecting tests, interpreting results, and sharing results with patients.
  Diagnostic management teams aimed at collaboration to improve diagnosis in a certain area are pioneering examples for future innovations.⁷¹ Of course, very real cultural and structural barriers exist to this deep engagement between healthcare team members, and too many examples show ineffective collaboration in healthcare. Thus, it is imperative that programs aimed at improving collaboration engage those in practice as well as those in formative educational programs to ensure that all members of the team are equipped to be effective team members. 
• AHRQ's TeamSTEPPS program presents a well-tested approach to improve team functioning in practice, and a new module (TeamSTEPPS for Diagnosis Improvement) specifically addresses teamwork in diagnosis.⁷² High-functioning, effective diagnostic teams should be studied and the practices disseminated. 

How To Teach It

Advances in the learning sciences provide key suggestions for improving diagnosis-related education.⁷³ These include suggestions to provide not only prototypical case presentations, but also repeated exposures to atypical presentations and other conditions and treatments that could cause the same symptoms. In addition, these experiences should be presented in authentic environments, with feedback and opportunities for reflection.⁷⁴ Therefore, diagnosis education must take place, as much as possible, in the clinical setting in which care is provided and learners should be substantively exposed to conditions and diagnostic processes in multiple different contexts.

Approaching chest pain, for example, is different in the nursing home, patient's home, emergency department, or federally qualified health center, and exposure across contexts is fundamental for learner competence. In addition, we must ensure that during their training, learners encounter the conditions for which they need to be competent. Most clinical curricula assign patients to learners in a relatively random way, meaning that a learner may not be exposed to even common conditions. Clinical curricula must be designed to ensure broad, intentional exposure to important conditions and be adaptive rather than fixed to ensure learner competence.

In addition, this education must be interprofessional as a rule, not an exception or extracurricular activity. An important paradigm shift to consider is for the default to become learning together across professions unless there is a good reason not to do so. If clinical practice is meant to be interprofessional, then education must be also, including preclinical and clinical educational experiences. For this change to occur, massive structural and cultural aspects must be addressed. However, we cannot rely on the same educational processes we have always used to get better results in the future. The goal is to create the team-based diagnosis that is envisioned for the future.^16,75,76

Such comprehensive approaches to team-based diagnosis education have not yet been implemented, but key aspects will guide development and implementation. Existing frameworks for interprofessional training are a helpful starting point. Team-based diagnosis education must equip learners for the roles they will play in relation to the roles of other healthcare professionals. In addition, learners need to be able to identify the contributions of different team members in different contexts. Finally, they need to be able to identify how interactions between people and between people and systems influence decision making, as well as how power structures, hierarchy, and unconscious bias affect diagnosis in practice.

How To Assess It

The ultimate test of competency in clinical reasoning is the ability to solve cases in real-world settings aligned with patients’ goals and values. Assessment of clinical reasoning has evolved in concert with our understanding of its complexity, and a range of instruments now exist that will be useful.^58,77-81 We must develop robust and longitudinal programs of assessment, both of learning and for learning, that measure diagnostic competence in the above domains across a breadth of conditions and contexts.

The removal of the United States Medical Licensing Examination Step 2 Clinical Skills Examination has left an opportunistic void for medical education programs to proactively design assessments that actually measure diagnostic competence in a robust and valid way, and early ideas are promising.^82,83 We have an opportunity to develop a national program of assessment that aims to ensure diagnostic competence and entrustment for practice across educational transitions. While the format and structure of such a program have yet to be developed, it is appealing to consider an approach that would ensure learners can perform their professions' activities with respect to diagnosis across a body of cases and contexts shared across institutions.

Technological innovations such as virtual patient cases may allow more rapid dissemination and uptake of such a program, although important gaps (especially around clinical skills) would remain. Further, the education community must ensure that we measure diagnostic competence as part of continuing professional development programs so that those in clinical practice are incentivized to keep pace with the rapidly changing field of diagnosis. This approach involves programs to review cases and obtain feedback about clinical reasoning in practice, robust peer review programs aimed at improvement, and educational activities to ensure clinicians are competent in emerging areas of diagnosis.

Instruments to assess team-based clinical reasoning are starting to emerge,^63,84 and more are needed to help understand the sources of variance in teams' diagnostic performance. The literature surrounding teamwork in healthcare, especially with a focus on quality and safety, is especially robust. It is clear that implementation of curricula focused on improving teamwork has substantial potential to benefit quality, safety, and operational functioning, although the impact on diagnostic performance remains to be studied.^85,86