Background: The advancement of large language models (LLMs) offers significant opportunities for health care, particularly in the generation of medical documentation. However, challenges related to ensuring the accuracy and reliability of LLM outputs, coupled with the absence of established quality standards, have raised concerns about their clinical application. Objective: This study aimed to develop and validate an evaluation framework for assessing the accuracy and clinical applicability of LLM-generated emergency department (ED) records, aiming to enhance artificial intelligence integration in health care documentation. Methods: We organized the Healthcare Prompt-a-thon, a competitive event designed to explore the capabilities of LLMs in generating accurate medical records. The event involved 52 participants who generated 33 initial ED records using HyperCLOVA X, a Korean-specialized LLM. We applied a dual evaluation approach. First, clinical evaluation: 4 medical professionals evaluated the records using a 5-point Likert scale across 5 criteria—appropriateness, accuracy, structure/format, conciseness, and clinical validity. Second, quantitative evaluation: We developed a framework to categorize and count errors in the LLM outputs, identifying 7 key error types. Statistical methods, including Pearson correlation and intraclass correlation coefficients (ICC), were used to assess consistency and agreement among evaluators. Results: The clinical evaluation demonstrated strong interrater reliability, with ICC values ranging from 0.653 to 0.887 (P<.001), and a test-retest reliability Pearson correlation coefficient of 0.776 (P<.001). Quantitative analysis revealed that invalid generation errors were the most common, constituting 35.38% of total errors, while structural malformation errors had the most significant negative impact on the clinical evaluation score (Pearson r=–0.654; P<.001). A strong negative correlation was found between the number of quantitative errors and clinical evaluation scores (Pearson r=–0.633; P<.001), indicating that higher error rates corresponded to lower clinical acceptability. Conclusions: Our research provides robust support for the reliability and clinical acceptability of the proposed evaluation framework. It underscores the framework’s potential to mitigate clinical burdens and foster the responsible integration of artificial intelligence technologies in health care, suggesting a promising direction for future research and practical applications in the field.

Yongjin Yi, Hayne Park, Jaehun Jung. Kidney Res Clin Pract 1999;0:. https://doi.org/10.23876/j.krcp.18.0044

The ability to accurately predict the early progression of hemorrhagic fever with renal syndrome (HFRS) is crucial for reducing morbidity and mortality rates in severely affected patients. However, the utility of biomarkers for predicting clinical outcomes remains elusive in HFRS. The aims of the current study were to analyze the serum levels of immune function-related proteins and identify novel biomarkers that may help ascertain clinical outcomes of HFRS. Enzyme-linked immunosorbent assay, Luminex, and bioanalyzer assays were used to quantitatively detect 15 biomarkers in 49 serum samples of 26 patients with HFRS. High hemoglobin (HGB) and low urine output (UO) levels were identified as potential biomarkers associated with the acute HFRS. The serum soluble urokinase plasminogen activator receptor (suPAR) and C-X-C motif chemokine ligand 10 (CXCL10) values increased in the early phase of diseases. Elevated suPAR, interleukin-10 (IL-10), CXCL10, and decreased transforming growth factor-beta 3 (TGF-β3) were representative predictors of the disease severity. Upregulation of the HGB showed a significant correlation with high levels of suPAR and CXCL10. Reduced UO positively correlated with increased suPAR, CXCL10, and TGF-β2, and decreased vascular endothelial growth factor and TGF-β3. The changing HGB and UO criteria, high suPAR, IL-10, CXCL10, and low TGF-β3 of HFRS raise significant awareness for physicians regarding prospective biomarkers for monitoring early warning signs of HFRS. This study provides critical insights into the clinical and immunological biomarkers for disease severity and progression in patients with HFRS to identify early predictions of fatal outcomes.

Background: Hantavirus infection occurs through the inhalation of aerosolized excreta, including urine, feces, and saliva of infected rodents. The presence of Hantaan virus (HTNV) RNA or infectious particles in urine specimens of patient with hemorrhagic fever with renal syndrome (HFRS) remains to be investigated. Methodology/Principal findings: We collected four urine and serum specimens of Republic of Korea Army (ROKA) patients with HFRS. We performed multiplex PCR-based next-generation sequencing (NGS) to obtain the genome sequences of clinical HTNV in urine specimens containing ultra-low amounts of viral genomes. The epidemiological and phylogenetic analyses of HTNV demonstrated geographically homogenous clustering with those in Apodemus agrarius captured in highly endemic areas, indicating that phylogeographic tracing of HTNV genomes reveals the potential infection sites of patients with HFRS. Genetic exchange analyses showed a genetic configuration compatible with HTNV L segment exchange in nature. Conclusion/Significance: Our results suggest that whole or partial genome sequences of HTNV from the urine enabled to track the putative infection sites of patients with HFRS by phylogeographically linking to the zoonotic HTNV from the reservoir host captured at endemic regions. This report raises awareness among physicians for the presence of HTNV in the urine of patients with HFRS.

Background: Acute kidney injury (AKI) is a significant complication in patients with trauma. The early identification of AKI in these patients poses challenges. This study aimed to predict AKI in trauma patients 24 or 48 hours in advance using an extreme gradient boosting (XGBoost) model. Methods: We analyzed 17 859 trauma patients admitted to a regional trauma center between January 2015 and July 2023. Demographic, clinical, and laboratory parameters were collected. The model was developed using data until July 2021 and validated using data from August 2021. We developed models to predict AKI stages 1–3 and AKI stages 2 and 3 occurring 48 and 24 hours later and measured predictive performance in the validation group. The models' performance was evaluated using the area under the receiver operating characteristic curve (AUROC), and feature importance was assessed through SHapley Additive exPlanations values. Results: The study population exhibited an incidence of AKI of 6.6% in the development group and 5.4% in the validation group. The models demonstrated predictive performance with AUROCs of 0.864 and 0.886 for 48-hour predictions of AKI stages 1–3 and stages 2 and 3, and 0.904 and 0.903 for 24-hour predictions of AKI stages 1–3 and stages 2 and 3, respectively. Key features influencing model predictions included baseline and in-hospital serum creatinine values, injury severity score, age, lactate dehydrogenase, D-dimer, platelets, albumin, and C-reactive protein levels. Conclusions: The XGBoost models effectively predicted AKI in trauma patients up to 48 hours in advance using clinical data.

Backgrounds: Recently, alternative surrogate endpoints such as a 30% or 40% decline in estimated glomerular filtration rate (eGFR) or eGFR slope over 2 to 3 years have been proposed for predicting renal outcomes. However, the impact of GFR estimation methods on the accuracy and effectiveness of surrogate markers is unknown. Methods: We retrospectively enrolled participants in health screening programs at three hospitals from 1995 to 2009. We defined two different participant groups as YR1 and YR3, which had available 1-year or 3-year eGFR values along with their baseline eGFR levels. We compared the effectiveness of eGFR percentage change or slope to estimate end-stage renal disease (ESRD) risk according to two estimating equations (modified Modification of Diet in Renal Disease equation [eGFRm] and Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) equation [eGFRc]) for GFR. Results: In the YR1 and YR3 groups, 9,971 and 10,171 candidates were enrolled and ESRD incidence during follow-up was 0.26% and 0.19%, respectively. The eGFR percentage change was more effective than eGFR slope in estimating ESRD risk, regardless of the method of estimation. A 40% of decline in eGFR was better than 30%, and a 3-year baseline period was better than a 1-year period for prediction accuracy. Although some diagnostic indices from the CKD-EPI equation were better, we found no significant differences in the discriminative ability and hazard ratios for incident ESRD between eGFRc and eGFRm in either eGFR percentage change or eGFR slope. Conclusion: There were no significant differences in the prediction accuracy of GFR percentage change or eGFR slope between eGFRc and eGFRm in the general population.