Journal of clinical monitoring and computing
-
J Clin Monit Comput · Oct 2022
Multicenter StudyPredicting hypoglycemia in critically Ill patients using machine learning and electronic health records.
Hypoglycemia is a common occurrence in critically ill patients and is associated with significant mortality and morbidity. We developed a machine learning model to predict hypoglycemia by using a multicenter intensive care unit (ICU) electronic health record dataset. Machine learning algorithms were trained and tested on patient data from the publicly available eICU Collaborative Research Database. ⋯ The best-performing predictive model was the eXtreme gradient boosting model (XGBoost), which achieved an area under the received operating curve (AUROC) of 0.85, a sensitivity of 0.76, and a specificity of 0.76. The machine learning model developed has strong discrimination and calibration for the prediction of hypoglycemia in ICU patients. Prospective trials of these models are required to evaluate their clinical utility in averting hypoglycemia within critically ill patient populations.
-
J Clin Monit Comput · Oct 2022
Implementing a Rapid Response System in a tertiary-care hospital. A cost-effectiveness study.
The occurrence of adverse events (AE) in hospitalized patients substancially increases the risk of disability or death, having a major negative clinical and economic impact on public health. For early identification of patients at risk and to establish preventive measures, different healthcare systems have implemented rapid response systems (RRS). The aim of this study was to carry out a cost-effectiveness analysis of implementing a RRS in a tertiary-care hospital. ⋯ The present analysis shows the RRS as a dominant, less costly and more effective structure compared to the non-RRS.
-
J Clin Monit Comput · Oct 2022
A capaciflector provides continuous and accurate respiratory rate monitoring for patients at rest and during exercise.
Respiratory rate (RR) is a marker of critical illness, but during hospital care, RR is often inaccurately measured. The capaciflector is a novel sensor that is small, inexpensive, and flexible, thus it has the potential to provide a single-use, real-time RR monitoring device. We evaluated the accuracy of continuous RR measurements by capaciflector hardware both at rest and during exercise. ⋯ Accuracy and continuity of monitoring were upheld even during vigorous CPET exercise, often with narrower limits of agreement than those reported for comparable technologies. We provide a unique clinical demonstration of the capaciflector as an accurate breathing monitor, which may have the potential to become a simple and affordable medical device. Clinical trial number: NCT03832205 https://clinicaltrials.gov/ct2/show/NCT03832205 registered February 6th, 2019.
-
J Clin Monit Comput · Oct 2022
Processed intraoperative burst suppression and postoperative cognitive dysfunction in a cohort of older noncardiac surgery patients.
Postoperative cognitive dysfunction (POCD) is a decline in cognitive test performance which persists months after surgery. There has been great interest in the anesthesia community regarding whether variables generated by commercially available processed EEG monitors originally marketed to prevent awareness under anesthesia can be used to guide intraoperative anesthetic management to prevent POCD. Processed EEG monitors represent an opportunity for anesthesiologists to directly monitor the brain even if they have not been trained to interpret EEG waveforms. ⋯ Our finding may be a limitation of the monitor's ability to detect burst suppression. The consistent trend towards more intraoperative burst suppression in patients who developed POCD suggests that future studies are needed to investigate the relationship of raw intraoperative burst suppression and POCD. Trial registry Clinical trial number and registry URL: Optimizing Postoperative Cognitive Dysfunction in the Elderly-PRESERVE, Clinical Trials Gov# NCT02650687; https://clinicaltrials.gov/ct2/show/NCT02650687 .