RESUMEN
Background: Hypoxia is an important risk factor and indicator for the declining health of inpatients. Predicting future hypoxic events using machine learning is a prospective area of study to facilitate time-critical interventions to counter patient health deterioration. Objective: This systematic review aims to summarize and compare previous efforts to predict hypoxic events in the hospital setting using machine learning with respect to their methodology, predictive performance, and assessed population. Methods: A systematic literature search was performed using Web of Science, Ovid with Embase and MEDLINE, and Google Scholar. Studies that investigated hypoxia or hypoxemia of hospitalized patients using machine learning models were considered. Risk of bias was assessed using the Prediction Model Risk of Bias Assessment Tool. Results: After screening, a total of 12 papers were eligible for analysis, from which 32 models were extracted. The included studies showed a variety of population, methodology, and outcome definition. Comparability was further limited due to unclear or high risk of bias for most studies (10/12, 83%). The overall predictive performance ranged from moderate to high. Based on classification metrics, deep learning models performed similar to or outperformed conventional machine learning models within the same studies. Models using only prior peripheral oxygen saturation as a clinical variable showed better performance than models based on multiple variables, with most of these studies (2/3, 67%) using a long short-term memory algorithm. Conclusions: Machine learning models provide the potential to accurately predict the occurrence of hypoxic events based on retrospective data. The heterogeneity of the studies and limited generalizability of their results highlight the need for further validation studies to assess their predictive performance.
RESUMEN
BACKGROUND AND PURPOSE: Anesthesia regimen in patients undergoing mechanical thrombectomy (MT) is still an unresolved issue. METHODS: We compared the effect of anesthesia regimen using data from the German Stroke Registry-Endovascular Treatment (GSR-ET) between June 2015 and December 2019. Degree of disability was rated by the modified Rankin Scale (mRS), and good outcome was defined as mRS 0-2. Successful reperfusion was assumed when the modified thrombolysis in cerebral infarction scale was 2b-3. RESULTS: Out of 6,635 patients, 67.1% (n=4,453) patients underwent general anesthesia (GA), 24.9% (n=1,650) conscious sedation (CS), and 3.3% (n=219) conversion from CS to GA. Rate of successful reperfusion was similar across all three groups (83.0% vs. 84.2% vs. 82.6%, P=0.149). Compared to the CA-group, the GA-group had a delay from admission to groin (71.0 minutes vs. 61.0 minutes, P<0.001), but a comparable interval from groin to flow restoration (41.0 minutes vs. 39.0 minutes). The CS-group had the lowest rate of periprocedural complications (15.0% vs. 21.0% vs. 28.3%, P<0.001). The CS-group was more likely to have a good outcome at follow-up (42.1% vs. 34.2% vs. 33.5%, P<0.001) and a lower mortality rate (23.4% vs. 34.2% vs. 26.0%, P<0.001). In multivariable analysis, GA was associated with reduced achievement of good functional outcome (odds ratio [OR], 0.82; 95% confidence interval [CI], 0.71 to 0.94; P=0.004) and increased mortality (OR, 1.42; 95% CI, 1.23 to 1.64; P<0.001). Subgroup analysis for anterior circulation strokes (n=5,808) showed comparable results. CONCLUSIONS: We provide further evidence that CS during MT has advantages over GA in terms of complications, time intervals, and functional outcome.
