Is Virtual Learning Here to Stay? A Multispecialty Survey of Residents, Fellows, and Faculty.

The transition to virtual learning during the coronavirus disease 2019 pandemic marks a paradigm shift in graduate medical education (GME). From June to September 2021, we conducted a dual-center, multispecialty survey of residents, fellows, and faculty members to determine overall perceptions about virtual learning and assess its benefits, drawbacks, and future role in GME. We discovered a mainly positive view of virtual education among trainees (138/207, 0.67, 95% CI 0.59-0.73) and faculty (180/278, 0.65, 0.59-0.70). Large group sessions, such as didactic lectures, grand rounds, and national conferences, were ranked best-suited for the virtual environment, whereas small groups and procedural training were the lowest ranked. Major benefits and drawbacks to virtual learning was identified. A hybrid approach, combining in-person and virtual sessions, was the preferred format among trainees (167/207, 0.81, 0.75-0.86) and faculty (229/278, 0.82, 0.77-0.87). Virtual learning offers a valuable educational experience that should be retained in postpandemic GME curriculums.

Novel Machine Learning Approach to Predict and Personalize Length of Stay for Patients Admitted with Syncope from the Emergency Department.

Background: Syncope, a common problem encountered in the emergency department (ED), has a multitude of causes ranging from benign to life-threatening. Hospitalization may be required, but the management can vary substantially depending on specific clinical characteristics. Models predicting admission and hospitalization length of stay (LoS) are lacking. The purpose of this study was to design an effective, exploratory model using machine learning (ML) technology to predict LoS for patients presenting with syncope. Methods: This was a retrospective analysis using over 4 million patients from the National Emergency Department Sample (NEDS) database presenting to the ED with syncope between 2016−2019. A multilayer perceptron neural network with one hidden layer was trained and validated on this data set. Results: Receiver Operator Characteristics (ROC) were determined for each of the five ANN models with varying cutoffs for LoS. A fair area under the curve (AUC of 0.78) to good (AUC of 0.88) prediction performance was achieved based on sequential analysis at different cutoff points, starting from the same day discharge and ending at the longest analyzed cutoff LoS ≤7 days versus >7 days, accordingly. The ML algorithm showed significant sensitivity and specificity in predicting short (≤48 h) versus long (>48 h) LoS, with an AUC of 0.81. Conclusions: Using variables available to triaging ED clinicians, ML shows promise in predicting hospital LoS with fair to good performance for patients presenting with syncope.

Timing variability and midfrontal ~4 Hz rhythms correlate with cognition in Parkinson's disease.

Patients with Parkinson's disease (PD) can have significant cognitive dysfunction; however, the mechanisms for these cognitive symptoms are unknown. Here, we used scalp electroencephalography (EEG) to investigate the cortical basis for PD-related cognitive impairments during interval timing, which requires participants to estimate temporal intervals of several seconds. Time estimation is an ideal task demand for investigating cognition in PD because it is simple, requires medial frontal cortical areas, and recruits basic executive processes such as working memory and attention. However, interval timing has never been systematically studied in PD patients with cognitive impairments. We report three main findings. First, 71 PD patients had increased temporal variability compared to 37 demographically matched controls, and this variability correlated with cognitive dysfunction as measured by the Montreal Cognitive Assessment (MOCA). Second, PD patients had attenuated ~4 Hz EEG oscillatory activity at midfrontal electrodes in response to the interval-onset cue, which was also predictive of MOCA. Finally, trial-by-trial linear mixed-effects modeling demonstrated that cue-triggered ~4 Hz power predicted subsequent temporal estimates as a function of PD and MOCA. Our data suggest that impaired cue-evoked midfrontal ~4 Hz activity predicts increased timing variability that is indicative of cognitive dysfunction in PD. These findings link PD-related cognitive dysfunction with cortical mechanisms of cognitive control, which could advance novel biomarkers and neuromodulation for PD.