Adoption of LLM AI tools in everyday tasks: A multi-site cross-sectional qualitative study of Chinese hospital administrators (Preprint)

Journal of Medical Internet Research, Год журнала: 2025, Номер unknown

Опубликована: Янв. 2, 2025

Large Language Model (LLM) artificial intelligence (AI) tools have the potential to streamline healthcare administration by enhancing efficiency in document drafting, resource allocation, and communication tasks. Despite this potential, adoption of such among hospital administrators remains understudied, particularly at individual level. To explore factors influencing utilization LLM AI China, focusing on enablers, barriers, practical applications daily administrative A multi-center, cross-sectional, descriptive qualitative design was employed. Three tertiary hospitals located Beijing (Site 1), Shenzhen 2), Chengdu 3) were selected represent diverse geographic regions institutional profiles. Middle-level recruited using purposive sampling. Data collected from June 11 August 16, 2024 through face-to-face semi-structured interviews guided a collaboratively developed piloted interview guide. Each audio-recorded transcribed verbatim. Colaizzi's method employed for thematic analysis. saturation determined per-site basis continuously reviewing transcripts during biweekly meetings until no new themes emerged additional interviews. total 31 participants 1: 9; Site 2: 10; 3: 12) completed lasting an average 27.3 min (range: 21-39 min). Only 22.6% reported high familiarity with tools, 25.8% frequent users while 45.2% rare users. Adoption varied site. 3 had highest proportion high-familiarity who consistently used more frequently. Qualitative analysis revealed that positive early experiences prior technological expertise facilitated adoption, whereas mistrust tool accuracy, limited prompting skills, insufficient training significant barriers. Participants predominantly drafting strongly advocated structured tutorials support enhance broader utilization. Familiarity technology, experiences, openness innovation may facilitate barriers as knowledge, skills can hinder use. are now primarily basic tasks application advanced functionalities due lack confidence. Structured needed usability integration. Targeted programs, combined organizational strategies build trust improve accessibility, could rates broaden usage. Future quantitative investigations should validate rate factors.

Язык: Английский

---

Understanding natural language: Potential application of large language models to ophthalmology

Asia-Pacific Journal of Ophthalmology, Год журнала: 2024, Номер 13(4), С. 100085 - 100085

Опубликована: Июль 1, 2024

Large language models (LLMs), a natural processing technology based on deep learning, are currently in the spotlight. These closely mimic comprehension and generation. Their evolution has undergone several waves of innovation similar to convolutional neural networks. The transformer architecture advancement generative artificial intelligence marks monumental leap beyond early-stage pattern recognition via supervised learning. With expansion parameters training data (terabytes), LLMs unveil remarkable human interactivity, encompassing capabilities such as memory retention comprehension. advances make particularly well-suited for roles healthcare communication between medical practitioners patients. In this comprehensive review, we discuss trajectory their potential implications clinicians For clinicians, can be used automated documentation, given better inputs extensive validation, may able autonomously diagnose treat future. patient care, triage suggestions, summarization documents, explanation patient's condition, customizing education materials tailored level. limitations possible solutions real-world use also presented. Given rapid advancements area, review attempts briefly cover many that play ophthalmic space, with focus improving quality delivery.

Язык: Английский

---

Large Language Model Prompting Techniques for Advancement in Clinical Medicine

Journal of Clinical Medicine, Год журнала: 2024, Номер 13(17), С. 5101 - 5101

Опубликована: Авг. 28, 2024

Large Language Models (LLMs have the potential to revolutionize clinical medicine by enhancing healthcare access, diagnosis, surgical planning, and education. However, their utilization requires careful, prompt engineering mitigate challenges like hallucinations biases. Proper of LLMs involves understanding foundational concepts such as tokenization, embeddings, attention mechanisms, alongside strategic prompting techniques ensure accurate outputs. For innovative solutions, it is essential maintain ongoing collaboration between AI technology medical professionals. Ethical considerations, including data security bias mitigation, are critical application. By leveraging supplementary resources in research education, we can enhance learning support knowledge-based inquiries, ultimately advancing quality accessibility care. Continued development necessary fully realize transforming healthcare.

Язык: Английский

---

The Role of Prompt Engineering for Multimodal LLM Glaucoma Diagnosis

medRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 1, 2024

Abstract Background and Aim This study evaluates the diagnostic performance of multimodal large language models (LLMs), GPT-4o Claude Sonnet 3.5, in detecting glaucoma from fundus images. We specifically assess impact prompt engineering use reference images on model performance. Methods utilized ACRIMA public dataset, comprising 705 labeled images, designed four types, ranging simple instructions to more refined prompts with The two were tested across 5640 API runs, accuracy, sensitivity, specificity, PPV, NPV assessed through non-parametric statistical tests. Results 3.5 achieved a highest sensitivity 94.92%, specificity 73.46%, F1 score 0.726. reached 81.47%, 50.49%, 0.645. incorporation improved GPT-4o’s accuracy by 39.8% 3.5’s 64.2%, significantly enhancing both models’ Conclusion Multimodal LLMs demonstrated potential diagnosing glaucoma, achieving far exceeding 22% reported for primary care physicians literature. Prompt engineering, especially As become integrated into medical practice, efficient design may be key, training doctors these tools effectively could enhance clinical outcomes.

Язык: Английский

---

Assessing the performance of AI chatbots in answering patients’ common questions about low back pain

Annals of the Rheumatic Diseases, Год журнала: 2024, Номер 84(1), С. 143 - 149

Опубликована: Сен. 19, 2024

The aim of this study was to assess the accuracy and readability answers generated by large language model (LLM)-chatbots common patient questions about low back pain (LBP). This cross-sectional analysed responses 30 LBP-related questions, covering self-management, risk factors treatment. were developed experienced clinicians researchers piloted with a group consumer representatives lived experience LBP. inquiries inputted in prompt form into ChatGPT 3.5, Bing, Bard (Gemini) 4.0. Responses evaluated relation their accuracy, presence disclaimers health advice. assessed comparing recommendations main guidelines for two independent reviewers classified as accurate, inaccurate or unclear. Readability measured Flesch Reading Ease Score (FRES). Out 120 yielding 1069 recommendations, 55.8% 42.1% 1.9% Treatment self-management domains showed highest while had most inaccuracies. Overall, LLM-chatbots provided that 'reasonably difficult' read, mean (SD) FRES score 50.94 (3.06). Disclaimer advice present around 70%-100% produced. use tools education counselling LBP shows promising but variable results. These chatbots generally provide moderately accurate recommendations. However, may vary depending on topic each question. reliability level inadequate, potentially affecting patient's ability comprehend information.

Язык: Английский

---

How do GPs Want Large Language Models to be Applied in Primary Care, and What Are Their Concerns? A Cross‐Sectional Survey

Journal of Evaluation in Clinical Practice, Год журнала: 2025, Номер 31(4)

Опубликована: Май 14, 2025

ABSTRACT Introduction Although the potential utility of large language models (LLMs) in medicine and healthcare is substantial, no assessment has been made to date how GPs want LLMs be applied primary care, or which issues are most concerned about regarding implementation into their clinical practice. This study's objective was generate preliminary evidence that answers these questions, relevant because themselves will ultimately harness power care. Methods Non‐probability sampling utilised: practicing UK who were members one two Facebook groups (one containing a community care staff, other GMC‐registered doctors UK) invited complete an online survey, ran from 06 13 November 2024. Results The survey received 113 responses, 107 UK. When LLM accuracy safety assumed guaranteed, broad enthusiasm for carrying out various nonclinical tasks reported. single task respondents supportive listening consultation writing notes real‐time GP review, edit, save (44.0%), identifying outstanding actioning them (51.0%), respectively. Respondents with range being embedded systems, patient commonly reported issue concern (36.2%). Discussion study generated those developing use Further research required expand this base further inform development technologies, ensure they acceptable them.

Язык: Английский

---

Use of AI in family medicine publications: a joint editorial from journal editors

Evidence-Based Practice, Год журнала: 2025, Номер 28(1), С. 1 - 4

Опубликована: Янв. 1, 2025

Schrager, Sarina MD, MS; Seehusen, Dean A. MPH; Sexton, Sumi M. MD; Richardson, Caroline Neher, Jon Pimlott, Nicholas Bowman, Marjorie Rodíguez, José Morley, Christopher P. PhD; Li, Li PhD, Dera, James Dom MD Author Information

Язык: Английский

---

Use of AI in Family Medicine Publications: A Joint Editorial From Journal Editors

Family Medicine, Год журнала: 2025, Номер 57(1), С. 1 - 5

Опубликована: Янв. 13, 2025

Язык: Английский

---

Use of AI in Family Medicine Publications: A Joint Editorial From Journal Editors

PRiMER, Год журнала: 2025, Номер 9

Опубликована: Янв. 13, 2025

There are multiple guidelines from publishers and organizations on the use of artiXcial intelligence (AI) in publishing.However, none speciXc to family medicine.Most journals have some basic AI recommendations for authors, but more explicit direction is needed, as not all tools same.

Язык: Английский

---

Use of AI in family medicine publications: a joint editorial from journal editors

Family Medicine and Community Health, Год журнала: 2025, Номер 13(1), С. e003238 - e003238

Опубликована: Янв. 1, 2025

There are multiple guidelines from publishers and organisations on the use of artificial intelligence (AI) in publishing.[1–5][1] However, none specific to family medicine. Most journals have some basic AI recommendations for authors, but more explicit direction is needed, as not all

Язык: Английский

---