Presentation
Understanding Resilience in Pediatric Primary Care Diagnosis
DescriptionIntroduction
Pediatric misdiagnosis, the failure to provide an accurate and timely explanation of a child’s health problem, is common, harmful, and understudied. A majority (54%) of general pediatricians report encountering a misdiagnosis at least once a month. In private pediatric practices, hypertension and adolescent depression are missed 50% of the time. Misdiagnosis can severely harm children due to delayed treatment. Forty-three percent of pediatric medical malpractice claims involve outpatient care, with 32% being diagnosis-related and resulting in injury or death. Misdiagnosis can therefore have an enormous and lasting impact on families’ psychological and financial health. Despite this, the diagnostic process in ambulatory pediatrics has not been studied in depth to identify opportunities for improvement and to understand the drivers of diagnostic excellence.
Traditionally, efforts to improve diagnostic safety have focused on identifying and preventing diagnostic errors (Safety-I). However, many have since realized that a singular focus on diagnostic missteps is insufficient. Exclusively targeting diagnostic error ignores the variability in diagnostic performance enabling success as well as system design ensuring that performance variability is safe. Shifting to a resilience engineering perspective that considers how individuals and systems manage complexity and uncertainty under dynamic conditions (Safety-II) will allow us to develop interventions that build on the capacity of the outpatient work system to adapt to the everyday challenges of pediatric diagnosis. Focusing on Safety II, we conducted an ethnographic study of ambulatory clinic visits in three pediatric healthcare systems to illustrate how healthcare teams, patients/families, and organizations contribute to resilience in diagnosis.
Methods
We performed focused ethnography (direct observation and semi-structured interviews) of clinic visits of children with multiple chronic conditions who have new or ongoing problems. Ethnography was conducted at general primary care, urgent care, and complex care clinics in three academic pediatric healthcare systems. Observation and interview guides based on the SEIPS 3.0 framework and Safety-II principles were developed to characterize variations in the diagnostic process (“work-as-done”) during information gathering, information integration and interpretation, and formulation/communication of the working diagnosis (and diagnostic uncertainty, if present). Purposive sampling was performed to ensure inclusion of patients with diverse racial, ethnic, spoken language, and socio-economic experiences. A trained researcher present in the exam room directly observed and audio-recorded all interactions between the patient/family and the clinical team during the visit. The main visit provider and a parent/guardian of the patient were each interviewed in-person or by phone by the same researcher within two weeks of the clinical encounter. Interviews were audio-recorded and transcribed. Parents/guardians who spoke languages other than English were interviewed with certified interpreters and English translations of their responses were transcribed.
We developed a codebook deductively based on resilience engineering concepts and inductively based on emerging data. The codes captured how clinicians and patients/families anticipate, monitor, respond, and learn throughout the diagnostic process and how robustness and graceful extensibility were demonstrated as they adapted to the various conditions under which diagnosis unfolded. Qualitative coding was performed by trained codersprimary and secondary coders were assigned to code field notes and interview transcripts generated from each patient. Discrepancies in coding were resolved by group consensus. Directed content analysis was used to characterize adaptations in diagnostic performance.
Results
We have conducted interviews and observations of 21patients as of September 19, 2024 (recruitment is ongoing). Thirteen (62%) patients were male, 8 (38%) were Black, 7 (33%) were White, and 5 (24%) were Hispanic. Sixteen patient families (76%) spoke English as their primary language, 2 (10%) spoke Spanish, while the remaining 3 each spoke Haitian Creole, Arabic, or Vietnamese. We identified the following key themes:
Anticipation. Based on their working diagnoses, clinicians anticipated possible disease courses and potential complications, which informed the nature and urgency of their next diagnostic actions. Clinicians also anticipated factors that may affect the diagnostic process (e.g., lab operating hours, home resources, patients’/families’ preferences) and tailored their diagnostic plans accordingly. Clinicians were able to anticipate potential problems more effectively if they had: 1) adequate medical knowledge, 2) substantial experience of the range of clinical presentations and outcomes, and 3) relevant clinical or situational information provided by other clinicians. Anticipation enabled advance planning of potential responses, increasing efficiency and timeliness of the diagnostic process. Using effective communication, patients/families were engaged by clinicians to become partners in monitoring the disease course and responding appropriately when necessary.
Monitoring. Monitoring patients’ response to treatment and their expected clinical course was integral to diagnosis, especially for those with diagnostic uncertainty. In an outpatient setting, monitoring required substantial resources (e.g., nurse staffing for patient follow-up). Monitoring patients required teamwork across the primary care team, subspecialty consultants, and patients/families. After receiving instructions from the clinician, most monitoring was performed by families at home and they were responsible for communicating concerns to the clinical team. Monitoring also allowed clinicians to learn the outcomes of their diagnostic decisions, providing feedback on their diagnostic performance.
Response. Information from the history, physical exam, test results, and clinical course informed how clinicians and patients/families responded, which may be in the form of additional testing, subspecialty consultation, and/or second opinion referrals. Responses were tailored to the individual patient’s circumstance and were a product of shared decision-making between the clinician and patients/families.
Learning. Clinicians learned to approach diagnosis in different ways through past experience, prior encounters with different patients/families, exposure to their colleagues’ varying diagnostic styles, and from available clinical decision support tools (e.g. institutional clinical pathways). Experience with diagnostic error prompted both individual clinician and organizational learning leading to changes in practice. Patients/families also learned from previous healthcare encounters, which shape the way they approach subsequent interactions.
Robustness. Individual clinicians contributed to robustness in diagnosis by confirming thoroughness in information gathering, ensuring that high risk patients received appropriate return instructions and follow-up, and reflecting on diagnostic decisions. Most clinicians considered their colleagues’ support as the single greatest source of robustness in outpatient diagnosis. Diagnostic excellence was institutionally supported by providing clinical teams with vital clinical decision support and communication tools (e.g., interpreter services), and by allowing adequate time per patient visit.
Graceful Extensibility. Clinical teams commonly extended themselves to ensure that the diagnostic process went smoothly, which included involving parents by phone during visits, seeing patients even after arriving late, and spending as much time as needed commensurate to the patient’s needs. Clinicians also noted that their colleagues generously extended their help for cases with significant diagnostic uncertainty.
Conclusions
In pediatric primary care, adaptations of clinicians, patient/families, and organizations are an integral part of the diagnostic process. These adaptations are dependent on the knowledge/expertise of clinicians, require significant organizational resources, involve shared decision-making with patients/families, and rely on patient/family engagement to implement effectively. Pediatric outpatient diagnosis is made more robust and extended to meet the needs of patients at both the individual clinician and organizational level. This challenges the idea that misdiagnosis occurs as a result of a single cognitive failure or one failed action of the clinician in an otherwise reliable system, with “one right way” to diagnose and linear relationships between process and outcome. It is the constant adaptations and interactions between clinicians, patients, families, and the wider organization that lead to diagnostic success or failure. This work represents an initial investigation of diagnostic performance variation and resilience in pediatric outpatient diagnosis. Further study is needed to understand specific adaptations that lead to improved diagnostic outcomes. Doing so will inform targeted interventions which build on the capacity of pediatric primary care systems to successfully adapt to the challenges of diagnosis.
Pediatric misdiagnosis, the failure to provide an accurate and timely explanation of a child’s health problem, is common, harmful, and understudied. A majority (54%) of general pediatricians report encountering a misdiagnosis at least once a month. In private pediatric practices, hypertension and adolescent depression are missed 50% of the time. Misdiagnosis can severely harm children due to delayed treatment. Forty-three percent of pediatric medical malpractice claims involve outpatient care, with 32% being diagnosis-related and resulting in injury or death. Misdiagnosis can therefore have an enormous and lasting impact on families’ psychological and financial health. Despite this, the diagnostic process in ambulatory pediatrics has not been studied in depth to identify opportunities for improvement and to understand the drivers of diagnostic excellence.
Traditionally, efforts to improve diagnostic safety have focused on identifying and preventing diagnostic errors (Safety-I). However, many have since realized that a singular focus on diagnostic missteps is insufficient. Exclusively targeting diagnostic error ignores the variability in diagnostic performance enabling success as well as system design ensuring that performance variability is safe. Shifting to a resilience engineering perspective that considers how individuals and systems manage complexity and uncertainty under dynamic conditions (Safety-II) will allow us to develop interventions that build on the capacity of the outpatient work system to adapt to the everyday challenges of pediatric diagnosis. Focusing on Safety II, we conducted an ethnographic study of ambulatory clinic visits in three pediatric healthcare systems to illustrate how healthcare teams, patients/families, and organizations contribute to resilience in diagnosis.
Methods
We performed focused ethnography (direct observation and semi-structured interviews) of clinic visits of children with multiple chronic conditions who have new or ongoing problems. Ethnography was conducted at general primary care, urgent care, and complex care clinics in three academic pediatric healthcare systems. Observation and interview guides based on the SEIPS 3.0 framework and Safety-II principles were developed to characterize variations in the diagnostic process (“work-as-done”) during information gathering, information integration and interpretation, and formulation/communication of the working diagnosis (and diagnostic uncertainty, if present). Purposive sampling was performed to ensure inclusion of patients with diverse racial, ethnic, spoken language, and socio-economic experiences. A trained researcher present in the exam room directly observed and audio-recorded all interactions between the patient/family and the clinical team during the visit. The main visit provider and a parent/guardian of the patient were each interviewed in-person or by phone by the same researcher within two weeks of the clinical encounter. Interviews were audio-recorded and transcribed. Parents/guardians who spoke languages other than English were interviewed with certified interpreters and English translations of their responses were transcribed.
We developed a codebook deductively based on resilience engineering concepts and inductively based on emerging data. The codes captured how clinicians and patients/families anticipate, monitor, respond, and learn throughout the diagnostic process and how robustness and graceful extensibility were demonstrated as they adapted to the various conditions under which diagnosis unfolded. Qualitative coding was performed by trained codersprimary and secondary coders were assigned to code field notes and interview transcripts generated from each patient. Discrepancies in coding were resolved by group consensus. Directed content analysis was used to characterize adaptations in diagnostic performance.
Results
We have conducted interviews and observations of 21patients as of September 19, 2024 (recruitment is ongoing). Thirteen (62%) patients were male, 8 (38%) were Black, 7 (33%) were White, and 5 (24%) were Hispanic. Sixteen patient families (76%) spoke English as their primary language, 2 (10%) spoke Spanish, while the remaining 3 each spoke Haitian Creole, Arabic, or Vietnamese. We identified the following key themes:
Anticipation. Based on their working diagnoses, clinicians anticipated possible disease courses and potential complications, which informed the nature and urgency of their next diagnostic actions. Clinicians also anticipated factors that may affect the diagnostic process (e.g., lab operating hours, home resources, patients’/families’ preferences) and tailored their diagnostic plans accordingly. Clinicians were able to anticipate potential problems more effectively if they had: 1) adequate medical knowledge, 2) substantial experience of the range of clinical presentations and outcomes, and 3) relevant clinical or situational information provided by other clinicians. Anticipation enabled advance planning of potential responses, increasing efficiency and timeliness of the diagnostic process. Using effective communication, patients/families were engaged by clinicians to become partners in monitoring the disease course and responding appropriately when necessary.
Monitoring. Monitoring patients’ response to treatment and their expected clinical course was integral to diagnosis, especially for those with diagnostic uncertainty. In an outpatient setting, monitoring required substantial resources (e.g., nurse staffing for patient follow-up). Monitoring patients required teamwork across the primary care team, subspecialty consultants, and patients/families. After receiving instructions from the clinician, most monitoring was performed by families at home and they were responsible for communicating concerns to the clinical team. Monitoring also allowed clinicians to learn the outcomes of their diagnostic decisions, providing feedback on their diagnostic performance.
Response. Information from the history, physical exam, test results, and clinical course informed how clinicians and patients/families responded, which may be in the form of additional testing, subspecialty consultation, and/or second opinion referrals. Responses were tailored to the individual patient’s circumstance and were a product of shared decision-making between the clinician and patients/families.
Learning. Clinicians learned to approach diagnosis in different ways through past experience, prior encounters with different patients/families, exposure to their colleagues’ varying diagnostic styles, and from available clinical decision support tools (e.g. institutional clinical pathways). Experience with diagnostic error prompted both individual clinician and organizational learning leading to changes in practice. Patients/families also learned from previous healthcare encounters, which shape the way they approach subsequent interactions.
Robustness. Individual clinicians contributed to robustness in diagnosis by confirming thoroughness in information gathering, ensuring that high risk patients received appropriate return instructions and follow-up, and reflecting on diagnostic decisions. Most clinicians considered their colleagues’ support as the single greatest source of robustness in outpatient diagnosis. Diagnostic excellence was institutionally supported by providing clinical teams with vital clinical decision support and communication tools (e.g., interpreter services), and by allowing adequate time per patient visit.
Graceful Extensibility. Clinical teams commonly extended themselves to ensure that the diagnostic process went smoothly, which included involving parents by phone during visits, seeing patients even after arriving late, and spending as much time as needed commensurate to the patient’s needs. Clinicians also noted that their colleagues generously extended their help for cases with significant diagnostic uncertainty.
Conclusions
In pediatric primary care, adaptations of clinicians, patient/families, and organizations are an integral part of the diagnostic process. These adaptations are dependent on the knowledge/expertise of clinicians, require significant organizational resources, involve shared decision-making with patients/families, and rely on patient/family engagement to implement effectively. Pediatric outpatient diagnosis is made more robust and extended to meet the needs of patients at both the individual clinician and organizational level. This challenges the idea that misdiagnosis occurs as a result of a single cognitive failure or one failed action of the clinician in an otherwise reliable system, with “one right way” to diagnose and linear relationships between process and outcome. It is the constant adaptations and interactions between clinicians, patients, families, and the wider organization that lead to diagnostic success or failure. This work represents an initial investigation of diagnostic performance variation and resilience in pediatric outpatient diagnosis. Further study is needed to understand specific adaptations that lead to improved diagnostic outcomes. Doing so will inform targeted interventions which build on the capacity of pediatric primary care systems to successfully adapt to the challenges of diagnosis.
Authors
Event Type
Oral Presentations
TimeMonday, March 312:15pm - 2:37pm EDT
LocationQueens Quay
Patient Safety and Research Initiatives (PS)