Dataset Information

Black Box Warning: Large Language Models and the Future of Infectious Diseases Consultation.

ABSTRACT: Large language models (LLMs) are artificial intelligence systems trained by deep learning algorithms to process natural language and generate text responses to user prompts. Some approach physician performance on a range of medical challenges, leading some proponents to advocate for their potential use in clinical consultation and prompting some consternation about the future of cognitive specialties. However, LLMs currently have limitations that preclude safe clinical deployment in performing specialist consultations, including frequent confabulations, lack of contextual awareness crucial for nuanced diagnostic and treatment plans, inscrutable and unexplainable training data and methods, and propensity to recapitulate biases. Nonetheless, considering the rapid improvement in this technology, growing calls for clinical integration, and healthcare systems that chronically undervalue cognitive specialties, it is critical that infectious diseases clinicians engage with LLMs to enable informed advocacy for how they should-and shouldn't-be used to augment specialist care.

SUBMITTER: Schwartz IS

PROVIDER: S-EPMC11006107 | biostudies-literature | 2024 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Black Box Warning: Large Language Models and the Future of Infectious Diseases Consultation.

Schwartz Ilan S IS Link Katherine E KE Daneshjou Roxana R Cortés-Penfield Nicolás N

Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 20240401 4

Large language models (LLMs) are artificial intelligence systems trained by deep learning algorithms to process natural language and generate text responses to user prompts. Some approach physician performance on a range of medical challenges, leading some proponents to advocate for their potential use in clinical consultation and prompting some consternation about the future of cognitive specialties. However, LLMs currently have limitations that preclude safe clinical deployment in performing s ...[more]

PMID: 37971399

Similar Datasets

Project description:BackgroundInfectious diseases have consistently been a significant concern in public health, requiring proactive measures to safeguard societal well-being. In this regard, regular monitoring activities play a crucial role in mitigating the adverse effects of diseases on society. To monitor disease trends, various organizations, such as the World Health Organization (WHO) and the European Centre for Disease Prevention and Control (ECDC), collect diverse surveillance data and make them publicly accessible. However, these platforms primarily present surveillance data in English, which creates language barriers for non-English-speaking individuals and global public health efforts to accurately observe disease trends. This challenge is particularly noticeable in regions such as the Middle East, where specific infectious diseases, such as Middle East respiratory syndrome coronavirus (MERS-CoV), have seen a dramatic increase. For such regions, it is essential to develop tools that can overcome language barriers and reach more individuals to alleviate the negative impacts of these diseases.ObjectiveThis study aims to address these issues; therefore, we propose InfectA-Chat, a cutting-edge large language model (LLM) specifically designed for the Arabic language but also incorporating English for question and answer (Q&A) tasks. InfectA-Chat leverages its deep understanding of the language to provide users with information on the latest trends in infectious diseases based on their queries.MethodsThis comprehensive study was achieved by instruction tuning the AceGPT-7B and AceGPT-7B-Chat models on a Q&A task, using a dataset of 55,400 Arabic and English domain-specific instruction-following data. The performance of these fine-tuned models was evaluated using 2770 domain-specific Arabic and English instruction-following data, using the GPT-4 evaluation method. A comparative analysis was then performed against Arabic LLMs and state-of-the-art models, including AceGPT-13B-Chat, Jais-13B-Chat, Gemini, GPT-3.5, and GPT-4. Furthermore, to ensure the model had access to the latest information on infectious diseases by regularly updating the data without additional fine-tuning, we used the retrieval-augmented generation (RAG) method.ResultsInfectA-Chat demonstrated good performance in answering questions about infectious diseases by the GPT-4 evaluation method. Our comparative analysis revealed that it outperforms the AceGPT-7B-Chat and InfectA-Chat (based on AceGPT-7B) models by a margin of 43.52%. It also surpassed other Arabic LLMs such as AceGPT-13B-Chat and Jais-13B-Chat by 48.61%. Among the state-of-the-art models, InfectA-Chat achieved a leading performance of 23.78%, competing closely with the GPT-4 model. Furthermore, the RAG method in InfectA-Chat significantly improved document retrieval accuracy. Notably, RAG retrieved more accurate documents based on queries when the top-k parameter value was increased.ConclusionsOur findings highlight the shortcomings of general Arabic LLMs in providing up-to-date information about infectious diseases. With this study, we aim to empower individuals and public health efforts by offering a bilingual Q&A system for infectious disease monitoring.

Project description:BackgroundThere is still a relatively serious disease burden of infectious diseases and the warning time for different infectious diseases before implementation of interventions is important. The logistic differential equation models can be used for predicting early warning of infectious diseases. The aim of this study is to compare the disease fitting effects of the logistic differential equation (LDE) model and the generalized logistic differential equation (GLDE) model for the first time using data on multiple infectious diseases in Jilin Province and to calculate the early warning signals for different types of infectious diseases using these two models in Jilin Province to solve the disease early warning schedule for Jilin Province throughout the year.MethodsCollecting the incidence of 22 infectious diseases in Jilin Province, China. The LDE and GLDE models were used to calculate the recommended warning week (RWW), the epidemic acceleration week (EAW) and warning removed week (WRW) for acute infectious diseases with seasonality, respectively.ResultsFive diseases were selected for analysis based on screening principles: hemorrhagic fever with renal syndrome (HFRS), shigellosis, mumps, Hand, foot and mouth disease (HFMD), and scarlet fever. The GLDE model fitted the above diseases better (0.80 ≤ R2 ≤ 0.94, P < 0. 005) than the LDE model. The estimated warning durations (per year) of the LDE model for the above diseases were: weeks 12-23 and 40-50; weeks 20-36; weeks 15-24 and 43-52; weeks 26-34; and weeks 16-25 and 41-50. While the durations of early warning (per year) estimated by the GLDE model were: weeks 7-24 and 36-51; weeks 13-37; weeks 11-26 and 39-54; weeks 23-35; and weeks 12-26 and 40-50.ConclusionsCompared to the LDE model, the GLDE model provides a better fit to the actual disease incidence data. The RWW appeared to be earlier when estimated with the GLDE model than the LDE model. In addition, the WRW estimated with the GLDE model were more lagged and had a longer warning time.

Project description:The United States Food and Drug Administration issued a Black Box warning in October 2004 after placebo-controlled trials of antidepressant medications found an increased risk of suicidal thoughts and behaviors among children and adolescents taking antidepressant medications relative to placebo. Subsequently, some researchers have concluded that the Black Box warning caused severe unintended consequences; specifically, they have argued that the warning led to reduced use of antidepressants among youth, which led to more suicides. In this paper, we critically examine research regarding the Black Box warning's alleged deleterious consequences. One study claimed that controlled trials did not actually find an increased risk of suicidality among youth taking fluoxetine relative to those taking placebo, but its measure of suicidality is likely invalid. We found that ecological time series studies claiming that decreasing antidepressant prescriptions are linked to higher rates of suicide attempts or actual suicides among youth were methodologically weak. These studies exhibited shortcomings including: selective use of time points, use of only a short-term time series, lack of performing statistical analysis, not examining level of severity/impairment among participants, inability to control confounding variables, and/or use of questionable measures of suicide attempts. Further, while some time-series studies claim that increased antidepressant prescriptions are related to fewer youth suicides, more recent data suggests that increasing antidepressant prescriptions are related to more youth suicide attempts and more completed suicides among American children and adolescents. We also note that case-control studies show increased risk of suicide attempts and suicide among youth taking antidepressants, even after controlling for some relevant confounds. As clinical trials have the greatest ability to control relevant confounds, it is important to remember such trials demonstrated increased risk of suicidality adverse events among youth taking antidepressants. The Black Box warning is firmly rooted in solid data whereas attempts to claim the warning has caused harm are based on quite weak evidence.

Dataset Information

Black Box Warning: Large Language Models and the Future of Infectious Diseases Consultation.

Publications

Black Box Warning: Large Language Models and the Future of Infectious Diseases Consultation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets