Kelly Michelle Smith
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Kelly Michelle Smith has a total of 195 co-authors. Here are the top co-authors by number of shared publications:
- Kristen Miller
MedStar Health Center for Diagnostic Systems Safety, Columbia, USA
8 shared publications - Traber Davis Giardina
Michael E. DeBakey VA Medical Center
6 shared publications - Helen Haskell
Mothers Against Medical Error, Columbia, SC, USA
5 shared publications - Heather Mary Arthur
Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada
4 shared publications - Mary A Hill
Institute of Health Policy Management and Evaluation, University of Toronto, Toronto, Ontario, Canada
4 shared publications - Alberta Tran
MedStar Health Research Institute Columbia Maryland USA
3 shared publications
Latest Publications
- Open Access
- 8 May 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 13, 1858637
Editorial: Patient-centered care: strengthening trust and communication in healthcare relationships
The Research Topic, patient-centered care: strengthening trust and communication in healthcare relationships, brings together twenty-three articles spanning diverse settings and populations, illustrating how trust and communication within patient-centered care are shaped across these levels, from individual experiences and interpersonal communication to organizational structures and broader societal conditions, and how improving health requires alignment across them.At the individual level, patient-centered care is shaped by patients' knowledge, beliefs, preferences, prior experiences, and capacity to engage in their care. 20 These studies highlight how these factors influence patients' information needs, decision-making preferences, and how patients interpret, respond to, and participate in healthcare. This work shows that trust and communication begin before any clinical interaction takes place, underscoring the need for communication approaches that are responsive to individual context, capacity.Hendrix et al. leveraged a co-design approach to develop patient education interventions aimed at improving medication safety, identifying knowledge gaps and information asymmetry as key barriers. They found that patients showed greater interest in partnership-focused educational content and that addressing information asymmetry may support more active patient engagement and shared decision-making. Fusiak et al. found that preferences for involvement in care are not fixed and may vary depending on how they are elicited, pointing to the importance of using thoughtful methods to understand what role patients want to play in decision-making. Extending this focus beyond patients alone, the review by Ofosu et al. captured the emotional, ethical, and practical complexity caregivers face when making high-stakes decisions under uncertainty, showing how limited support and power imbalances can shape participation and trust. Finally, Zhang et al. provided a window into the lived realities of managing chronic illness, illustrating how ongoing physical, emotional, and social burdens shape how individuals experience care and interact with healthcare providers.At the interpersonal level, patient-centered care is enacted through relationships between patients, clinicians, and caregivers, where communication serves as the primary vehicle for care delivery. These studies focus on the dynamics of these interactions, including how communication strategies and behaviors of both patients and clinicians, as well as relational cues, influence understanding and engagement. Together, they demonstrate that trust is built through interaction, not intention. Both verbal and nonverbal communication, as well as respect, responsiveness, and inclusion in decision-making (i.e. Shared Decision Making), shape whether patients feel heard, valued, and confident in their care, ultimately influencing satisfaction, adherence, and outcomes.In an opinion piece, Faytong-Haro examined how patients' symptoms and concerns may be dismissed or minimized across clinical contexts such as chronic, poorly understood conditions, mental health, and women's health, contributing to emotional distress, delayed diagnosis, and erosion of trust, particularly among marginalized populations. The author concludes that these patterns are driven by a combination of information asymmetry, bias, and systemic constraints, and emphasizes the need for improved communication, clinician awareness, and system-level changes to rebuild trust and support safer care. The review by Piątkowska et al. highlighted both the promise and challenges of implementing Shared Decision Making in the context of asthma and chronic respiratory disease management, emphasizing the need for approaches that are feasible and responsive to real-world clinical settings. Efforts to better capture and evaluate communication are reflected in the original research by Shao et al., which introduced a structured, patient-reported scale to assess doctorpatient communication quality, including how communication is experienced and its impact on engagement and health behaviors. At the community and organizational level, patient-centered care is shaped by the environments in which care is delivered, including care delivery models, access to services, and the broader social and cultural context. These studies highlight how care delivery structures, continuity, and community-based resources shape the context in which patient-clinician communication occurs. Collectively, this work illustrates that trust and communication are not solely the responsibility of individual clinicians but are enabled or constrained by system design. Access, coordination, and organizational culture shape the conditions under which meaningful communication can occur and ultimately determine whether patient-centered care can be consistently delivered.At the level of care delivery experience shaped by organizational context, Zhang et al. demonstrated that how patients experience and evaluate care strongly influences whether they engage with and consistently use primary care services. Similarly, in a brief research report, Moreno & Cornejo highlighted the importance of sustained relationships in care among older adults in primary care settings, linking continuity not only to patient experience but also to broader well-being, including mental health outcomes. Expanding beyond individual encounters, Mitnik et al. illustrated how community-based oncology centers can improve access to care by bringing services closer to patients' homes, while also highlighting the need for strong coordination between community and hospital teams to maintain patient confidence, ensure safety, and support continuity of care.At the organizational level, Alshahrani et al. demonstrated that workforce experience directly influences communication quality and patient trust, with supportive work environments enabling more effective and patient-centered...
- 8 May 2026
- Published by Frontiers Media SA in Frontiers in Medicine
- Vol. 13, 1858637
Open AccessEditorial: Patient-centered care: strengthening trust and communication in healthcare relationships
The Research Topic, patient-centered care: strengthening trust and communication in healthcare relationships, brings together twenty-three articles spanning diverse settings and populations, illustrating how trust and communication within patient-centered care are shaped across these levels, from individual experiences and interpersonal communication to organizational structures and broader societal conditions, and how improving health requires alignment across them.At the individual level, patient-centered care is shaped by patients' knowledge, beliefs, preferences, prior experiences, and capacity to engage in their care. 20 These studies highlight how these factors influence patients' information needs, decision-making preferences, and how patients interpret, respond to, and participate in healthcare. This work shows that trust and communication begin before any clinical interaction takes place, underscoring the need for communication approaches that are responsive to individual context, capacity.Hendrix et al. leveraged a co-design approach to develop patient education interventions aimed at improving medication safety, identifying knowledge gaps and information asymmetry as key barriers. They found that patients showed greater interest in partnership-focused educational content and that addressing information asymmetry may support more active patient engagement and shared decision-making. Fusiak et al. found that preferences for involvement in care are not fixed and may vary depending on how they are elicited, pointing to the importance of using thoughtful methods to understand what role patients want to play in decision-making. Extending this focus beyond patients alone, the review by Ofosu et al. captured the emotional, ethical, and practical complexity caregivers face when making high-stakes decisions under uncertainty, showing how limited support and power imbalances can shape participation and trust. Finally, Zhang et al. provided a window into the lived realities of managing chronic illness, illustrating how ongoing physical, emotional, and social burdens shape how individuals experience care and interact with healthcare providers.At the interpersonal level, patient-centered care is enacted through relationships between patients, clinicians, and caregivers, where communication serves as the primary vehicle for care delivery. These studies focus on the dynamics of these interactions, including how communication strategies and behaviors of both patients and clinicians, as well as relational cues, influence understanding and engagement. Together, they demonstrate that trust is built through interaction, not intention. Both verbal and nonverbal communication, as well as respect, responsiveness, and inclusion in decision-making (i.e. Shared Decision Making), shape whether patients feel heard, valued, and confident in their care, ultimately influencing satisfaction, adherence, and outcomes.In an opinion piece, Faytong-Haro examined how patients' symptoms and concerns may be dismissed or minimized across clinical contexts such as chronic, poorly understood conditions, mental health, and women's health, contributing to emotional distress, delayed diagnosis, and erosion of trust, particularly among marginalized populations. The author concludes that these patterns are driven by a combination of information asymmetry, bias, and systemic constraints, and emphasizes the need for improved communication, clinician awareness, and system-level changes to rebuild trust and support safer care. The review by Piątkowska et al. highlighted both the promise and challenges of implementing Shared Decision Making in the context of asthma and chronic respiratory disease management, emphasizing the need for approaches that are feasible and responsive to real-world clinical settings. Efforts to better capture and evaluate communication are reflected in the original research by Shao et al., which introduced a structured, patient-reported scale to assess doctorpatient communication quality, including how communication is experienced and its impact on engagement and health behaviors. At the community and organizational level, patient-centered care is shaped by the environments in which care is delivered, including care delivery models, access to services, and the broader social and cultural context. These studies highlight how care delivery structures, continuity, and community-based resources shape the context in which patient-clinician communication occurs. Collectively, this work illustrates that trust and communication are not solely the responsibility of individual clinicians but are enabled or constrained by system design. Access, coordination, and organizational culture shape the conditions under which meaningful communication can occur and ultimately determine whether patient-centered care can be consistently delivered.At the level of care delivery experience shaped by organizational context, Zhang et al. demonstrated that how patients experience and evaluate care strongly influences whether they engage with and consistently use primary care services. Similarly, in a brief research report, Moreno & Cornejo highlighted the importance of sustained relationships in care among older adults in primary care settings, linking continuity not only to patient experience but also to broader well-being, including mental health outcomes. Expanding beyond individual encounters, Mitnik et al. illustrated how community-based oncology centers can improve access to care by bringing services closer to patients' homes, while also highlighting the need for strong coordination between community and hospital teams to maintain patient confidence, ensure safety, and support continuity of care.At the organizational level, Alshahrani et al. demonstrated that workforce experience directly influences communication quality and patient trust, with supportive work environments enabling more effective and patient-centered...
- Open Access
- 27 April 2026
- Published by JMIR Publications Inc. in JMIR AI
Identifying Diagnostic Breakdowns in Patient Experience Survey Comments Using Natural Language Processing: An Evaluation Study (Preprint)
- 27 April 2026
- Published by JMIR Publications Inc. in JMIR AI
Open AccessIdentifying Diagnostic Breakdowns in Patient Experience Survey Comments Using Natural Language Processing: An Evaluation Study (Preprint)
- Open Access
- 23 December 2025
- Published by BMJ in BMJ Quality & Safety
- Vol. 35 (8) , 552-564
Scoping review of patient and family engagement interventions in diagnosis: a paradox of too much, yet so little
Introduction: Actively engaging patients is essential for diagnostic excellence and patient safety. Objectives: To (1) identify and synthesise interventions facilitating patient and family engagement (PFE) across the diagnostic process, and (2a) assess patient involvement and (2b) equity considerations in their design or implementation. Design: This scoping review followed Arksey and O’Malley’s framework and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Review) guidelines. An advisory panel guided the review. We searched Medline, Embase, CINAHL, PsycInfo and Northern Light for peer-reviewed literature and conducted grey literature searches using DuckDuckGo and targeted websites. Search terms focused on PFE and diagnostic error. Eligible interventions were published in English between January 1999 and July 2024 and supported PFE in at least one step of the National Academies of Sciences, Engineering, and Medicine (NASEM) diagnostic process. Narrative reviews, case studies and editorials were excluded. Interventions were mapped to the NASEM steps; data were extracted on patient involvement and equity. Results: Of the 11 630 studies screened, 250 were included, representing 260 interventions. Most (n=213; 85.2%) were from the grey literature, and patients were primary users (n=166; 63.8%). Interventions spanned all diagnostic process steps but were most common in treatment (n=122; 46.9%) and history taking (n=100; 38.5%), with few in referrals (n=10, 3.8%) and physical examinations (n=6, 2.3%). The evidence base was weak: grey literature interventions lacked high-quality studies, and among the 37 peer-reviewed studies, three were randomised controlled trials, each limited by small samples or high attrition. Only 63 interventions (24.2%) were designed with patients, and 48 (18.5%) incorporated equity. Conclusion: PFE interventions exist across the diagnostic process, but few target referrals and physical examinations. The evidence remains weak, and current interventions cannot be considered effective. Future research should prioritise equity, patient involvement and rigorous evaluation.
- 23 December 2025
- Published by BMJ in BMJ Quality & Safety
- Vol. 35 (8) , 552-564
Open AccessScoping review of patient and family engagement interventions in diagnosis: a paradox of too much, yet so little
Introduction: Actively engaging patients is essential for diagnostic excellence and patient safety. Objectives: To (1) identify and synthesise interventions facilitating patient and family engagement (PFE) across the diagnostic process, and (2a) assess patient involvement and (2b) equity considerations in their design or implementation. Design: This scoping review followed Arksey and O’Malley’s framework and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Review) guidelines. An advisory panel guided the review. We searched Medline, Embase, CINAHL, PsycInfo and Northern Light for peer-reviewed literature and conducted grey literature searches using DuckDuckGo and targeted websites. Search terms focused on PFE and diagnostic error. Eligible interventions were published in English between January 1999 and July 2024 and supported PFE in at least one step of the National Academies of Sciences, Engineering, and Medicine (NASEM) diagnostic process. Narrative reviews, case studies and editorials were excluded. Interventions were mapped to the NASEM steps; data were extracted on patient involvement and equity. Results: Of the 11 630 studies screened, 250 were included, representing 260 interventions. Most (n=213; 85.2%) were from the grey literature, and patients were primary users (n=166; 63.8%). Interventions spanned all diagnostic process steps but were most common in treatment (n=122; 46.9%) and history taking (n=100; 38.5%), with few in referrals (n=10, 3.8%) and physical examinations (n=6, 2.3%). The evidence base was weak: grey literature interventions lacked high-quality studies, and among the 37 peer-reviewed studies, three were randomised controlled trials, each limited by small samples or high attrition. Only 63 interventions (24.2%) were designed with patients, and 48 (18.5%) incorporated equity. Conclusion: PFE interventions exist across the diagnostic process, but few target referrals and physical examinations. The evidence remains weak, and current interventions cannot be considered effective. Future research should prioritise equity, patient involvement and rigorous evaluation.
- 27 November 2025
- Published by Elsevier in Australian Critical Care
- Vol. 39 (1) , 101488
Barriers to and facilitators for implementing a Care Partner program for adult patients with persistent critical illness
- 27 November 2025
- Published by Elsevier in Australian Critical Care
- Vol. 39 (1) , 101488
Barriers to and facilitators for implementing a Care Partner program for adult patients with persistent critical illness
- Top Cited PapersOpen Access
- 24 July 2025
- Published by JMIR Publications Inc. in Journal of Participatory Medicine
- Vol. 17, e69564
Patient Perspectives on Artificial Intelligence in Health Care: Focus Group Study for Diagnostic Communication and Tool Implementation
Background: Artificial intelligence (AI) is rapidly transforming health care, offering potential benefits in diagnosis, treatment, and workflow efficiency. However, limited research explores patient perspectives on AI, especially in its role in diagnosis and communication. This study examines patient perceptions of various AI applications, focusing on the diagnostic process and communication. Objective: This study aimed to examine patient perspectives on AI use in health care, particularly in diagnostic processes and communication, identifying key concerns, expectations, and opportunities to guide the development and implementation of AI tools. Methods: This study used a qualitative focus group methodology with co-design principles to explore patient and family member perspectives on AI in clinical practice. A single 2-hour session was conducted with 17 adult participants. The session included interactive activities and breakout sessions focused on five specific AI scenarios relevant to diagnosis and communication: (1) portal messaging, (2) radiology review, (3) digital scribe, (4) virtual human, and (5) decision support. The session was audio-recorded and transcribed, with facilitator notes and demographic questionnaires collected. Data were analyzed using inductive thematic analysis by 2 independent researchers (GF and JB), with discrepancies resolved via consensus. Results: Participants reported varying comfort levels with AI applications contingent on the level of patient interaction, with digital scribe (average 4.24, range 2-5) and radiology review (average 4.00, range 2-5) being the highest, and human (average 1.68, range 1-4) being the lowest. In total, five cross-cutting themes emerged: (1) validation (concerns about model reliability), (2) usability (impact on diagnostic processes), (3) transparency (expectations for disclosing AI usage), (4) opportunities (potential for AI to improve care), and (5) privacy (concerns about data security). Participants valued the co-design session and felt they had a significant say in the discussions. Conclusions: This study highlights the importance of incorporating patient perspectives in the design and implementation of AI tools in health care. Transparency, human oversight, clear communication, and data privacy are crucial for patient trust and acceptance of AI in diagnostic processes. These findings inform strategies for individual clinicians, health care organizations, and policy makers to ensure responsible and patient-centered AI deployment in health care.
- 24 July 2025
- Published by JMIR Publications Inc. in Journal of Participatory Medicine
- Vol. 17, e69564
Top Cited PapersOpen AccessPatient Perspectives on Artificial Intelligence in Health Care: Focus Group Study for Diagnostic Communication and Tool Implementation
Background: Artificial intelligence (AI) is rapidly transforming health care, offering potential benefits in diagnosis, treatment, and workflow efficiency. However, limited research explores patient perspectives on AI, especially in its role in diagnosis and communication. This study examines patient perceptions of various AI applications, focusing on the diagnostic process and communication. Objective: This study aimed to examine patient perspectives on AI use in health care, particularly in diagnostic processes and communication, identifying key concerns, expectations, and opportunities to guide the development and implementation of AI tools. Methods: This study used a qualitative focus group methodology with co-design principles to explore patient and family member perspectives on AI in clinical practice. A single 2-hour session was conducted with 17 adult participants. The session included interactive activities and breakout sessions focused on five specific AI scenarios relevant to diagnosis and communication: (1) portal messaging, (2) radiology review, (3) digital scribe, (4) virtual human, and (5) decision support. The session was audio-recorded and transcribed, with facilitator notes and demographic questionnaires collected. Data were analyzed using inductive thematic analysis by 2 independent researchers (GF and JB), with discrepancies resolved via consensus. Results: Participants reported varying comfort levels with AI applications contingent on the level of patient interaction, with digital scribe (average 4.24, range 2-5) and radiology review (average 4.00, range 2-5) being the highest, and human (average 1.68, range 1-4) being the lowest. In total, five cross-cutting themes emerged: (1) validation (concerns about model reliability), (2) usability (impact on diagnostic processes), (3) transparency (expectations for disclosing AI usage), (4) opportunities (potential for AI to improve care), and (5) privacy (concerns about data security). Participants valued the co-design session and felt they had a significant say in the discussions. Conclusions: This study highlights the importance of incorporating patient perspectives in the design and implementation of AI tools in health care. Transparency, human oversight, clear communication, and data privacy are crucial for patient trust and acceptance of AI in diagnostic processes. These findings inform strategies for individual clinicians, health care organizations, and policy makers to ensure responsible and patient-centered AI deployment in health care.
- Open Access
- 29 May 2025
- Published by JMIR Publications Inc. in JMIR Rehabilitation and Assistive Technologies
- Vol. 12, e70339
Use of ChatGPT for Urinary Symptom Management Among People With Spinal Cord Injury or Disease: Qualitative Study
Background: Individuals with spinal cord injury or disease (SCI/D) experience disproportionately high rates of recurrent urinary tract infections, which are often complicated by atypical symptoms and delayed diagnoses. Patient-centered tools, like the Urinary Symptom Questionnaires for Neurogenic Bladder (USQNB), have been developed to support symptom assessment yet remain underused. Generative artificial intelligence tools such as ChatGPT may offer a more usable approach to improving symptom management by providing real-time, tailored health information directly to patients. Objective: This study explores the role of ChatGPT (version 3.5) in supporting urinary symptom management for individuals with SCI/D, focusing on its perceived accuracy, usefulness, and impact on health care engagement and self-management practices. Methods: A total of 30 individuals with SCI/D were recruited through advocacy groups and health care networks. Using realistic, scenario-based testing derived from validated tools for symptom management with SCI/D, such as the USQNB, participants interacted with ChatGPT to seek advice for urinary symptoms. Follow-up interviews were conducted remotely to assess individuals’ experiences using ChatGPT for urinary symptom management. Data were analyzed using inductive content analysis, with themes refined iteratively through a consensus-based process. Results: People with SCI/D reported high levels of trust in ChatGPT’s recommendations, with all 30 participants agreeing or strongly agreeing with the advice provided. ChatGPT’s responses were perceived as clear and comparable to professional medical advice. Participants mentioned concerns about the lack of sources and integration with patient-specific data. ChatGPT influenced individuals’ decision-making by supporting symptom assessment and guiding participants on when to seek professional care or pursue self-management strategies. Conclusions: ChatGPT is a promising tool for symptom assessment and managing chronic conditions such as urinary symptoms in individuals with SCI/D. While ChatGPT enhances accessibility to health information, further research is needed to improve its transparency and integration with personalized health data to be a more usable tool in making informed health decisions.
- 29 May 2025
- Published by JMIR Publications Inc. in JMIR Rehabilitation and Assistive Technologies
- Vol. 12, e70339
Open AccessUse of ChatGPT for Urinary Symptom Management Among People With Spinal Cord Injury or Disease: Qualitative Study
Background: Individuals with spinal cord injury or disease (SCI/D) experience disproportionately high rates of recurrent urinary tract infections, which are often complicated by atypical symptoms and delayed diagnoses. Patient-centered tools, like the Urinary Symptom Questionnaires for Neurogenic Bladder (USQNB), have been developed to support symptom assessment yet remain underused. Generative artificial intelligence tools such as ChatGPT may offer a more usable approach to improving symptom management by providing real-time, tailored health information directly to patients. Objective: This study explores the role of ChatGPT (version 3.5) in supporting urinary symptom management for individuals with SCI/D, focusing on its perceived accuracy, usefulness, and impact on health care engagement and self-management practices. Methods: A total of 30 individuals with SCI/D were recruited through advocacy groups and health care networks. Using realistic, scenario-based testing derived from validated tools for symptom management with SCI/D, such as the USQNB, participants interacted with ChatGPT to seek advice for urinary symptoms. Follow-up interviews were conducted remotely to assess individuals’ experiences using ChatGPT for urinary symptom management. Data were analyzed using inductive content analysis, with themes refined iteratively through a consensus-based process. Results: People with SCI/D reported high levels of trust in ChatGPT’s recommendations, with all 30 participants agreeing or strongly agreeing with the advice provided. ChatGPT’s responses were perceived as clear and comparable to professional medical advice. Participants mentioned concerns about the lack of sources and integration with patient-specific data. ChatGPT influenced individuals’ decision-making by supporting symptom assessment and guiding participants on when to seek professional care or pursue self-management strategies. Conclusions: ChatGPT is a promising tool for symptom assessment and managing chronic conditions such as urinary symptoms in individuals with SCI/D. While ChatGPT enhances accessibility to health information, further research is needed to improve its transparency and integration with personalized health data to be a more usable tool in making informed health decisions.
- Open Access
- 22 January 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Exploring the Role of ChatGPT in Urinary Symptom Management for Individuals with Spinal Cord Injury/Disease: A Qualitative Study (Preprint)
BACKGROUND Individuals with spinal cord injury or disease (SCI/D) experience disproportionately high rates of recurrent urinary tract infections (UTIs), which are often complicated by atypical symptoms and delayed diagnoses. Patient-centered tools, like the Urinary Symptom Questionnaires for Neurogenic Bladder (USQNB), have been developed to support symptom assessment yet remain underutilized. Generative AI tools, such as ChatGPT, may offer a more usable approach to improving symptom management by providing real-time, tailored health information directly to patients. OBJECTIVE This study explores the role of ChatGPT (version 3.5) in supporting urinary symptom management for individuals with SCI/D, focusing on its perceived accuracy, usefulness, and impact on healthcare engagement and self-management practices. METHODS Thirty individuals with SCI/D were recruited through advocacy groups and healthcare networks. Using realistic, scenario-based testing derived from validated tools for symptom management with SCI/D, such as the USQNB, participants interacted with ChatGPT to seek advice for urinary symptoms. Follow-up interviews were conducted remotely to assess individuals’ experiences using ChatGPT for urinary symptom management. Data were analyzed using inductive content analysis, with themes refined iteratively through a consensus-based process. RESULTS People with SCI/D reported high levels of trust in ChatGPT’s recommendations, with all 30 participants agreeing or strongly agreeing with the advice provided. ChatGPT’s responses were perceived as clear and comparable to professional medical advice. Participants mentioned concerns about the lack of sources and integration with patient-specific data. ChatGPT influenced individuals’ decision-making by supporting symptom assessment and guiding participants on when to seek professional care or pursue self-management strategies. CONCLUSIONS ChatGPT is a promising tool for symptom assessment and managing chronic conditions, such as urinary symptoms in individuals with SCI/D. While ChatGPT enhances accessibility to health information, further research is needed to improve its transparency and integration with personalized health data to be a more usable tool in making informed health decisions.
- 22 January 2025
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessExploring the Role of ChatGPT in Urinary Symptom Management for Individuals with Spinal Cord Injury/Disease: A Qualitative Study (Preprint)
BACKGROUND Individuals with spinal cord injury or disease (SCI/D) experience disproportionately high rates of recurrent urinary tract infections (UTIs), which are often complicated by atypical symptoms and delayed diagnoses. Patient-centered tools, like the Urinary Symptom Questionnaires for Neurogenic Bladder (USQNB), have been developed to support symptom assessment yet remain underutilized. Generative AI tools, such as ChatGPT, may offer a more usable approach to improving symptom management by providing real-time, tailored health information directly to patients. OBJECTIVE This study explores the role of ChatGPT (version 3.5) in supporting urinary symptom management for individuals with SCI/D, focusing on its perceived accuracy, usefulness, and impact on healthcare engagement and self-management practices. METHODS Thirty individuals with SCI/D were recruited through advocacy groups and healthcare networks. Using realistic, scenario-based testing derived from validated tools for symptom management with SCI/D, such as the USQNB, participants interacted with ChatGPT to seek advice for urinary symptoms. Follow-up interviews were conducted remotely to assess individuals’ experiences using ChatGPT for urinary symptom management. Data were analyzed using inductive content analysis, with themes refined iteratively through a consensus-based process. RESULTS People with SCI/D reported high levels of trust in ChatGPT’s recommendations, with all 30 participants agreeing or strongly agreeing with the advice provided. ChatGPT’s responses were perceived as clear and comparable to professional medical advice. Participants mentioned concerns about the lack of sources and integration with patient-specific data. ChatGPT influenced individuals’ decision-making by supporting symptom assessment and guiding participants on when to seek professional care or pursue self-management strategies. CONCLUSIONS ChatGPT is a promising tool for symptom assessment and managing chronic conditions, such as urinary symptoms in individuals with SCI/D. While ChatGPT enhances accessibility to health information, further research is needed to improve its transparency and integration with personalized health data to be a more usable tool in making informed health decisions.
- Open Access
- 26 August 2024
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 26 (1) , e50935
Evaluation of a Natural Language Processing Approach to Identify Diagnostic Errors and Analysis of Safety Learning System Case Review Data: Retrospective Cohort Study
Background: Diagnostic errors are an underappreciated cause of preventable mortality in hospitals and pose a risk for severe patient harm and increase hospital length of stay. Objective: This study aims to explore the potential of machine learning and natural language processing techniques in improving diagnostic safety surveillance. We conducted a rigorous evaluation of the feasibility and potential to use electronic health records clinical notes and existing case review data. Methods: Safety Learning System case review data from 1 large health system composed of 10 hospitals in the mid-Atlantic region of the United States from February 2016 to September 2021 were analyzed. The case review outcome included opportunities for improvement including diagnostic opportunities for improvement. To supplement case review data, electronic health record clinical notes were extracted and analyzed. A simple logistic regression model along with 3 forms of logistic regression models (ie, Least Absolute Shrinkage and Selection Operator, Ridge, and Elastic Net) with regularization functions was trained on this data to compare classification performances in classifying patients who experienced diagnostic errors during hospitalization. Further, statistical tests were conducted to find significant differences between female and male patients who experienced diagnostic errors. Results: In total, 126 (7.4%) patients (of 1704) had been identified by case reviewers as having experienced at least 1 diagnostic error. Patients who had experienced diagnostic error were grouped by sex: 59 (7.1%) of the 830 women and 67 (7.7%) of the 874 men. Among the patients who experienced a diagnostic error, female patients were older (median 72, IQR 66-80 vs median 67, IQR 57-76; P=.02), had higher rates of being admitted through general or internal medicine (69.5% vs 47.8%; P=.01), lower rates of cardiovascular-related admitted diagnosis (11.9% vs 28.4%; P=.02), and lower rates of being admitted through neurology department (2.3% vs 13.4%; P=.04). The Ridge model achieved the highest area under the receiver operating characteristic curve (0.885), specificity (0.797), positive predictive value (PPV; 0.24), and F1-score (0.369) in classifying patients who were at higher risk of diagnostic errors among hospitalized patients. Conclusions: Our findings demonstrate that natural language processing can be a potential solution to more effectively identifying and selecting potential diagnostic error cases for review and therefore reducing the case review burden.
- 26 August 2024
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 26 (1) , e50935
Open AccessEvaluation of a Natural Language Processing Approach to Identify Diagnostic Errors and Analysis of Safety Learning System Case Review Data: Retrospective Cohort Study
Background: Diagnostic errors are an underappreciated cause of preventable mortality in hospitals and pose a risk for severe patient harm and increase hospital length of stay. Objective: This study aims to explore the potential of machine learning and natural language processing techniques in improving diagnostic safety surveillance. We conducted a rigorous evaluation of the feasibility and potential to use electronic health records clinical notes and existing case review data. Methods: Safety Learning System case review data from 1 large health system composed of 10 hospitals in the mid-Atlantic region of the United States from February 2016 to September 2021 were analyzed. The case review outcome included opportunities for improvement including diagnostic opportunities for improvement. To supplement case review data, electronic health record clinical notes were extracted and analyzed. A simple logistic regression model along with 3 forms of logistic regression models (ie, Least Absolute Shrinkage and Selection Operator, Ridge, and Elastic Net) with regularization functions was trained on this data to compare classification performances in classifying patients who experienced diagnostic errors during hospitalization. Further, statistical tests were conducted to find significant differences between female and male patients who experienced diagnostic errors. Results: In total, 126 (7.4%) patients (of 1704) had been identified by case reviewers as having experienced at least 1 diagnostic error. Patients who had experienced diagnostic error were grouped by sex: 59 (7.1%) of the 830 women and 67 (7.7%) of the 874 men. Among the patients who experienced a diagnostic error, female patients were older (median 72, IQR 66-80 vs median 67, IQR 57-76; P=.02), had higher rates of being admitted through general or internal medicine (69.5% vs 47.8%; P=.01), lower rates of cardiovascular-related admitted diagnosis (11.9% vs 28.4%; P=.02), and lower rates of being admitted through neurology department (2.3% vs 13.4%; P=.04). The Ridge model achieved the highest area under the receiver operating characteristic curve (0.885), specificity (0.797), positive predictive value (PPV; 0.24), and F1-score (0.369) in classifying patients who were at higher risk of diagnostic errors among hospitalized patients. Conclusions: Our findings demonstrate that natural language processing can be a potential solution to more effectively identifying and selecting potential diagnostic error cases for review and therefore reducing the case review burden.
- 12 July 2024
- Published by Emerald Publishing in Journal of Integrated Care
- Vol. 32 (3) , 303-312
Community-engaged co-design of a quality improvement capacity building program within an integrated health system in Ontario, Canada
Purpose: An integrated care system identified quality improvement (QI) capacity as a gap in advancing their integrated quality care priorities and improvement efforts. Here we describe the design and implementation of a QI capacity building program that aimed to (1) build QI capacity amongst diverse integrated care system members and (2) apply QI principles to advance integrated quality care priorities.Design/methodology/approach: The integrated care system leaders, including community members, partnered with the University of Toronto Centre for Quality Improvement and Patient Safety to co-design and deliver the QI capacity building program focused on improving cancer screening rates. An existing acute care capacity building program was adapted. Content included QI tools, data to identify and monitor QI priorities, equity considerations, and empowering participants as change agents.Findings: Participants were satisfied with the content and delivery of the program. Some described using QI tools and strategies in practice following the workshop. Challenges to using the tools included the current pressures facing primary care and the health system, resources, and data availability.Practical implications: This QI capacity building program was challenging but feasible. Clarifying the target audience, being attentive to co-design, acknowledging post-pandemic system challenges and proactively addressing variable knowledge and barriers to QI work in practice will inform future iterations of this program.Originality/value: While many examples of QI education programs exist, the majority target a single healthcare sector. We describe a novel QI capacity building model that bridges healthcare sectors and includes patient partners and community members as teachers and participants.
- 12 July 2024
- Published by Emerald Publishing in Journal of Integrated Care
- Vol. 32 (3) , 303-312
Community-engaged co-design of a quality improvement capacity building program within an integrated health system in Ontario, Canada
Purpose: An integrated care system identified quality improvement (QI) capacity as a gap in advancing their integrated quality care priorities and improvement efforts. Here we describe the design and implementation of a QI capacity building program that aimed to (1) build QI capacity amongst diverse integrated care system members and (2) apply QI principles to advance integrated quality care priorities.Design/methodology/approach: The integrated care system leaders, including community members, partnered with the University of Toronto Centre for Quality Improvement and Patient Safety to co-design and deliver the QI capacity building program focused on improving cancer screening rates. An existing acute care capacity building program was adapted. Content included QI tools, data to identify and monitor QI priorities, equity considerations, and empowering participants as change agents.Findings: Participants were satisfied with the content and delivery of the program. Some described using QI tools and strategies in practice following the workshop. Challenges to using the tools included the current pressures facing primary care and the health system, resources, and data availability.Practical implications: This QI capacity building program was challenging but feasible. Clarifying the target audience, being attentive to co-design, acknowledging post-pandemic system challenges and proactively addressing variable knowledge and barriers to QI work in practice will inform future iterations of this program.Originality/value: While many examples of QI education programs exist, the majority target a single healthcare sector. We describe a novel QI capacity building model that bridges healthcare sectors and includes patient partners and community members as teachers and participants.
- Open Access
- 28 December 2023
- Published by Ubiquity Press, Ltd. in International Journal of Integrated Care
- Vol. 23 (S1) , 166
Implementation and Evaluation of Innovation in Integrated Care: Strategies to Overcome the Challenge of Work as Imagined
Health system learning is an iterative process. The six phases of a learning health system (LHS) are problem identification, design, implementation, evaluation, modification, and dissemination. In theory, evaluation is supposed to inform design and implementation and in turn implementation process and modification as the direct result of the evaluation endeavour set the direction for future evaluation. In other words, evaluation and implementation go hand-in-hand. In practice, however, when designing an integrated model of care, more often than not in the fast-paced environment of the healthcare where lives are literally at stake, both evaluation and implementation planning are sacrificed. This happens for many reasons, from lack of resources to lack of expertise. One main reason that work as imagined is perceived as work as done is that those who run the program are smart and dedicated people who are well aware of the needs of their communities and clients, and have their best interest at heart. East Toronto Health Partners (ETHP) has invested in evaluations since 2018 to create a learning health system by embedding rapid cycles of evaluation to support learning, knowledge transfer, and decision making for scale and spread of their new models of care. One of the main findings of our evaluation work was that implementation planning and monitoring the implementation process raised the likelihood of a successful evaluation which in turn resulted in tangible improvement in the program. In our poster presentation, we will focus on both implementation and evaluation. We showcase the link between the logic model/evaluation plan and the implementation plan. We will cover the following questions: Why having an implementation plan is necessary to bring an idea for an intervention to a reality? How an implementation plan helps you to carry out your evaluation plan more successfully? What questions you should be thinking about when you are developing your implementation plan? We will share examples of best practices. The poster will also provide practical implementation templates that can be used in different setting.
- 28 December 2023
- Published by Ubiquity Press, Ltd. in International Journal of Integrated Care
- Vol. 23 (S1) , 166
Open AccessImplementation and Evaluation of Innovation in Integrated Care: Strategies to Overcome the Challenge of Work as Imagined
Health system learning is an iterative process. The six phases of a learning health system (LHS) are problem identification, design, implementation, evaluation, modification, and dissemination. In theory, evaluation is supposed to inform design and implementation and in turn implementation process and modification as the direct result of the evaluation endeavour set the direction for future evaluation. In other words, evaluation and implementation go hand-in-hand. In practice, however, when designing an integrated model of care, more often than not in the fast-paced environment of the healthcare where lives are literally at stake, both evaluation and implementation planning are sacrificed. This happens for many reasons, from lack of resources to lack of expertise. One main reason that work as imagined is perceived as work as done is that those who run the program are smart and dedicated people who are well aware of the needs of their communities and clients, and have their best interest at heart. East Toronto Health Partners (ETHP) has invested in evaluations since 2018 to create a learning health system by embedding rapid cycles of evaluation to support learning, knowledge transfer, and decision making for scale and spread of their new models of care. One of the main findings of our evaluation work was that implementation planning and monitoring the implementation process raised the likelihood of a successful evaluation which in turn resulted in tangible improvement in the program. In our poster presentation, we will focus on both implementation and evaluation. We showcase the link between the logic model/evaluation plan and the implementation plan. We will cover the following questions: Why having an implementation plan is necessary to bring an idea for an intervention to a reality? How an implementation plan helps you to carry out your evaluation plan more successfully? What questions you should be thinking about when you are developing your implementation plan? We will share examples of best practices. The poster will also provide practical implementation templates that can be used in different setting.