RESUMO
INTRODUCTION: The efficiency of multidisciplinary teams (MDTs) in cancer care hinges on facilitating clinicians' cognitive processes as they navigate complex and uncertain judgements during treatment planning. When systems and workflows are not designed to adequately support human judgement and decision-making, even experts are prone to fallible reasoning due to cognitive biases. Incomplete integration of information or biased interpretations of patient data can lead to clinical errors and delays in the implementation of treatment recommendations. Though their impact is intuitively recognised, there is currently a paucity of empirical work on cognitive biases in MDT decision-making. Our study aims to explicate the impact of such biases on treatment planning and establish a foundation for targeted investigations and interventions to mitigate their negative effects. METHODS AND ANALYSIS: This is a qualitative, observational study. We employ cognitive ethnography, informed by the Distributed Cognition for Teamwork framework to assess and evaluate MDT decision-making processes. The study involves in-person and virtual field observations of hepatopancreaticobiliary and upper gastrointestinal MDTs and interviews with their members over several months. The data generated will be analysed in a hybrid inductive/deductive fashion to develop a comprehensive map of potential cognitive biases in MDT decision processes identifying antecedents and risk factors of suboptimal treatment planning processes. Further, we will identify components of the MDT environment that can be redesigned to support decision-making via development of an MDT workspace evaluation tool. ETHICS AND DISSEMINATION: This project has received management and ethical approvals from NHS Lothian Research and Development (2023/0245) and the University of Edinburgh Medical School ethical review committee (23-EMREC-049). Findings will be shared with participating MDTs and disseminated via a PhD thesis, international conference presentations and relevant scientific journals.
Assuntos
Antropologia Cultural , Tomada de Decisão Clínica , Cognição , Neoplasias , Equipe de Assistência ao Paciente , Humanos , Escócia , Neoplasias/terapia , Pesquisa Qualitativa , Projetos de Pesquisa , Estudos Observacionais como Assunto , Tomada de Decisões , ViésRESUMO
Objective: Hospital at Home (HaH) programs currently lack decision support tools to help efficiently navigate the complex decision-making process surrounding HaH as a care option. We assessed user needs and perspectives to guide early prototyping and co-creation of 4PACS (Partnering Patients and Providers for Personalized Acute Care Selection), a decision support app to help patients make an informed decision when presented with discrete hospitalization options. Methods: From December 2021 to January 2022, we conducted semi-structured interviews via telephone with patients and caregivers recruited from Atrium Health's HaH program and physicians and a nurse with experience referring patients to HaH. Interviews were evaluated using thematic analysis. The findings were synthesized to create illustrative user descriptions to aid 4PACS development. Results: In total, 12 stakeholders participated (3 patients, 2 caregivers, 7 providers [physicians/nurse]). We identified 4 primary themes: attitudes about HaH; 4PACS app content and information needs; barriers to 4PACS implementation; and facilitators to 4PACS implementation. We characterized 3 user descriptions (one per stakeholder group) to support 4PACS design decisions. User needs included patient selection criteria, clear program details, and descriptions of HaH components to inform care expectations. Implementation barriers included conflict between app recommendations and clinical judgement, inability to adequately represent patient-risk profile, and provider burden. Implementation facilitators included ease of use, auto-populating features, and appropriate health literacy. Conclusions: The findings indicate important information gaps and user needs to help inform 4PACS design and barriers and facilitators to implementing 4PACS in the decision-making process of choosing between hospital-level care options.
RESUMO
Introduction: Language that assumes gender and sex are binary and aligned is pervasive in medicine and is often used when teaching on physiology and pathology. Information presented through this lens oversimplifies disease mechanisms and poorly addresses the health of gender and sexually diverse (GSD) individuals. We developed a training session to help faculty reference gender and sex in a manner that would be accurate and inclusive of GSD health. Methods: The 1-hour session for undergraduate and graduate medical educators highlighted cisgender and binary biases in medical teachings and introduced a getting-to-the-root mindset that prioritized teaching the processes underlying differences in disease profiles among gender and sex subpopulations. The training consisted of 30 minutes of didactic teaching and 20 minutes of small-group discussion. Medical education faculty attended and self-reported knowledge and awareness before and after the training. Results were compared using paired t tests. Expenses included fees for consultation and catering. Results: Forty faculty participated (pretraining survey n = 36, posttraining survey n = 21). After the training, there was a significant increase in self-reported awareness of the difference between gender and sex (p = .002), perceived relevance of gender to teachings (p = .04), and readiness to discuss physiological drivers of sex-linked disease (p = .005). Discussion: Participants reported increased understanding and consideration of gender and sex in medical education; feedback emphasized a desire for continued guidance. This easily adaptable session can provide an introduction to a series of medical teachings on gender and sex.
Assuntos
Docentes de Medicina , Humanos , Inquéritos e Questionários , Masculino , Feminino , Educação Médica/métodos , Identidade de Gênero , Adulto , Minorias Sexuais e de GêneroRESUMO
The decision to euthanize a compromised pig can be challenging for pig farmers and veterinarians. To understand more about the challenges in Germany, a cross-sectional online survey was conducted. Based on a hybrid design, the responses of 39 veterinarians and 62 pig farmers were analyzed to generate a list of common clinical signs associated with the euthanasia of sows, fatteners, and piglets. Moreover, a set of influences on the farm, due to economic and personal considerations, were found to shape the decision-making process. The two most salient reasons outlined for the delay of timely euthanasia were uncertainty and misinterpretation of the chance for healing. The lack of valid clinical signs or a sound justification was most frequently mentioned as a challenge to the general decision-making process. In summation, this study highlights the need to generate a valid taxonomy for clinical signs that includes their development in a compromised pig over time. Future studies should elaborate on the justification of euthanasia decisions to facilitate the resolution of ethical dilemmas among the involved pig farmers and veterinarians. Lastly, the results suggest that clinical reasoning and consultation skills should be included when decision-making behavior is to be trained.
RESUMO
Background: the aim of this study is to understand the diagnostic process undertaken by psychiatrists and psychologists regarding adjustment disorder (AD) in their clinical practice and how they differentiate it from major depressive episode (MDE).Methods: A hermeneutic study using grounded theory techniques was carried out. Semi-structured interviews were conducted with twelve psychiatrists and eight psychologists in Colombia, and transcribed verbatim. Initial line-by-line coding was performed, followed by focused and axial coding to construct categories explaining the professionals' reasoning process.Results: The clinical reasoning of professionals regarding AD was understood through four major categories. (1) Difficulty in addressing the experience of stressful events, as there is a risk of pathologizing and medicalizing them. (2) Mental health diagnoses are necessary but not apodictic. (3) The diagnostic category of AD allows for the description of a fluctuating depressive and anxious syndrome occurring in reaction to a stressful event, whose abnormality criteria are based on intersubjective knowledge of the patient's life history and consequential reasoning regarding the need for professional support. (4) The AD label could potentially protect against overdiagnosis of MDE and overuse of antidepressants. Many clinicians in their practice thus subordinate the diagnosis of MDE to ensuring it is not AD, contrary to what is outlined in diagnostic manuals.Conclusion: This study allowed us to understand the clinical reasoning of psychiatrists and psychologists about AD as a diagnosis that inherently indicates the need to work on coping and intervene in the stressor and should be considered as a diagnostic possibility in the same hierarchy as MDE in reactive syndromes, rather than a residual category.
Clinicians use consequential and intersubjective reasoning to diagnose Adjustment Disorder (AD).Systemic pressures lead to overdiagnosis of Major Depressive Episode (MDE) and excessive antidepressant use.AD should be recognized as a valid non-residual diagnostic category.
Assuntos
Transtornos de Adaptação , Raciocínio Clínico , Teoria Fundamentada , Psiquiatria , Humanos , Feminino , Transtornos de Adaptação/diagnóstico , Transtornos de Adaptação/psicologia , Masculino , Adulto , Transtorno Depressivo Maior/diagnóstico , Psicologia , Colômbia , Pessoa de Meia-Idade , Pesquisa Qualitativa , Entrevistas como Assunto , Diagnóstico Diferencial , PsiquiatrasRESUMO
AIMS: To investigate the relationships among communication competence, professional autonomy and clinical reasoning and to identify the factors that influence clinical reasoning competence in oncology nurses. DESIGN: Cross-sectional descriptive design. METHODS: Participants included 147 oncology nurses with more than a year of clinical experience in cancer wards. The Global Interpersonal Communication Competence Scale, Schutzenhofer Professional Autonomy Scale and Nurses Clinical Reasoning Scale (NCRS) were used to collect data. Data were analysed using descriptive statistics, independent t-test, one-way ANOVA, Pearson correlation analysis and hierarchical multiple regression analysis. RESULTS: Communication competence (r = 0.59) and professional autonomy (r = 0.46) showed significant positive relationships with clinical reasoning competence. Clinical experience, communication competence, age and professional autonomy were statistically significant predictors and explained 48.6% of clinical reasoning competence. CONCLUSIONS: The clinical reasoning competence of oncology nurses increases proportionally with their communication competence and professional autonomy. Therefore, oncology nurses must reinforce their communication competence and professional autonomy to enhance their clinical reasoning competence. IMPLICATIONS FOR THE PROFESSION AND/OR PATIENT CARE: The reinforcement of communication competence and professional autonomy is necessary for oncology nurses to enhance their clinical reasoning competence. In order to improve nurses' communication competence, practical-focused communication education programmes must be designed and deployed systematically and periodically. In addition, to increase nurses' professional autonomy, it is necessary to expand their clinical experiences through the regular rotation of working units and to make institutional efforts to retain experienced nurses. REPORTING METHOD: We have adhered to STROBE checklist. PATIENT OR PUBLIC CONTRIBUTION: Participants in the study were recruited online. They were informed of the study's purpose, method and usability and the survey could only be conducted if they consented to participate voluntarily.
Assuntos
Competência Clínica , Comunicação , Enfermagem Oncológica , Autonomia Profissional , Humanos , Estudos Transversais , Feminino , Masculino , Adulto , Enfermagem Oncológica/educação , Competência Clínica/normas , Inquéritos e Questionários , Raciocínio Clínico , Pessoa de Meia-Idade , Enfermeiras e Enfermeiros/psicologia , Competência Profissional/normasRESUMO
INTRODUCTION: Standardized patient (SP) encounters allow medical students to practice physical examination skills and clinical reasoning. SP cases are used for learning and assessment, but recorded encounters can also be valuable curriculum evaluation tools. We aimed to review SP encounters to improve abdominal examination skills and the broader physical examination curriculum. METHODS: We reviewed recorded SP encounters of third-year medical students on surgery clerkship rotation. Students examined a cisgender woman presenting with acute right lower abdominal pain. We observed abdominal examinations to determine which maneuvers were attempted and completed correctly. We then used these outcomes to develop targeted clerkship training for the subsequent student cohort. Our intervention targeted abdominal examination gaps by explaining how to integrate abdominal examination findings with a focused history for surgical patients. We evaluated the intervention's impact on abdominal examination skills with third-year medical students in comparison (2021-2022, n = 119) and intervention (2022-2023, n = 132) groups. RESULTS: In both the comparison and intervention groups, nearly all students attempted at least 1 general examination maneuver like auscultation, palpation, percussion, or rebound tenderness. Only 40% of students in the comparison group attempted an advanced maneuver like the Rovsing, Psoas, or Obturator sign. After the intervention, 75% of students in the intervention group attempted an advanced maneuver (χ2(1, 251) = 31.0, p < .001). Cohorts did not gain skills over time through the clerkship. Rebound tenderness was frequently assessed incorrectly by students in both groups, with many avoiding the right lower quadrant entirely. CONCLUSIONS: This project highlights how medical students struggle to utilize abdominal examination maneuvers and integrate findings. The results also showed that students did not consistently learn advanced examination skills either before or during clerkship rotation, which may be commonly assumed by clinical faculty. Finally, this work demonstrates how SP encounters can be used to evaluate and improve surgical education curriculum.
RESUMO
BACKGROUND: Creative activities as intervention (CaI) in mental health promotes recovery and is an important part of occupational therapy practice. Yet few studies have explored occupational therapists' reasoning about using CaI to trace tacit knowledge. AIMS/OBJECTIVES: The aim of this study was to explore occupational therapists' reasoning on rationale and motives in co-creating a model (the CreActivity model) for CaI as a resource in mental health rehabilitation. MATERIAL AND METHODS: Co-operative inquiry and narrative theory provided the methodology for exploring eight occupational therapists' narrative reasoning in the process of co-creating a model. Data were generated via ethnographic methods from seven co-operative inquiry group meetings and analysed through narrative analysis. RESULTS: The occupational therapists' reasoning in using creative activities comprised building relationships and opportunity for activity. Motivating and engagement by finding 'cracks' and stories bridging action with the past and future using creative activities were used to create reflection, development, and empowerment. CONCLUSIONS AND SIGNIFICANCE: Identifying the narrative layers of occupational therapists' reasoning on achieving core dimensions of occupational therapy may enable and support occupational therapists in the use of creative activities as intervention.
Assuntos
Criatividade , Narração , Terapeutas Ocupacionais , Terapia Ocupacional , Humanos , Terapeutas Ocupacionais/psicologia , Terapia Ocupacional/métodos , Motivação , Feminino , Masculino , Adulto , Pesquisa Qualitativa , Pessoa de Meia-IdadeRESUMO
The International Classification of Function, Disability, and Health (ICF) is known to be a valuable classification model in interprofessional neurorehabilitation, as it can lead to more patient-centered and self-determined treatment. To acquire the competencies implementing the ICF in the field of neurorehabilitation, it is important to anchor essential skills in the basic training of healthcare professionals. The Institute of Physiotherapy at FH JOANNEUM in Graz/Austria developed a concept to help students learn the necessary skills for implementing the ICF in a structured way. In the area of neurorehabilitation, we linked the ICF model with the Clinical Reasoning Model (CR). Competences are acquired over six semesters. Besides the general topics relating to the ICF (such as history, intention, and language) and CR that are taught in the first year, we focus in the later semesters explicitly on transferring these skills to neurorehabilitation. Therefore, we use interprofessional group work and problem-based courses as essential didactic elements for this transfer of skills. This article aims to show how the ICF could be implemented in Bachelor's degree programs for physiotherapy as well as in other healthcare programs. The authors' experiences are described and some best practice examples when working with the ICF in this field are given.
RESUMO
Background: The script concordance test (SCT) has been shown to be an effective tool to assess the clinical reasoning skills of nursing students. Various nursing studies have demonstrated the construct validity of this test. However, studies on the barriers that may impede construct validity during the development process are limited. Objective: This evaluation describes the barriers to the development of SCT for Bachelor's nursing students and the lessons learned regarding construct validity. Methods: We conducted a descriptive evaluation of the SCT development and a validation process was performed. The evaluation was based on written comments during the assessment (N = 327), a Student's Perspective Questionnaire (N = 100), and student feedback during three live review sessions (N = 27). Results: Despite consideration of the guidelines during SCT development, we encountered three main barriers that may impede construct validity. We undertook the necessary efforts to recruit an appropriate expert panel. We overestimated the experts' and students' understanding of the SCT methodology. Additionally, four potential causes of invalid item construction were identified. These possible causes were 'questionable intervention, hypothesis, or investigation', 'blurred data in new information', 'regression to the middle', and 'misinterpretation of the midpoint'. Conclusion: The three lessons learned are as follows: 1) The recruitment of an appropriate expert panel must not be underestimated. Besides clinical expertise, experts need training in SCT methodology, including awareness of possible pitfalls; 2) SCT training is a prerequisite for SCT as an assessment; and 3) student feedback may offer a deeper understanding of potential hidden script errors and causes for misinterpretation of SCT. Further studies are necessary to identify additional causes which may impede the construct validity of SCT in nursing education.
RESUMO
The decision-making process of experts in forensic psychiatric investigations (FPI) is complex and reasoning regarding psychiatric diagnosis and severe mental disorder (SMD, the judicial concept central to legal exemption in Swedish law) has severe ramifications. Nevertheless, the qualitative aspects of FPI experts' decision-making process have seldom been studied systematically. METHOD: The participants (N = 41) were FPI experts: forensic psychiatrists (n = 15), forensic psychologists (n = 15) and forensic social workers (n = 11). Using three case vignettes and qualitative content analysis, it was explored how case-specific characteristics could affect which hypotheses FPI experts generated regarding a) psychiatric diagnosis and b) severe mental disorder and c) which information sources they required. Each case vignette described a diagnostically ambiguous case but indicated emphasis on: psychotic symptoms (case 1); personality disorder symptoms (case 2) and neurodevelopmental disorder symptoms (case 3). RESULTS: Experts reasoned in a similar manner regarding generating hypotheses and required information, but also in a case-adapted manner. Experts considered various diagnostic alternatives, and some (e.g. psychosis) were mentioned for all three cases. Other diagnoses were only suggested as hypotheses in certain cases (e.g. case 3: intellectual disability). DISCUSSION: In Sweden, a core basis for SMD is psychotic-like functioning, and psychosis was suggested as a hypothesis for all three cases. Experts reasoned in similar ways regarding SMD in all cases, considering various perspectives for and against SMD. Some case-specific arguments for and against SMD adapted to the psychopathological circumstances were found. These could be related to aspects of the SMD concept that become important to ascertain when the type of psychopathology indicated in the case vignette was present; for example, ascertaining reality monitoring for a person with potential delusions of being followed by a criminal gang requires investigation of criminal history and related conflicts. Taken together, FPI-experts considered a broad range of psychiatric diagnoses in various cases. Their reasoning regarding SMD was both based on general and case-specific (or psychopathology-specific) factors.
RESUMO
This article examines the integration of a Clinical Nurse Reasoning (CNR) model into advanced nursing practice training to develop the clinical reasoning skills of advanced practice nurses. The article explores the contribution of Callista Roy's conceptual model of adaptation and presents a global analysis of a complex care situation that justifies the introduction of an CNR model paired with nursing knowledge to ensure comprehensive, quality nursing care.
Assuntos
Prática Avançada de Enfermagem , Raciocínio Clínico , Teoria de Enfermagem , Estudantes de Enfermagem , Humanos , Prática Avançada de Enfermagem/educação , Competência Clínica/normas , Modelos de EnfermagemRESUMO
Virtual patients (VPs) have long been used to teach and assess clinical reasoning. VPs can be programmed to simulate authentic patient-clinician interactions and to reflect a variety of contextual permutations. However, their use has historically been limited by the high cost and logistical challenges of large-scale implementation. We describe a novel globally-accessible approach to develop low-cost VPs at scale using artificial intelligence (AI) large language models (LLMs). We leveraged OpenAI Generative Pretrained Transformer (GPT) to create and implement two interactive VPs, and created permutations that differed in contextual features. We used systematic prompt engineering to refine a prompt instructing ChatGPT to emulate the patient for a given case scenario, and then provide feedback on clinician performance. We implemented the prompts using GPT-3.5-turbo and GPT-4.0, and created a simple text-only interface using the OpenAI API. GPT-4.0 was far superior. We also conducted limited testing using another LLM (Anthropic Claude), with promising results. We provide the final prompt, case scenarios, and Python code. LLM-VPs represent a 'disruptive innovation' - an innovation that is unmistakably inferior to existing products but substantially more accessible (due to low cost, global reach, or ease of implementation) and thereby able to reach a previously underserved market. LLM-VPs will lay the foundation for global democratization via low-cost-low-risk scalable development of educational and clinical simulations. These powerful tools could revolutionize the teaching, assessment, and research of management reasoning, shared decision-making, and AI evaluation (e.g. 'software as a medical device' evaluations).
RESUMO
BACKGROUND: Bleeding following cardiac surgery is common and serious, yet a gap persists in understanding how experienced intensive care nurses identify and respond to such complications. AIM: To describe the clinical decision-making of experienced intensive care unit nurses in addressing bleeding after cardiac surgery. STUDY DESIGN: This qualitative study adopted the Recognition-Primed Decision Model as its theoretical framework. Thirty-nine experienced nurses from four adult intensive care units participated in semi-structured interviews based on the critical decision method. The interviews explored their clinical judgements and decisions in bleeding situations, and data were analysed through dimensional analysis, an alternative to grounded theory. RESULTS: Participants maintained consistent vigilance towards post-cardiac surgery bleeding, recognizing it through a haemorrhagic dimension associated with blood loss and chest drainage and a hypovolemic dimension focusing on the repercussions of reduced blood volume. These dimensions organized their understanding of bleeding types (i.e., normal, medical, surgical, tamponade) and necessary actions. Their decision-making encompassed monitoring bleeding, identifying the cause, stopping the bleeding, stabilizing haemodynamic and supporting the patient and family. Participants also adapted their actions to specific circumstances, including local practices, professional autonomy, interprofessional dynamics and resource availability. CONCLUSIONS: Nurses' decision-making was shaped by their personal attributes, the patient's condition and contextual circumstances, underscoring their expertise and pivotal role in anticipating actions and adapting to diverse conditions. The concept of actionability emerged as the central dimension explaining their decision-making, defined as the capability to implement actions towards specific goals within the possibilities and constraints of a situation. RELEVANCE TO CLINICAL PRACTICE: This study underscores the need for continual updates to care protocols to align with current evidence and for quality improvement initiatives to close existing practice gaps. Exploring the concept of actionability further, developing adaptability-focused educational programmes, and understanding decision-making intricacies are crucial for informing nursing education and decision-support systems.
RESUMO
Background There has been an explosion of commentary and discussion about the ethics and utility of using artificial intelligence in medicine, and its practical use in medical education is still being debated. Through qualitative research methods, this study aims to highlight the advantages and pitfalls of using ChatGPT in the development of clinical reasoning cases for medical student education. Methods Five highly experienced faculty in medical education were provided instructions to create unique clinical reasoning cases for three different chief concerns using ChatGPT 3.0. Faculty were then asked to reflect on and review the created cases. Finally, a focus group was conducted to further analyze and describe their experiences with the new technology. Results Overall, faculty found the use of ChatGPT in the development of clinical reasoning cases easy to use but difficult to get to certain objectives and largely incapable of being creative enough to create complexity for student use without heavy editing. The created cases did provide a helpful starting point and were extremely efficient; however, faculty did experience some medical inaccuracies and fact fabrication. Conclusion There is value to using ChatGPT to develop curricular content, especially for clinical reasoning cases, but it needs to be comprehensively reviewed and verified. To efficiently and effectively utilize the tool, educators will need to develop a framework that can be easily translatable into simple prompts that ChatGPT can understand. Future work will need to strongly consider the risks of recirculating biases and misinformation.
RESUMO
BACKGROUND AND PURPOSE: Clinical decision-making (CDM) is crucial in pharmacy practice, necessitating effective teaching in undergraduate and postgraduate pharmacy education. This study aims to explore undergraduates and postgraduates' perceptions of how a new teaching model supports their CDM when addressing patient cases. EDUCATIONAL ACTIVITY AND SETTING: Implemented in a full-day CDM course for pharmacy students and a half-day course for pharmacists in the Netherlands, the model, accompanied by a learning guide, facilitated CDM in patient cases. Eight courses were conducted between September 2022 to June 2023, followed by an online survey measuring participants' agreement on how the model supported their CDM, using a 5-point Likert scale. Additionally, three open-ended questions were included to elicit learning outcomes and self-development opportunities. FINDINGS: Of 175 invited participants, 159 (91%) completed the survey. Most agreed the teaching model supported their CDM, particularly in considering the patient's healthcare needs and context (96%), and exploring all available options (96%). Participants found the model provided a clear structure (97%), and fostered critical thinking (93%). The most frequently mentioned learning outcomes and self-development opportunities included collecting sufficient relevant information, maintaining a broad perspective, and decelerating the process to avoid premature closure. SUMMARY: Participants agreed that the teaching model helped them to make clinical decisions. Both undergraduate and postgraduate pharmacy education could possibly benefit from the teaching model's implementation in supporting pharmacy students and pharmacists conducting CDM in pharmacy practice.
RESUMO
OBJECTIVES: To investigate approaches of reasoning with large language models (LLMs) and to propose a new prompting approach, ensemble reasoning, to improve medical question answering performance with refined reasoning and reduced inconsistency. MATERIALS AND METHODS: We used multiple choice questions from the USMLE Sample Exam question files on 2 closed-source commercial and 1 open-source clinical LLM to evaluate our proposed approach ensemble reasoning. RESULTS: On GPT-3.5 turbo and Med42-70B, our proposed ensemble reasoning approach outperformed zero-shot chain-of-thought with self-consistency on Steps 1, 2, and 3 questions (+3.44%, +4.00%, and +2.54%) and (2.3%, 5.00%, and 4.15%), respectively. With GPT-4 turbo, there were mixed results with ensemble reasoning again outperforming zero-shot chain-of-thought with self-consistency on Step 1 questions (+1.15%). In all cases, the results demonstrated improved consistency of responses with our approach. A qualitative analysis of the reasoning from the model demonstrated that the ensemble reasoning approach produces correct and helpful reasoning. CONCLUSION: The proposed iterative ensemble reasoning has the potential to improve the performance of LLMs in medical question answering tasks, particularly with the less powerful LLMs like GPT-3.5 turbo and Med42-70B, which may suggest that this is a promising approach for LLMs with lower capabilities. Additionally, the findings show that our approach helps to refine the reasoning generated by the LLM and thereby improve consistency even with the more powerful GPT-4 turbo. We also identify the potential and need for human-artificial intelligence teaming to improve the reasoning beyond the limits of the model.
Assuntos
Processamento de Linguagem Natural , Humanos , Avaliação Educacional/métodosRESUMO
OBJECTIVE: The recent surge in large language models (LLMs) across various fields has yet to be fully realized in traditional Chinese medicine (TCM). This study aims to bridge this gap by developing a large language model tailored to TCM knowledge, enhancing its performance and accuracy in clinical reasoning tasks such as diagnosis, treatment, and prescription recommendations. MATERIALS AND METHODS: This study harnessed a wide array of TCM data resources, including TCM ancient books, textbooks, and clinical data, to create 3 key datasets: the TCM Pre-trained Dataset, the Traditional Chinese Patent Medicine (TCPM) Question Answering Dataset, and the Spleen and Stomach Herbal Prescription Recommendation Dataset. These datasets underpinned the development of the Lingdan Pre-trained LLM and 2 specialized models: the Lingdan-TCPM-Chat Model, which uses a Chain-of-Thought process for symptom analysis and TCPM recommendation, and a Lingdan Prescription Recommendation model (Lingdan-PR) that proposes herbal prescriptions based on electronic medical records. RESULTS: The Lingdan-TCPM-Chat and the Lingdan-PR Model, fine-tuned on the Lingdan Pre-trained LLM, demonstrated state-of-the art performances for the tasks of TCM clinical knowledge answering and herbal prescription recommendation. Notably, Lingdan-PR outperformed all state-of-the-art baseline models, achieving an improvement of 18.39% in the Top@20 F1-score compared with the best baseline. CONCLUSION: This study marks a pivotal step in merging advanced LLMs with TCM, showcasing the potential of artificial intelligence to help improve clinical decision-making of medical diagnostics and treatment strategies. The success of the Lingdan Pre-trained LLM and its derivative models, Lingdan-TCPM-Chat and Lingdan-PR, not only revolutionizes TCM practices but also opens new avenues for the application of artificial intelligence in other specialized medical fields. Our project is available at https://github.com/TCMAI-BJTU/LingdanLLM.
Assuntos
Raciocínio Clínico , Medicina Tradicional Chinesa , Humanos , Medicamentos de Ervas Chinesas/uso terapêutico , Registros Eletrônicos de Saúde , Processamento de Linguagem Natural , Conjuntos de Dados como AssuntoRESUMO
PURPOSE: The objective of our IAEA-coordinated international study was to assess CT practices and radiation doses from multiple hospitals across several African countries. METHODS: The study included 13 hospitals from Africa which contributed information on minimum of 20 consecutive patients who underwent head, chest, and/or abdomen-pelvis CT. Prior to the data recording step, all hospitals had a mandatory one-hour training on the best practices in recording the relevant data elements. The recorded data elements included patient age, weight, protocol name, scanner information, acquisition parameters, and radiation dose descriptors including phase-specific CT dose index volume (CTDIvol in mGy) and dose length product (DLP in mGy.cm). We estimated the median and interquartile range of body-region specific CTDIvol and DLP and compared data across sites and countries using the Kruskal-Wallis H Test for non-normal distribution, analysis of variance. RESULTS: A total of 1061 patients (mean age 50 ± 19 years) were included in the study. 16 % of CT exams had no stated clinical indications for CT examinations of the head (32/343, 9 %), chest (50/281, 18 %), abdomen-pelvis (67/243, 28 %), and/or chest-abdomen-pelvis CT (24/194, 12 %). Most hospitals used multiphase CT protocols for abdomen-pelvis (9/11 hospitals) and chest CT (10/12 hospitals), regardless of clinical indications. Total median DLP values for head (953 mGy.cm), chest (405 mGy.cm), and abdomen-pelvis (1195 mGy.cm) CT were above the UK, German, and American College of Radiology Diagnostic Reference Levels (DRLs). CONCLUSIONS: Concerning variations in CT practices and protocols across several hospitals in Africa were demonstrated, emphasizing the need for better protocol optimization to improve patient safety.