Background: Whether GPT-4, the conversational artificial intelligence, can accurately diagnose and triage health conditions and whether it presents racial and ethnic biases in its decisions remain unclear. Objective: We aim to assess the accuracy of GPT-4 in the diagnosis and triage of health conditions and whether its performance varies by patient race and ethnicity. Methods: We compared the performance of GPT-4 and physicians, using 45 typical clinical vignettes, each with a correct diagnosis and triage level, in February and March 2023. For each of the 45 clinical vignettes, GPT-4 and 3 board-certified physicians provided the most likely primary diagnosis and triage level (emergency, nonemergency, or self-care). Independent reviewers evaluated the diagnoses as “correct” or “incorrect.” Physician diagnosis was defined as the consensus of the 3 physicians. We evaluated whether the performance of GPT-4 varies by patient race and ethnicity, by adding the information on patient race and ethnicity to the clinical vignettes. Results: The accuracy of diagnosis was comparable between GPT-4 and physicians (the percentage of correct diagnosis was 97.8% (44/45; 95% CI 88.2%-99.9%) for GPT-4 and 91.1% (41/45; 95% CI 78.8%-97.5%) for physicians; P=.38). GPT-4 provided appropriate reasoning for 97.8% (44/45) of the vignettes. The appropriateness of triage was comparable between GPT-4 and physicians (GPT-4: 30/45, 66.7%; 95% CI 51.0%-80.0%; physicians: 30/45, 66.7%; 95% CI 51.0%-80.0%; P=.99). The performance of GPT-4 in diagnosing health conditions did not vary among different races and ethnicities (Black, White, Asian, and Hispanic), with an accuracy of 100% (95% CI 78.2%-100%). P values, compared to the GPT-4 output without incorporating race and ethnicity information, were all .99. The accuracy of triage was not significantly different even if patients’ race and ethnicity information was added. The accuracy of triage was 62.2% (95% CI 46.5%-76.2%; P=.50) for Black patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for White patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for Asian patients, and 62.2% (95% CI 46.5%-76.2%; P=.69) for Hispanic patients. P values were calculated by comparing the outputs with and without conditioning on race and ethnicity. Conclusions: GPT-4’s ability to diagnose and triage typical clinical vignettes was comparable to that of board-certified physicians. The performance of GPT-4 did not vary by patient race and ethnicity. These findings should be informative for health systems looking to introduce conversational artificial intelligence to improve the efficiency of patient diagnosis and triage.

Nuclear accumulation of β-catenin is a key event for the development of colorectal cancer. Little is known, however, about the mechanisms underlying translocation of β-catenin from the cytoplasm or the membrane to the nucleus. The present study examined whether signal transducers and activators of transcription 3 (STAT3) activation is involved in the nuclear accumulation of β-catenin in colorectal cancer cells. Of the 90 primary colorectal cancer tissues, 40 (44.4%) were positive for nuclear staining of p-STAT3 and 63 (70.0%) were positive for nuclear staining of β-catenin. The nuclear staining of both p-STAT3 and β-catenin were observed predominantly in the periphery of the cancer tissues. Importantly, of the 40 tumors with p-STAT3 nuclear staining, 37 (92.5%) were also positive for nuclear β-catenin staining and there was a significant correlation between p-STAT3 and β-catenin nuclear staining (P < 0.01). Coexpression of nuclear p-STAT3 and β-catenin was associated with lower patient survival (P < 0.01). In an in vitro study using a human colon cancer cell line, SW480, inhibition of STAT3 by dominant negative STAT3 or the Janus kinase inhibitor, AG490, induced translocation of β-catenin from the nucleus to the cytoplasm or membrane. Luciferase assays revealed that STAT3 inhibition resulted in significant suppression of β-catenin/T-cell factor transcription in association with significant inhibition of cell proliferation (P < 0.05). These findings suggest that in colorectal cancer, STAT3 activation is involved in the nuclear accumulation of β-catenin, resulting in poor patient survival. (Cancer Res 2006; 66(6): 2913-7)

Introduction: Post-intensive care syndrome (PICS) is an emerging problem in critically ill patients and the prevalence and risk factors are unclear in patients with severe coronavirus disease 2019 (COVID-19). This multicenter prospective observational study aimed to investigate the prevalence and risk factors of PICS in ventilated patients with COVID-19 after ICU discharge. Methods: Questionnaires were administered twice in surviving patients with COVID-19 who had required mechanical ventilation, concerning Barthel Index, Short-Memory Questionnaire, and Hospital Anxiety and Depression Scale scores. The risk factors for PICS were examined using a multivariate logistic regression analysis. Results: The first and second PICS surveys were obtained at 5.5 and 13.5 months (mean) after ICU discharge, with 251 and 209 patients completing the questionnaires and with a prevalence of PICS of 58.6% and 60.8%, respectively, along with the highest percentages of cognitive impairment. Delirium (with an odds ratio of (OR) 2.34, 95% CI 1.1–4.9, and p = 0.03) and the duration of mechanical ventilation (with an OR of 1.29, 95% CI 1.05–1.58, and p = 0.02) were independently identified as the risk factors for PICS in the first PICS survey. Conclusion: Approximately 60% of the ventilated patients with COVID-19 experienced persistent PICS, especially delirium, and required longer mechanical ventilation.

Background: Post-intensive care syndrome (PICS) affects the quality of life (QOL) of survivors of critical illness. Although PICS persists for a long time, the longitudinal changes in each component and their interrelationships over time both remain unclear. This multicenter prospective study investigated the 2-year trajectory of PICS and its components as well as factors contributing to deterioration or recovery in mechanically ventilated patients with coronavirus disease 2019 (COVID-19), and also attempted to identify possible countermeasures. Methods: Patients who survived COVID-19 requiring mechanical ventilation completed questionnaires on the Barthel index, Short-Memory Questionnaire, Hospital Anxiety and Depression Scale, and EuroQol 5 dimensions 5-level every six months over a two-year period. Scores were weighted to account for dropouts, and the trajectory of each functional impairment was evaluated with alluvial diagrams. The prevalence of PICS and factors impairing or restoring function were examined using generalized estimating equations considering trajectories. Results: Among 334 patients, PICS prevalence rates in the four completed questionnaires were 72.1, 78.5, 77.6, and 82.0%, with cognitive impairment being the most common and lower QOL being noted when multiple impairments coexisted. Physical function and QOL indicated that many patients exhibited consistent trends of either recovery or deterioration. In contrast, cognitive function and mental health revealed considerable variability, with many patients showing fluctuating ratings in the later surveys. Delirium was associated with worse physical and mental health and poor QOL, while prolonged ventilation was associated with poor QOL. Living with family was associated with the recovery of all functions and QOL, while extracorporeal membrane oxygenation (ECMO) was associated with the recovery of cognitive function and mental health. Conclusions: Critically ill patients had PICS for a long period and followed different trajectories for each impairment component. Based on trajectories, known PICS risk factors such as prolonged ventilation and delirium were associated with impaired recovery, while ECMO and the presence of family were associated with recovery from PICS. In critically ill COVID-19 patients, delirium management and family interventions may play an important role in promoting recovery from PICS. Trial registration number: UMIN000041276, August 01, 2020.

We have developed a technique using laser-light scattering for monitoring and controlling precipitate formation during pulsed-laser deposition of ${\mathrm{YBa}}_2{\mathrm{Cu}}_3{\mathrm{O}}_{7\mathrm{-}\mathrm{\delta }}$ (YBCO) films. By laser ablating a stoichiometric YBCO target under optimum conditions for preparing high-crystallinity YBCO films, it has been almost inevitable to have precipitates on the film surface. The dominant phases of the precipitates analyzed by μ-Auger mapping were Y-deficient compounds (CuO and BaCu${\mathrm{O}}_2$). The progress of precipitate formation at the very beginning of the film growth could be monitored in situ by measuring the intensity of the light scattered to a nonspecular direction after being impinged obliquely on the film surface. The increase of scattered light intensity was verified to result mainly from the increase of the precipitate size from the qualitative analysis of ex situ atomic force microscope images and Rayleigh scattering theory. We have proposed that the precipitates were segregated from the highly crystalline 123 phase and their formation proceeded in two steps, i.e., nucleation and growth. Then, we could expect the precipitate segregation could be reduced either by a two-step process with the aid of optical scattering data to readjust the averaged film stoichiometry or preventing the nucleation to reduce the segregated precipitates. By switching the target to an yttrium-rich (50%) pellet $\left({\mathrm{Y}}_{1.5}{\mathrm{Ba}}_2{\mathrm{Cu}}_3{\mathrm{O}}_{7\mathrm{-}\mathrm{\delta }}\right)$ after the deposition of 10-unit-cell-thick YBCO having precipitates, a decrease in the scattered light intensity was observed to be consistent with the reduction of precipitate. This is attributed to the compensation of the yttrium deficiency and conversion of the Y-deficient precipitates into the YBCO phase by the reaction with Y-rich ablated species. Thus the measurement of light scattering is shown to be a useful tool not only for analyzing the surface reactions, but also for suppressing precipitate formation.