There is an increasing interest in developing artificial intelligence (AI) systems to process and interpret electronic health records (EHRs). Natural language processing (NLP) powered by pretrained language models is the key technology for medical AI systems utilizing clinical narratives. However, there are few clinical language models, the largest of which trained in the clinical domain is comparatively small at 110 million parameters (compared with billions of parameters in the general domain). It is not clear how large clinical language models with billions of parameters can help medical AI systems utilize unstructured EHRs. In this study, we develop from scratch a large clinical language model—GatorTron—using >90 billion words of text (including >82 billion words of de-identified clinical text) and systematically evaluate it on five clinical NLP tasks including clinical concept extraction, medical relation extraction, semantic textual similarity, natural language inference (NLI), and medical question answering (MQA). We examine how (1) scaling up the number of parameters and (2) scaling up the size of the training data could benefit these NLP tasks. GatorTron models scale up the clinical language model from 110 million to 8.9 billion parameters and improve five clinical NLP tasks (e.g., 9.6% and 9.5% improvement in accuracy for NLI and MQA), which can be applied to medical AI systems to improve healthcare delivery. The GatorTron models are publicly available at:https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/models/gatortron_og.

There are enormous enthusiasm and concerns in applying large language models (LLMs) to healthcare. Yet current assumptions are based on general-purpose LLMs such as ChatGPT, which are not developed for medical use. This study develops a generative clinical LLM, GatorTronGPT, using 277 billion words of text including (1) 82 billion words of clinical text from 126 clinical departments and approximately 2 million patients at the University of Florida Health and (2) 195 billion words of diverse general English text. We train GatorTronGPT using a GPT-3 architecture with up to 20 billion parameters and evaluate its utility for biomedical natural language processing (NLP) and healthcare text generation. GatorTronGPT improves biomedical natural language processing. We apply GatorTronGPT to generate 20 billion words of synthetic text. Synthetic NLP models trained using synthetic text generated by GatorTronGPT outperform models trained using real-world clinical text. Physicians’ Turing test using 1 (worst) to 9 (best) scale shows that there are no significant differences in linguistic readability (p = 0.22; 6.57 of GatorTronGPT compared with 6.93 of human) and clinical relevance (p = 0.91; 7.0 of GatorTronGPT compared with 6.97 of human) and that physicians cannot differentiate them (p < 0.001). This study provides insights into the opportunities and challenges of LLMs for medical research and healthcare.

Although Short Form (SF)-12 × 2® has been extensively studied and used as a valid measure of health-related quality of life in a variety of population groups, no systematic studies have described the reliability of the measure in patients with behavioral conditions or serious mental illness (SMI). We assessed the internal consistency, split-half reliability and annual test-retest correlations in a sample of 1587 participants with either a combination of physical and behavioral conditions or SMI. The Mosier’s alpha was 0.70 for the Physical Composite Scale (PCS) and 0.69 for the Mental Health Composite Scale (MCS), indicating good internal consistency. We observed strong correlations between physical functioning, physical role and body pain scales (r = 0.55–0.56), and between social functioning, emotional role, and mental health (r = 0.53–0.58). We calculated split-half reliabilities to be 0.74 for physical functioning, 0.75 for physical role, 0.73 for emotional role and 0.65 for mental health respectively. We assessed the annual test-retest correlation using intraclass correlation (ICC) and found an ICC of 0.61 for PCS and 0.57 for MCS composite scores, adjusting for age, sex, race/ethnicity, and CRG. We found no decline in the correlations between baseline and the following study years until year 3. Our results encourage using SF-12v2® to assess health-related quality of life in the Medicaid population with combined physical and behavioral conditions or similar cohorts. The WIN study was registered with clinicaltrials.gov on April 22, 2015. Trial registration number: NCT02440906 . Retrospectively registered.

The post-acute sequelae of SARS-CoV-2 infection (PASC) refers to a broad spectrum of symptoms and signs that are persistent, exacerbated or newly incident in the period after acute SARS-CoV-2 infection. Most studies have examined these conditions individually without providing evidence on co-occurring conditions. In this study, we leveraged the electronic health record data of two large cohorts, INSIGHT and OneFlorida+, from the national Patient-Centered Clinical Research Network. We created a development cohort from INSIGHT and a validation cohort from OneFlorida+ including 20,881 and 13,724 patients, respectively, who were SARS-CoV-2 infected, and we investigated their newly incident diagnoses 30–180 days after a documented SARS-CoV-2 infection. Through machine learning analysis of over 137 symptoms and conditions, we identified four reproducible PASC subphenotypes, dominated by cardiac and renal (including 33.75% and 25.43% of the patients in the development and validation cohorts); respiratory, sleep and anxiety (32.75% and 38.48%); musculoskeletal and nervous system (23.37% and 23.35%); and digestive and respiratory system (10.14% and 12.74%) sequelae. These subphenotypes were associated with distinct patient demographics, underlying conditions before SARS-CoV-2 infection and acute infection phase severity. Our study provides insights into the heterogeneity of PASC and may inform stratified decision-making in the management of PASC conditions.

This paper studies the temporal consistency of health care expenditures in a large state Medicaid program. Predictive machine learning models were used to forecast the expenditures, especially for the high-cost, high-need (HCHN) patients. We systematically tests temporal correlation of patient-level health care expenditures in both the short and long terms. The results suggest that medical expenditures are significantly correlated over multiple periods. Our work demonstrates a prevalent and strong temporal correlation and shows promise for predicting future health care expenditures using machine learning. Temporal correlation is stronger in HCHN patients and their expenditures can be better predicted. Including more past periods is beneficial for better predictive performance. This study shows that there is significant temporal correlation in health care expenditures. Machine learning models can help to accurately forecast the expenditures. These results could advance the field toward precise preventive care to lower overall health care costs and deliver care more efficiently.

Problem: Developing a national pragmatic clinical trial infrastructure is central to understanding the effectiveness of interventions applied under usual conditions and where people receive health care. To address this challenge, three Florida universities—the University of Florida Clinical and Translational Science Institute, Florida State University (with its community-based distributive medical education model), and the University of Miami—created (2010–2013) a statewide consortium, the OneFlorida Clinical Research Consortium, to support the conduct of pragmatic clinical trials and provide mentored research experiences for medical and graduate students in real-world practice settings. Approach: OneFlorida has four programs, which report to a steering committee with membership from each partner, community members, and the state Medicaid agency and Department of Health to ensure shared governance. The Clinical Research Program provides support to conduct research in the network and uses champions to engage community clinicians. The Citizen Scientist Program has community members who provide input on health topics of importance to them, study design, recruitment and retention strategies, and the interpretation of findings. The Data Trust Program contains electronic health record and health care claims data for 10.6 million Floridians. The Minority Education Program, in collaboration with three historically black colleges and universities, offers minority junior faculty mentoring in pragmatic clinical trials and implementation science. Outcomes: OneFlorida has implemented 27 studies with diverse patient populations and in diverse community practice settings. Next Steps: To identify evidence-based best practices from the clinical trials conducted in the network, foster their implementation, and expand research training opportunities.