Pediatric Diagnostic Safety: State of the Science and Future Directions

Research To Understand the Diagnostic Process and Missed Diagnostic Opportunities in Pediatric Care Settings

We have only just started to develop an understanding of the extent of the problem of MDOs, much less a comprehensive theory of the diagnostic process and its vulnerabilities in pediatric care. Additional work is needed to create conceptual models of the diagnostic process specific to the clinical settings where pediatric care is provided (ambulatory care, EDs, general inpatient wards, intensive care units).

Although NAM’s conceptual model of the diagnostic process is useful in many ways, adapting it to develop more tailored models will be essential to identify the barriers and facilitators of accurate and timely diagnosis unique to each setting. One example is the recent operational framework put forth by Mahajan, et al., to study the diagnostic process in the ED.¹³⁰

It is also essential that investigators study and learn from diagnostic process successes despite considerable variation in disease presentation and healthcare delivery (safety-II) in addition to understanding failure in the diagnostic process (safety-I).¹³¹ Within such conceptual models, potentially high-risk areas and factors investigators have started to address include:

• Pediatric transitions of care.¹³²
• Clinicians’ cognitive load.¹³³
• The role of diagnostic uncertainty.¹³⁴
• Teamwork and communication.¹³⁵
• The clinical work environment and culture.² 
• Disparities in diagnostic testing and outcomes.^136-138

Research on Interventions To Improve Pediatric Diagnosis

In tandem with biomedical discovery fueling the more precise diagnosis of new types and subtypes of pediatric disease, interventions addressing the sociotechnical aspects of pediatric diagnosis that cut across diseases must also be developed, implemented, and evaluated. Because diagnostic efforts cannot be extricated from the systems, circumstances, and environments within which patient care is delivered and healthcare teams must work,¹³⁹ an intense focus on the diagnostic process is needed. Such focus could target high-yield systems-based interventions for diagnostic safety and excellence.

Using a modified Delphi process to reach consensus, Singh, et al., determined 10 practical approaches that healthcare systems and organizations could implement to identify and support opportunities to improve diagnostic safety.¹⁴⁰ These include:

• Ensuring appropriate leadership engagement.
• Building a diagnostic safety culture that supports event reporting and learning.
• Developing infrastructure for MDO measurement and improvement activities.
• Engaging patients.

Teamwork, team cognition, safety culture, communication, and patient/family-centeredness are some of the important aspects of the sociotechnical milieu of pediatric diagnosis that will likely receive more attention in the future as diagnosis-focused interventions are developed and tested.

Promising advances that address these aspects of diagnosis have been developed initially in adult clinical settings, many of which have been translated effectively for use in pediatrics. These include interventions to improve clinicians’ calibration through feedback^98,116; tools to adjust how healthcare teams think about diagnosis, such as diagnostic timeouts¹⁴¹; and programs to improve communication about diagnostic uncertainty.⁹³

Interventions focused on clinicians likely apply across adult and pediatric settings because of the many similarities in clinician workflows regardless of the patient population. However, an important challenge for pediatrics would be to determine the applicability and adaptability of interventions that involve patient-centeredness and family involvement in the diagnostic process.^120,142 Patient-focused interventions in pediatrics require more emphasis on family partnerships because of the inherent universal involvement of parents and caregivers in the diagnostic process.¹⁴³

Many potential resources that can be translated into tools to improve these various aspects of the diagnostic process are now available. Technological innovations in health information technology have historically been powerful allies in making healthcare safer in pediatrics, but they are not without pitfalls,¹⁴⁴ as noted in the discussion of artificial intelligence (AI) later in this section.

Existing technology, such as the EHR and telemedicine, can be redesigned or redeployed in the service of diagnostic safety. For example, efforts to identify patients with diagnostic uncertainty and provide clinician feedback to improve calibration have made clever use of the EHR to serve both as an information source and information delivery vehicle.^96,98,145

The use of telemedicine in pediatrics has expanded during the COVID-19 pandemic.^146-148 Telemedicine has provided more equitable access to pediatric subspecialists for children who are underserved or living in remote locations.^149,150

New technology such as clinical decision support powered by AI can have a powerful impact on pediatric diagnosis.¹⁵¹ For example, recent work showing how AI can be used for pediatric diagnosis includes:

• Studies on AI’s ability to diagnose common pediatric conditions after being trained on large medical record-derived datasets.^152,153
• Web-based AI systems that can provide advice to parents of sick children (including whether their child requires a doctor’s visit).^46,154
• AI-driven vital sign parameters to prevent alert fatigue in pediatric acute care units.¹⁵⁵ 
• AI methodologies to automate or augment pediatric biomedical image interpretation.^156,157

However, as impressive as these initial forays are, AI-powered applications for pediatric diagnosis are still very much in their infancy. Similar to other new technologies, AI’s thoughtful implementation into complex clinical workflows must be carefully considered just as much as its intended functions. In addition, we must consider challenges such as developing clinician trust and mitigating possible adverse outcomes, including inaccuracy due to shortcomings in training datasets and propagation of harmful healthcare disparities and bias.^158,159

Partnerships for Pediatric Diagnostic Safety

Developing a deep understanding of the diagnostic process and implementing interventions to promote diagnostic safety will require the inclusion of diagnostic excellence-focused goals and programs within and strong partnerships across a multitude of stakeholder organizations.

Active efforts must be made to include diagnosis-focused clinical and health services research in the portfolios of pediatric research collaborative groups to ensure robust multicenter research. For example, the international Pediatric Emergency Research Network¹⁶⁰ served as the primary platform for a recent study delineating the types of MDOs reported by pediatric emergency providers.¹⁴

In tandem, pediatric patient safety and quality collaboratives must include diagnostic excellence as a key patient safety goal to prioritize projects that improve diagnosis. To illustrate, the Children’s Hospital Association, a national collaborative aimed at improving the safety and quality of pediatric hospital care, has developed a Diagnostic Safety Toolkit, which guides organizations in improving communication to prevent MDOs.

Likewise, subspecialty-specific pediatric professional societies and medical journals should ensure that their projects, programming, and publications appropriately include work focused on diagnostic excellence. In recent years, AAP, the Society of Critical Care Medicine, and their corresponding official journals, Hospital Pediatrics and Pediatric Critical Care Medicine, have prominently featured work on pediatric diagnostic safety.

Strong partnerships must be forged between investigators, pediatric research and professional networks, and pediatric patient safety collaboratives to ensure a throughline across discovery, implementation, and dissemination, resulting in more timely translation of findings to broadly benefit pediatric diagnosis.

Multistakeholder organizations, such as SIDM and the Society for Medical Decision Making, must continue to provide resources and support for a wide range of research and quality improvement initiatives to improve diagnosis. They must also ensure that pediatric-specific projects are well represented.

These societies are especially relevant when they provide venues for stakeholder engagement and collaboration. Engaged patients, families, and frontline clinicians are particularly important in pediatric diagnostic safety work. Thus, their input must be integrated into all initiatives aiming to improve diagnosis. One example is SIDM’s Patients Improving Research in Diagnosis (PAIRED) program. This program trains and connects patient partners (many of whom are parents and family members of children affected by diagnostic error) with investigators to ensure that patients’ and families’ voices are integrated into diagnostic safety research.

Funding organizations such as AHRQ and the Gordon and Betty Moore Foundation have been significant champions of projects to improve pediatric diagnosis. For example, AHRQ’s Measure Dx pilot implementation project designed to develop capacity for diagnostic safety recruited health systems to participate; nearly half of these systems were freestanding children’s hospitals.¹⁶¹

The Moore Foundation, through its diagnostic excellence initiative, has funded numerous fellowships, quality improvement projects, and clinical education through various organizations, including:

• SIDM.
• NAM.
• Society of Bedside Medicine.
• American Board of Medical Specialties.
• Council of Medical Specialty Societies (CMSS).
• Institute for Healthcare Improvement.

These grants and partnerships support researchers and clinicians (many of whom work in pediatric settings) in implementing a variety of initiatives to improve pediatric diagnosis. They also support efforts to expand the pediatric diagnostic safety research and quality improvement workforce. For example, CMSS awarded one of 11 grants funded by the Moore Foundation to AAP to promote diagnostic excellence in ambulatory pediatrics.¹⁶² These collaborations ensure that pediatric healthcare priorities align with and benefit from the rapidly expanding expertise in diagnostic safety.

Finally, to ensure the continued growth and development of the field, investments are needed in research, including the research workforce, and programs that will develop future leaders in pediatric diagnostic safety research and improvement. Diagnostic reasoning curricula must be incorporated not only into medical school education but also into allied healthcare professions’ education.

Pediatric-specific graduate medical training programs must include didactic and practical education on diagnostic safety, which can be incorporated into standard patient safety and quality improvement education. Postgraduate fellowships in diagnostic safety research and quality improvement must also be offered by academic institutions, professional societies, and nonprofit organizations and agencies. Priority should go to healthcare professionals interested in improving diagnosis.

Consistently achieving pediatric diagnostic excellence will involve marshaling and coordinating resources from a broad array of disciplines, stakeholders, and institutions. Increased pediatric advocacy efforts and research funding focused on diagnostic safety are also needed.

Interdisciplinary research is needed to elucidate vulnerabilities in the diagnostic process specific to pediatric clinical settings and to rigorously investigate corresponding interventions that can improve diagnostic outcomes for children. At the same time, targeted efforts at the frontlines of pediatric care are needed to efficiently translate evidence-based interventions into measurable ways of improving diagnosis and, most importantly, reducing harm.

To accomplish this goal effectively and equitably, partnerships will need to be forged with patients, families, and frontline clinicians. Resources will also need to be shared among existing national and international pediatric research networks, pediatric patient safety collaboratives, professional societies, institutions of higher learning, and federal and private funding agencies. Finally, ensuring continuity and growth of the field requires that programs be established to develop and support learners at all stages to become future leaders in pediatric diagnostic safety research and improvement.