Creation and Adoption of Large Language Models in Medicine.

Importance: There is increased interest in and potential benefits from using large language models (LLMs) in medicine. However, by simply wondering how the LLMs and the applications powered by them will reshape medicine instead of getting actively involved, the agency in shaping how these tools can be used in medicine is lost. Observations: Applications powered by LLMs are increasingly used to perform medical tasks without the underlying language model being trained on medical records and without verifying their purported benefit in performing those tasks. Conclusions and Relevance: The creation and use of LLMs in medicine need to be actively shaped by provisioning relevant training data, specifying the desired benefits, and evaluating the benefits via testing in real-world deployments.

Large language models encode clinical knowledge.

Large language models (LLMs) have demonstrated impressive capabilities, but the bar for clinical applications is high. Attempts to assess the clinical knowledge of models typically rely on automated evaluations based on limited benchmarks. Here, to address these limitations, we present MultiMedQA, a benchmark combining six existing medical question answering datasets spanning professional medicine, research and consumer queries and a new dataset of medical questions searched online, HealthSearchQA. We propose a human evaluation framework for model answers along multiple axes including factuality, comprehension, reasoning, possible harm and bias. In addition, we evaluate Pathways Language Model1 (PaLM, a 540-billion parameter LLM) and its instruction-tuned variant, Flan-PaLM2 on MultiMedQA. Using a combination of prompting strategies, Flan-PaLM achieves state-of-the-art accuracy on every MultiMedQA multiple-choice dataset (MedQA3, MedMCQA4, PubMedQA5 and Measuring Massive Multitask Language Understanding (MMLU) clinical topics6), including 67.6% accuracy on MedQA (US Medical Licensing Exam-style questions), surpassing the prior state of the art by more than 17%. However, human evaluation reveals key gaps. To resolve this, we introduce instruction prompt tuning, a parameter-efficient approach for aligning LLMs to new domains using a few exemplars. The resulting model, Med-PaLM, performs encouragingly, but remains inferior to clinicians. We show that comprehension, knowledge recall and reasoning improve with model scale and instruction prompt tuning, suggesting the potential utility of LLMs in medicine. Our human evaluations reveal limitations of today's models, reinforcing the importance of both evaluation frameworks and method development in creating safe, helpful LLMs for clinical applications.

Reflections on the design of an ultrasound study program in medical undergraduate / Reflexiones sobre el diseño de un programa de estudio de ultrasonido en el grado de medicina

Introduction: The use of didactic tools for teaching basic sciences in the medical career focuses on anatomical models, electrodiagnostic equipment, and simulation. Only some study programs incorporate images for teaching basic sciences; some of the reasons are the cost of the ultrasound equipment. However, many medical schools have the infrastructure to do so. Materials and methods: We conducted a review of the scientific literature in the Scopus, Web of Science, and Google Academic databases, after which the researchers conducted discussion sessions to select the main ideas that would help build the educational proposal. Results: Describe a proposal for curricular design for creating training programs and teacher training that allows maximizing the use of ultrasound as a teaching tool for the basic sciences of the medical career. Conclusion: The best way to strengthen the teaching of medical sciences is through constant academic training, both in disciplinary content and in teaching. Only in this way can we face the great need to train doctors who are very aware of their social responsibility.(AU)

Introducción: El uso de herramientas didácticas para la enseñanza de las ciencias básicas en la carrera de medicina se centra en modelos anatómicos, equipos de electrodiagnóstico y simulación. Solo algunos programas de estudio incorporan imágenes para la enseñanza de las ciencias básicas; algunas de las razones son el costo del equipo de ultrasonido. Sin embargo, muchas escuelas de medicina tienen la infraestructura para hacerlo. Materiales y métodos: Se realizó una revisión de la literatura científica en las bases de datos Scopus, Web of Science y Google Academic, tras lo cual los investigadores realizaron sesiones de discusión para seleccionar las ideas principales que ayudarían a construir la propuesta educativa. Resultados: Describir una propuesta de diseño curricular para la creación de programas de formación y formación docente que permita maximizar el uso de la ecografía como herramienta didáctica de las ciencias básicas de la carrera de medicina. Conclusión: La mejor manera de fortalecer la enseñanza de las ciencias médicas es a través de la formación académica constante, tanto en los contenidos disciplinares como en la docencia. Solo así podremos afrontar la gran necesidad de formar médicos muy conscientes de su responsabilidad social.(AU)

It Takes an Average of 17 Years for Evidence to Change Practice-the Burgeoning Field of Implementation Science Seeks to Speed Things Up.

This Medical News feature discusses how implementation science seeks to narrow the often yearslong gap between the development of evidence-based practices and their routine use in the real world.

Designing and describing an electronic referral system to facilitate direct hospital admissions.

BACKGROUND: In France, the progressive use of emergency departments (EDs) by primary care providers (PCPs) as a point of access to hospitalization for nonurgent patients is one of the many causes of their overcrowding. To increase the proportion of direct hospital admissions, it is necessary to improve coordination between PCPs and hospital specialists. The objective of our work was to describe the design and implementation of an electronic referral system aimed at facilitating direct hospital admissions. METHODS: This initiative was conducted in a French area (Hauts-de-Seine Sud) through a partnership between the Antoine-Béclère University Hospital, the Paris-Saclay University Department of General Medicine and the local health care network. The implementation was carried out in 3 stages, namely, conducting a survey of PCPs in the territory about their communication methods with the hospital, designing and implementing a web-based application called "SIPILINK" (Système d'Information de la Plateforme d'Intermédiation Link) and an innovative organization for hospital management of the requests, and analysing through descriptive statistics the platform use 9 months after launch. RESULTS: The e-referral platform was launched in November 2019. First, a PCP filled out an electronic form describing the reason for his or her request. Then, a hospital specialist worked to respond within 72 h. Nine months after the launch, 132 PCPs had registered for the SIPILINK platform, which represented 36.6% of PCPs in this area. Of the 124 requests made, 46.8% corresponded to a hospitalization request (conventional or day hospitalization). The most requested specialty was internal medicine (48.4% of requests). The median time to first response was 43 min, and 43.5% of these requests resulted in direct admission (conventional or day hospitalization). CONCLUSIONS: This type of system responds to a need for coordination in the primary-secondary care direction, which is less often addressed than in the secondary-primary care direction. The first results show the potential of the system to facilitate direct admissions within a short time frame. To make the system sustainable, the next step is to extend its use to other hospitals in the territory.

Temporal Cohort Logic.

We introduce a new logic, called Temporal Cohort Logic (TCL), for cohort specification and discovery in clinical and population health research. TCL is created to fill a conceptual gap in formalizing temporal reasoning in biomedicine, in a similar role that temporal logics play for computer science and its applications. We provide formal syntax and semantics for TCL and illustrate the various logical constructs using examples related to human health. Relationships and distinctions with existing temporal logical frameworks are discussed. Applications in electronic health record (EHR) and in neurophysiological data resource are provided. Our approach differs from existing temporal logics, in that we explicitly capture Allen's interval algebra as modal operators in a language of temporal logic (rather than addressing it in the semantic structure). This has two major implications. First, it provides a formal logical framework for reasoning about time in biomedicine, allowing general (i.e., higher-levels of abstraction) investigation into the properties of this approach (such as proof systems, completeness, expressiveness, and decidability) independent of a specific query language or a database system. Second, it puts our approach in the context of logical developments in computer science, allowing potential translation of existing results into the setting of TCL and its variants or subsystems so as to illuminate opportunities and computational challenges involved in temporal reasoning for biomedicine.

Early Impacts of the COVID-19 Pandemic on Telehealth Patterns in Primary Care, Mental Health, and Specialty Care Facilities in Texas.

OBJECTIVES: Since the onset of the coronavirus disease 2019 (COVID-19) pandemic, many US clinics have shifted some or all of their practice from in-person to virtual visits. In this study, we assessed the use of telehealth among primary care and specialty clinics, by targeting healthcare administrators via multiple channels. METHODS: Using an online survey, we assessed the use of, barriers to, and reimbursement for telehealth. Respondents included clinic administrators (chief executive officers, vice presidents, directors, and senior-level managers). RESULTS: A total of 85 complete responses were recorded, 79% of which represented solo or group practices and 63% reported a daily patient census >50. The proportion of clinics that delivered ≥50% of their consults using telehealth increased from 16% in March to 42% in April, 35% in May, and 30% in June. Clinics identified problems with telehealth reimbursement; although 63% of clinics reported that ≥75% of their telehealth consults were reimbursed, only 51% indicated that ≥75% of their telehealth visits were reimbursed at par with in-person office visits. Sixty-five percent of clinics reported having basic or foundational telehealth services, whereas only 9% of clinics reported advanced telehealth maturity. Value-based care participating clinics were more likely to report advanced telehealth services (27%), compared with non-value-based care clinics (3%). CONCLUSIONS: These findings highlight the adaptability of clinics to quickly transition and adopt telehealth. Uncertainty about reimbursement and policy changes may make the shift temporal, however.

Machine Learning in Epigenomics: Insights into Cancer Biology and Medicine.

The recent deluge of genome-wide technologies for the mapping of the epigenome and resulting data in cancer samples has provided the opportunity for gaining insights into and understanding the roles of epigenetic processes in cancer. However, the complexity, high-dimensionality, sparsity, and noise associated with these data pose challenges for extensive integrative analyses. Machine Learning (ML) algorithms are particularly suited for epigenomic data analyses due to their flexibility and ability to learn underlying hidden structures. We will discuss four overlapping but distinct major categories under ML: dimensionality reduction, unsupervised methods, supervised methods, and deep learning (DL). We review the preferred use cases of these algorithms in analyses of cancer epigenomics data with the hope to provide an overview of how ML approaches can be used to explore fundamental questions on the roles of epigenome in cancer biology and medicine.

Opening the black box of AI-Medicine.

One of the biggest challenges of utilizing artificial intelligence (AI) in medicine is that physicians are reluctant to trust and adopt something that they do not fully understand and regarded as a "black box." Machine Learning (ML) can assist in reading radiological, endoscopic and histological pictures, suggesting diagnosis and predict disease outcome, and even recommending therapy and surgical decisions. However, clinical adoption of these AI tools has been slow because of a lack of trust. Besides clinician's doubt, patients lacking confidence with AI-powered technologies also hamper development. While they may accept the reality that human errors can occur, little tolerance of machine error is anticipated. In order to implement AI medicine successfully, interpretability of ML algorithm needs to improve. Opening the black box in AI medicine needs to take a stepwise approach. Small steps of biological explanation and clinical experience in ML algorithm can help to build trust and acceptance. AI software developers will have to clearly demonstrate that when the ML technologies are integrated into the clinical decision-making process, they can actually help to improve clinical outcome. Enhancing interpretability of ML algorithm is a crucial step in adopting AI in medicine.