Knepper, Camille

Objective: To evaluate the feasibility and effectiveness of wrist-worn wearable devices combined with machine learning (ML) approaches for detecting a diverse array of seizure types beyond generalized tonic–clonic (GTC), including focal, generalized, and subclinical seizures. Materials and Methods: Twenty-eight patients undergoing inpatient video-EEG monitoring at Mayo Clinic were concurrently monitored using Empatica E4 wrist-worn devices. These devices captured accelerometry, blood volume pulse, electrodermal activity, skin temperature, and heart rate. Seizures were annotated by neurologists. The data were preprocessed to experiment with various segment lengths (10 s and 60 s) and multiple feature sets. Three ML strategies, XGBoost, deep learning models (LSTM, CNN, Transformer), and ROCKET, were evaluated using leave-one-patient-out cross-validation. Performance was assessed using area under the receiver operating characteristic curve (AUROC), seizure-wise recall (SW-Recall), and false alarms per hour (FA/h). Results: Detection performance varied by seizure type and model. GTC seizures were detected most reliably (AUROC = 0.86, SW-Recall = 0.81, FA/h = 3.03). Hyperkinetic and tonic seizures showed high SW-Recall but also high FA/h. Subclinical and aware-dyscognitive seizures exhibited the lowest SW-Recall and highest FA/h. MultiROCKET and XGBoost performed best overall, though no single model was optimal for all seizure types. Longer segments (60 s) generally reduced FA/h. Feature set effectiveness varied, with multi-biosignal sets improving performance across seizure types. Conclusions: Wrist-worn wearables combined with ML can extend seizure detection beyond GTC seizures, though performance remains limited for non-motor types. Optimizing model selection, feature sets, and segment lengths, and minimizing false alarms, are key to clinical utility and real-world adoption.

Background The COVID-19 pandemic introduced unique challenges to healthcare systems, particularly in relation to patient safety and adverse events during hospitalization. There is limited understanding of COVID-19’s association with some patient safety indicators (PSIs). Objectives This study aimed to investigate the impact of COVID-19 infection on the rate of PSI-3 events and its implications on quality metrics. We compared PSI-3 event rates between COVID-19–infected and uninfected patients and examined the clinical characteristics of COVID-19 patients experiencing PSI-3 events. Methods This is a retrospective study at Mayo Clinic hospitals between January 2020 and February 2022, analyzing patients meeting PSI-3 denominator eligibility criteria. PSI-3 events were identified using AHRQ WinQI software. Patients were categorized based on COVID-19 status. Patient demographics, characteristics, and PSI-3 rates were compared. A case series analysis described clinical details of COVID-19 patients with PSI-3 events. Results Of 126,781 encounters meeting PSI-3 criteria, 8674 (6.8%) had acute COVID-19 infection. COVID-19–infected patients were older, more likely to be male, non-white, and had private insurance. PSI-3 rates were significantly higher in COVID-19 patients (0.21% versus 0.06%, P < 0.0001), even after risk adjustment (adjusted risk ratio, 3.24, P < 0.0001). The case series of 17 COVID-19 patients with PSI-3 events showed distinctive clinical characteristics, including higher medical device–related pressure injuries, and greater predisposition for head, face, and neck region. Conclusions Acute COVID-19 infection correlates with higher PSI-3 event rates. Current quality indicators may require adaptation to address the pandemic’s complexities and impact on patient safety. Further research is needed to comprehensively understand the intricate relationship between COVID-19 and patient outcomes.