A Comparison of Five Algorithmic Methods and Machine Learning Pattern Recognition for Artifact Detection in Electronic Records of Five Different Vital Signs: A Retrospective Analysis.

BACKGROUND: Research on electronic health record physiologic data is common, invariably including artifacts. Traditionally, these artifacts have been handled using simple filter techniques. The authors hypothesized that different artifact detection algorithms, including machine learning, may be necessary to provide optimal performance for various vital signs and clinical contexts. METHODS: In a retrospective single-center study, intraoperative operating room and intensive care unit (ICU) electronic health record datasets including heart rate, oxygen saturation, blood pressure, temperature, and capnometry were included. All records were screened for artifacts by at least two human experts. Classical artifact detection methods (cutoff, multiples of SD [z-value], interquartile range, and local outlier factor) and a supervised learning model implementing long short-term memory neural networks were tested for each vital sign against the human expert reference dataset. For each artifact detection algorithm, sensitivity and specificity were calculated. RESULTS: A total of 106 (53 operating room and 53 ICU) patients were randomly selected, resulting in 392,808 data points. Human experts annotated 5,167 (1.3%) data points as artifacts. The artifact detection algorithms demonstrated large variations in performance. The specificity was above 90% for all detection methods and all vital signs. The neural network showed significantly higher sensitivities than the classic methods for heart rate (ICU, 33.6%; 95% CI, 33.1 to 44.6), systolic invasive blood pressure (in both the operating room [62.2%; 95% CI, 57.5 to 71.9] and the ICU [60.7%; 95% CI, 57.3 to 71.8]), and temperature in the operating room (76.1%; 95% CI, 63.6 to 89.7). The CI for specificity overlapped for all methods. Generally, sensitivity was low, with only the z-value for oxygen saturation in the operating room reaching 88.9%. All other sensitivities were less than 80%. CONCLUSIONS: No single artifact detection method consistently performed well across different vital signs and clinical settings. Neural networks may be a promising artifact detection method for specific vital signs.

Effect of perioperative levosimendan administration on postoperative N-terminal pro-B-type natriuretic peptide concentration in patients with increased cardiovascular risk factors undergoing non-cardiac surgery: protocol for the double-blind, randomised, placebo-controlled IMPROVE trial.

INTRODUCTION: Elevated N-terminal pro-brain natriuretic peptide (NT-pro-BNP) after non-cardiac surgery is a strong predictor for cardiovascular complications and reflects increased myocardial strain. NT-pro-BNP concentrations significantly rise after non-cardiac surgery within the first 3 days. Levosimendan is a potent inotropic drug that increases calcium sensitivity to cardiac myocytes, which results in improved cardiac contractility that last for approximately 7 days. Thus, we will test the effect of a pre-emptive perioperative administration of levosimendan on postoperative NT-pro-BNP concentration as compared with the administration of a placebo in patients undergoing moderate-risk to high-risk major abdominal surgery. METHODS AND ANALYSIS: We will conduct a double-blinded prospective randomised trial at the Medical University of Vienna, Vienna, Austria (and potentially a second centre in Germany), including 230 patients at-risk for cardiovascular complications undergoing moderate- to high-risk major abdominal surgery. Patients will be randomly assigned to receive a single dose of 12.5 mg levosimendan versus placebo after induction of anaesthesia. The primary outcome will be the postoperative maximum NT-pro-BNP concentration between both group within the first three postoperative days. Our secondary outcomes will be the incidence of myocardial ischaemia, myocardial injury after non-cardiac surgery and a composite of myocardial infarction and death within 30 days and 1 year after surgery between both groups. Our further secondary outcome will be stratification of NT-pro-BNP values according to previously thresholds to predict mortality of myocardial infarction after surgery. ETHICS AND DISSEMINATION: The study was approved by the Ethics Committee of the Medical University of Vienna on 14 July 2020 (EK 2187/2019). Written informed consent will be obtained from all patients a day before surgery. Results of this study will be submitted for publication in a peer-reviewed journal. TRIAL REGISTRATION NUMBER: NCT04329624.