Me-LLaMA: Foundation Large Language Models for Medical Applications.

Recent advancements in large language models (LLMs) such as ChatGPT and LLaMA have hinted at their potential to revolutionize medical applications, yet their application in clinical settings often reveals limitations due to a lack of specialized training on medical-specific data. In response to this challenge, this study introduces Me-LLaMA, a novel medical LLM family that includes foundation models - Me-LLaMA 13/70B, along with their chat-enhanced versions - Me-LLaMA 13/70B-chat, developed through continual pre-training and instruction tuning of LLaMA2 using large medical datasets. Our methodology leverages a comprehensive domain-specific data suite, including a large-scale, continual pre-training dataset with 129B tokens, an instruction tuning dataset with 214k samples, and a new medical evaluation benchmark (MIBE) across six critical medical tasks with 12 datasets. Our extensive evaluation using the MIBE shows that Me-LLaMA models achieve overall better performance than existing open-source medical LLMs in zero-shot, few-shot and supervised learning abilities. With task-specific instruction tuning, Me-LLaMA models outperform ChatGPT on 7 out of 8 datasets and GPT-4 on 5 out of 8 datasets. In addition, we investigated the catastrophic forgetting problem, and our results show that Me-LLaMA models outperform other open-source medical LLMs in mitigating this issue. Me-LLaMA is one of the largest open-source medical foundation LLMs that use both biomedical and clinical data. It exhibits superior performance across both general and medical tasks compared to other open-source medical LLMs, rendering it an attractive choice for medical AI applications. We release our models, datasets, and evaluation scripts at: https://github.com/BIDS-Xu-Lab/Me-LLaMA.

Web-Based Dashboard for the Interactive Visualization and Analysis of National Risk-Standardized Mortality Rates of Sepsis in the US.

Sepsis mortality is heavily influenced by the quality of care in hospitals. Comparing risk-standardized mortality rate (RSMR) of sepsis patients in different states in the United States has potentially important clinical and policy implications. In the current study, we aimed to compare national sepsis RSMR using an interactive web-based dashboard. We analyzed sepsis mortality using the National Inpatient Sample Database of the US. The RSMR was calculated by the hierarchical logistic regression model. We wrote the interactive web-based dashboard using the Shiny framework, an R package that integrates R-based statistics computation and graphics generation. Visual summarizations (e.g., heat map, and time series chart), and interactive tools (e.g., year selection, automatic year play, map zoom, copy or print data, ranking data by name or value, and data search) were implemented to enhance user experience. The web-based dashboard (https://sepsismap.shinyapps.io/index2/) is cross-platform and publicly available to anyone with interest in sepsis outcomes, health inequality, and administration of state/federal healthcare. After extrapolation to the national level, approximately 35 million hospitalizations were analyzed for sepsis mortality each year. Eight years of sepsis mortality data were summarized into four easy to understand dimensions: Sepsis Identification Criteria; Sepsis Mortality Predictors; RSMR Map; RSMR Trend. Substantial variation in RSMR was observed for different states in the US. This web-based dashboard allows anyone to visualize the substantial variation in RSMR across the whole US. Our work has the potential to support healthcare transparency, information diffusion, health decision-making, and the formulation of new public policies.

A Web-Based Clinical System for Cohort Surveillance of Specific Clinical Effectiveness and Safety Outcomes: A Cohort Study of Non-Vitamin K Antagonist Oral Anticoagulants and Warfarin.

BACKGROUND: Conventional systems of drug surveillance lack a seamless workflow, which makes it crucial to have an active drug surveillance system that proactively assesses adverse drug events. OBJECTIVE: The aim of this study was to develop a seamless, Web-based workflow for comparing the safety and effectiveness of drugs in a database of electronic medical records. METHODS: We proposed a comprehensive integration process for cohort surveillance using the National Taiwan University Hospital Clinical Surveillance System (NCSS). We studied a practical application of the NCSS that evaluates the drug safety and effectiveness of novel oral anticoagulants (NOACs) and warfarin by cohort tree analysis in an efficient and interoperable platform. RESULTS: We demonstrated a practical example of investigating the differences in effectiveness and safety between NOACs and warfarin in patients with nonvalvular atrial fibrillation (AF) using the NCSS. We efficiently identified 2357 patients with nonvalvular AF with newly prescribed oral anticoagulants between 2010 and 2015 and further developed 1 main cohort and 2 subcohorts for separately measuring ischemic stroke as the clinical effectiveness outcome and intracranial hemorrhage (ICH) as the safety outcome. In the subcohort of ischemic stroke, NOAC users exhibited a significantly lower risk of ischemic stroke than warfarin users after adjusting for age, sex, comorbidity, and comedication in an intention-to-treat (ITT) analysis (P=.01) but did not exhibit a significantly distinct risk in an as-treated (AT) analysis (P=.12) after the 2-year follow-up. In the subcohort of ICH, NOAC users did not exhibit a different risk of ICH both in ITT (P=.68) and AT analyses (P=.15). CONCLUSIONS: With a seamless and Web-based workflow, the NCSS can serve the critical role of forming associations between evidence and the real world at a medical center in Taiwan.

Identifying Unmet Treatment Needs for Patients With Osteoporotic Fracture: Feasibility Study for an Electronic Clinical Surveillance System.

BACKGROUND: Traditional clinical surveillance relied on the results from clinical trials and observational studies of administrative databases. However, these studies not only required many valuable resources but also faced a very long time lag. OBJECTIVE: This study aimed to illustrate a practical application of the National Taiwan University Hospital Clinical Surveillance System (NCSS) in the identification of patients with an osteoporotic fracture and to provide a high reusability infrastructure for longitudinal clinical data. METHODS: The NCSS integrates electronic medical records in the National Taiwan University Hospital (NTUH) with a data warehouse and is equipped with a user-friendly interface. The NCSS was developed using professional insight from multidisciplinary experts, including clinical practitioners, epidemiologists, and biomedical engineers. The practical example identifying the unmet treatment needs for patients encountering major osteoporotic fractures described herein was mainly achieved by adopting the computerized workflow in the NCSS. RESULTS: We developed the infrastructure of the NCSS, including an integrated data warehouse and an automatic surveillance workflow. By applying the NCSS, we efficiently identified 2193 patients who were newly diagnosed with a hip or vertebral fracture between 2010 and 2014 at NTUH. By adopting the filter function, we identified 1808 (1808/2193, 82.44%) patients who continued their follow-up at NTUH, and 464 (464/2193, 21.16%) patients who were prescribed anti-osteoporosis medications, within 3 and 12 months post the index date of their fracture, respectively. CONCLUSIONS: The NCSS systems can integrate the workflow of cohort identification to accelerate the survey process of clinically relevant problems and provide decision support in the daily practice of clinical physicians, thereby making the benefit of evidence-based medicine a reality.