Objective: Chatbots have the potential to improve user compliance in electronic Patient-Reported Outcome (ePRO) system. Compared to rule-based chatbots, Large Language Model (LLM) offers advantages such as simplifying the development process and increasing conversational flexibility. However, there is currently a lack of practical applications of LLMs in ePRO systems. Therefore, this study utilized ChatGPT to develop the Chat-ePRO system and designed a pilot study to explore the feasibility of building an ePRO system based on LLM.
Materials And Methods: This study employed prompt engineering and offline knowledge distillation to design a dialogue algorithm and built the Chat-ePRO system on the WeChat Mini Program platform. In order to compare Chat-ePRO with the form-based ePRO and rule-based chatbot ePRO used in previous studies, we conducted a pilot study applying the three ePRO systems sequentially at the Sir Run Run Shaw Hospital to collect patients' PRO data.
Result: Chat-ePRO is capable of correctly generating conversation based on PRO forms (success rate: 95.7 %) and accurately extracting the PRO data instantaneously from conversation (Macro-F1: 0.95). The majority of subjective evaluations from doctors (>70 %) suggest that Chat-ePRO is able to comprehend questions and consistently generate responses. Pilot study shows that Chat-ePRO demonstrates higher response rate (9/10, 90 %) and longer interaction time (10.86 s/turn) compared to the other two methods.
Conclusion: Our study demonstrated the feasibility of utilizing algorithms such as prompt engineering to drive LLM in completing ePRO data collection tasks, and validated that the Chat-ePRO system can effectively enhance patient compliance.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.jbi.2024.104651 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
mHealth Incentivized Adherence Plus Patient Navigation (MIAPP): protocol for a pilot randomized controlled trial to improve linkage and retention on buprenorphine for hospitalized patients with methamphetamine use and opioid use disorder.
Addict Sci Clin Pract
January 2025
Department of Medicine, Division of General Internal Medicine, University of Washington/Harborview Medical Center, 325 9Th Avenue, Box 359780, Seattle, WA, 98104, USA.
Background: Initiation of buprenorphine for treatment of opioid use disorder (OUD) in acute care settings improves access and outcomes, however patients who use methamphetamine are less likely to link to ongoing treatment. We describe the intervention and design from a pilot randomized controlled trial of an intervention to increase linkage to and retention in outpatient buprenorphine services for patients with OUD and methamphetamine use who initiate buprenorphine in the hospital.
Methods: The study is a two-arm pilot randomized controlled trial (N = 40) comparing the mHealth Incentivized Adherence Plus Patient Navigation (MIAPP) intervention to treatment as usual.
Unsupervised detection of novel SARS-CoV-2 mutations and lineages in wastewater samples using long-read sequencing.
BMC Genomics
January 2025
Department of Virology, Norwegian Institute of Public Health, Oslo, 0456, Norway.
The COVID-19 pandemic has underscored the importance of virus surveillance in public health and wastewater-based epidemiology (WBE) has emerged as a non-invasive, cost-effective method for monitoring SARS-CoV-2 and its variants at the community level. Unfortunately, current variant surveillance methods depend heavily on updated genomic databases with data derived from clinical samples, which can become less sensitive and representative as clinical testing and sequencing efforts decline.In this paper, we introduce HERCULES (High-throughput Epidemiological Reconstruction and Clustering for Uncovering Lineages from Environmental SARS-CoV-2), an unsupervised method that uses long-read sequencing of a single 1 Kb fragment of the Spike gene.
View Article and Find Full Text PDFThe impact of carbon emissions trading on green total factor productivity based on evidence from a quasi-natural experiment.
Sci Rep
January 2025
Faculty of Social Sciences, University of Lodz, ul. Prez, prez. Gabriela Narutowicza 68, 90-136, Łódź, Poland.
Based on a balanced panel dataset of 272 prefecture-level cities from 2000 to 2022, this paper systematically investigates the impact of the carbon emissions trading system on green total factor productivity and its underlying mechanisms from an integrated perspective of overall, dynamic, and spatial dimensions. The findings reveal that (1) the carbon emissions trading system significantly enhances regional total factor productivity, primarily by optimizing resource allocation efficiency and strengthening regional competitiveness. (2) From a dynamic perspective, the policy effect exhibited a U-shaped relationship: from 2013 to 2018, green total factor productivity was suppressed due to underdeveloped market mechanisms and the policy environment; after 2018, with market maturation and policy stabilization, the policy effects improved significantly.
View Article and Find Full Text PDFSuperior Outcomes with Ultrasound-Guided Hyaluronidase for Impending Filler-Induced Facial Skin Necrosis: A Systematic Review and Pilot Meta-Analysis.
Aesthetic Plast Surg
January 2025
Department of Plastic and Reconstructive Surgery, John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, UK.
Background: Hyaluronidase remains the mainstay treatment for impending filler-induced facial skin necrosis. Complete resolution of impending skin necrosis following hyaluronidase injection is estimated to be around 77.8%.
View Article and Find Full Text PDFAssociation of digital eye strain with central corneal thickness: a cross-sectional observational pilot study.
Int Ophthalmol
January 2025
Department of Ophthalmology, Command Hospital, Pune, 411040, India.
Purpose: This study aimed to evaluate whether the digital eye strain (DES) was associated with the low central corneal thickness (CCT).
Methods: This observational cross-sectional pilot study was conducted from April 2023 to October 2023 at a tertiary eye care centre in North India, where CCT values were compared between subjects with DES and those without DES. Two hundred and eighty subjects (n = 280) aged 15-40 years with clear corneas and lenses were initially included in this study.
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!