Dataset Information

Prediction of extubation outcome in mechanically ventilated patients: Development and validation of the Extubation Predictive Score (ExPreS).

ABSTRACT: Despite the best efforts of intensive care units (ICUs) professionals, the extubation failure rates in mechanically ventilated patients remain in the range of 5%-30%. Extubation failure is associated with increased risk of death and longer ICU stay. This study aimed to identify respiratory and non-respiratory parameters predictive of extubation outcome, and to use these predictors to develop and validate an "Extubation Predictive Score (ExPreS)" that could be used to predict likelihood of extubation success in patients receiving invasive mechanical ventilation (IMV). Derivation cohort was composed by patients aged ≥18 years admitted to the ICU and receiving IMV through an endotracheal tube for >24 hours. The weaning process followed the established ICU protocol. Clinical signs and ventilator parameters of patients were recorded during IMV, in the end phase of weaning in pressure support ventilation (PSV) mode, with inspiratory pressure of 7 cm H2O over the PEEP (positive end expiratory pressure). Patients who tolerated this ventilation were submitted to spontaneous breathing trial (SBT) with T-tube for 30 minutes. Those who passed the SBT and a subsequent cuff-leak test were extubated. The primary outcome of this study was extubation success at 48 hours. Parameters that showed statistically significant association with extubation outcome were further investigated using the receiver operating characteristics (ROC) analysis to assess their predictive value. The area under the curve (AUC) values were used to select parameters for inclusion in the ExPreS. Univariable logistic regression analysis and ROC analysis were performed to evaluate the performance of ExPreS. Patients' inclusion and statistical analyses for the prospective validation cohort followed the same criteria used for the derivation cohort and the decision to extubate was based on the ExPreS result. In the derivation cohort, a total of 110 patients were extubated: extubation succeeded in 101 (91.8%) patients and failed in 9 (8.2%) patients. Rapid shallow-breathing index (RSBI) in SBT, dynamic lung compliance, duration of IMV, muscle strength, estimated GCS, hematocrit, and serum creatinine were significantly associated with extubation outcome. These parameters, along with another parameter-presence of neurologic comorbidity-were used to create the ExPreS. The AUC value for the ExPreS was 0.875, which was higher than the AUCs of the individual parameters. The total ExPreS can range from 0 to 100. ExPreS ≥59 points indicated high probability of success (OR = 23.07), while ExPreS ≤44 points indicated low probability of success (OR = 0.82). In the prospective validation cohort, 83 patients were extubated: extubation succeeded in 81 (97.6%) patients and failed in 2 (2.4%) patients. The AUC value for the ExPreS in this cohort was 0.971. The multiparameter score that we propose, ExPreS, shows good accuracy to predict extubation outcome in patients receiving IMV in the ICU. In the prospective validation, the use of ExPreS decreased the extubation failure rate from 8.2% to 2.4%, even in a cohort of more severe patients.

SUBMITTER: Baptistella AR

PROVIDER: S-EPMC7971695 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundAssessment of preparedness of weaning has been recommended before extubation for mechanically ventilated patients. We aimed to understand the association of a structured assessment of weaning preparedness with successful liberation.MethodsWe retrospectively investigated patients with acute respiratory failure who experienced an extubation trial at the medical intensive care units of a medical center and compared the demographic and clinical characteristics between those patients with successful and failed extubation. A composite score to assess the preparedness of weaning, the WEANSNOW score, was generated consisting of eight components, including Weaning parameters, Endotracheal tube, Arterial blood gas analysis, Nutrition, Secretions, Neuromuscular-affecting agents, Obstructive airway problems and Wakefulness. The prognostic ability of the WEANSNOW score for extubation was then analyzed.ResultsOf the 205 patients included, 138 (67.3%) patients had successful extubation. Compared with the failure group, the success group had a significantly shorter duration of MV before the weaning attempt (11.2 ± 11.6 vs. 31.7 ± 26.2 days, p < 0.001), more with congestive heart failure (42.0% vs. 25.4%, p = 0.020), and had different distribution of the types of acute respiratory failure (p = 0.037). The failure group also had a higher WEANSNOW score (1.22 ± 0.85 vs. 0.51 ± 0.71, p < 0.001) and worse Rapid Shallow Breathing Index (93.9 ± 63.8 vs. 56.3 ± 35.1, p < 0.001). Multivariate logistic regression analysis showed that a WEANSNOW Score = 1 or higher (OR = 2.880 (95% CI [1.291-6.426]), p = 0.010) and intubation duration >21 days (OR = 7.752 (95% CI [3.560-16.879]), p < 0.001) were independently associated with an increased probability of extubation failure.ConclusionAssessing the pre-extubation status of intubated patients in a checklist-based approach using the WEANSNOW score might provide valuable insights into extubation failure in patients in a medical ICU for acute respiratory failure. Further prospective studies are warranted to elucidate the practice of assessing weaning preparedness.

Project description:BackgroundPersistent swallowing disorders (SD) are non-pulmonary complications of mechanical ventilation (MV). However, there are few clinical studies on persistent SD in critically ill patients undergoing tracheal intubation for MV. The aim of the present study was to assess the incidence and characteristics of clinical manifestations associated with persistent SD.MethodsWe prospectively evaluated in patients requiring more than 7 days of invasive MV the incidence and characteristics of clinical manifestations related to persistent SD. For this purpose, quality of swallowing was assessed within 24 h after extubation by an experienced physical therapist not directly involved in patient management. Swallowing assessment consisted in a specific standardized test combining a swallowing test and a full clinical evaluation of the cranial nerves involved in swallowing. In patients with SD on the first test, a second test was done within 48 h in order to discriminate between transient and persistent SD.ResultsAmong the 482 patients mechanically ventilated more than 7 days, 138 were enrolled in this study. The first test performed 24 h after extubation revealed SD in 35 patients (25%). According to the second test performed 48 h later, SD were considered transient in 21 (15%) and persistent in 14 (10%) cases. Patients with persistent SD were older (66 ± 16 vs 58 ± 15 years), had lower bodyweight at admission (76 ± 15 vs 87 ± 23 kg) and received less often neuromuscular blocking agents (36% vs 66%) compared to patients without or with only transient SD. Patients with persistent SD had longer duration of Intensive Care Unit (ICU) stay after first extubation and longer delay to oral feeding than patients without or with only transient SD, respectively, 11 ± 9 vs 7 ± 6 days and 23 ± 33 vs 5 ± 7 days.ConclusionsBased on a specific standardized clinical test, 25% of patients mechanically ventilated more than 7 days exhibited clinical manifestations of SD. However, SD were considered as persistent after extubation in only 10% of them. Persistent SD were associated with longer duration of ICU stay after extubation and longer time of enteral feeding.Trial registrationThe study is registered with Clinical Trials (NCT01360580).

Project description:ObjectivesTo evaluate the contribution of a preextubation chest X-ray (CXR) to identify the risk of extubation failure in mechanically ventilated patients.DesignRetrospective cohort study.SettingsICUs in a tertiary center (the Medical Information Mart for Intensive Care IV database).PatientsPatients greater than or equal to 18 years old who were mechanically ventilated and extubated after a spontaneous breathing trial.InterventionsNone.Measurements and main resultsAmong 1,066 mechanically ventilated patients, 132 patients (12%) experienced extubation failure, defined as reintubation or death within 48 hours of extubation. To predict extubation failure, we developed the following models based on deep learning (EfficientNet) and machine learning (LightGBM) with the training data: 1) model using only the rapid-shallow breathing index (RSBI), 2) model using RSBI and CXR, 3) model using all candidate clinical predictors (i.e., patient demographics, vital signs, laboratory values, and ventilator settings) other than CXR, and 4) model using all candidate clinical predictors with CXR. We compared the predictive abilities between models with the test data to investigate the predictive contribution of CXR. The predictive ability of the model using CXR as well as RSBI was not significantly higher than that of the model using only RSBI (c-statistics, 0.56 vs 0.56; p = 0.95). The predictive ability of the model using clinical predictors with CXR was not significantly higher than that of the model using all clinical predictors other than CXR (c-statistics, 0.71 vs 0.70; p = 0.12). Based on SHapley Additive exPlanations values to interpret the model using all clinical predictors with CXR, CXR was less likely to contribute to the predictive ability than other predictors (e.g., duration of mechanical ventilation, inability to follow commands, and heart rate).ConclusionsAdding CXR to a set of other clinical predictors in our prediction model did not significantly improve the predictive ability of extubation failure in mechanically ventilated patients.

Project description:BackgroundSeveral studies have assessed predictors of weaning and extubation outcome in short-term mechanically ventilated patients, but there are only few studies on predictors of weaning from prolonged mechanical ventilation.MethodsRetrospective, single-center, observational study at a specialized national weaning center in Germany. Patients' medical records were reviewed to obtain data on demographics, comorbidities, respiratory indices, and the result of a prospectively documented, standardized spontaneous breathing trial (SBT) upon admission to the weaning center. Respiratory indices assessed were the ventilatory ratio (VR) and parameters derived from calculated mechanical power (MP). Predictors associated with failure of prolonged weaning and failure of the SBT were assessed using a binary logistic regression model.ResultsA total of 263 prolonged mechanically ventilated, tracheotomized patients, treated over a 5-year period were analyzed. After 3 weeks of mechanical ventilation, patients with unsuccessful weaning failed a SBT more frequently and showed significantly increased values for inspiratory positive airway pressure, driving pressure, VR, absolute MP, and MP normalized to predicted body weight and dynamic lung-thorax compliance (LTC-MP). In the logistic regression analyses, variables independently correlated with weaning failure were female gender (adjusted odds ratio 0.532 [95% CI 0.291-0.973]; p = 0.040), obesity (body mass index ≥ 30 kg/m2) (2.595 [1.210-5.562]; p = 0.014), COPD (3.209 [1.563-6.589]; p = 0.002), LTC-MP (3.470 [1.067-11.284]; p = 0.039), PaCO2 on mechanical ventilation (1.101 [95% CI 1.034-1.173]; p = 0.003), and failure of the SBT (4.702 [2.250-9.825]; p < 0.001). In addition, female gender (0.401 [0.216-0.745]; p = 0.004), LTC-MP (3.017 [1.027-8.862]; p = 0.046), and PaCO2 on mechanical ventilation (1.157 [1.083-1.235]; p < 0.001) were independent risk factors for an unsuccessful SBT.ConclusionsIn the present study, the derived predictors of weaning point to a crucial role of the workload imposed on respiratory muscles during spontaneous breathing. Mechanical power normalized to lung-thorax compliance was independently correlated with weaning outcome and may identify patients at high risk for weaning failure.

Project description:BackgroundMechanical ventilation (MV) is vital for critically ill ICU patients but carries significant mortality risks. This study aims to develop a predictive model to estimate hospital mortality among MV patients, utilizing comprehensive health data to assist ICU physicians with early-stage alerts.MethodsWe developed a Machine Learning (ML) framework to predict hospital mortality in ICU patients receiving MV. Using the MIMIC-III database, we identified 25,202 eligible patients through ICD-9 codes. We employed backward elimination and the Lasso method, selecting 32 features based on clinical insights and literature. Data preprocessing included eliminating columns with over 90% missing data and using mean imputation for the remaining missing values. To address class imbalance, we used the Synthetic Minority Over-sampling Technique (SMOTE). We evaluated several ML models, including CatBoost, XGBoost, Decision Tree, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Logistic Regression, using a 70/30 train-test split. The CatBoost model was chosen for its superior performance in terms of accuracy, precision, recall, F1-score, AUROC metrics, and calibration plots.ResultsThe study involved a cohort of 25,202 patients on MV. The CatBoost model attained an AUROC of 0.862, an increase from an initial AUROC of 0.821, which was the best reported in the literature. It also demonstrated an accuracy of 0.789, an F1-score of 0.747, and better calibration, outperforming other models. These improvements are due to systematic feature selection and the robust gradient boosting architecture of CatBoost.ConclusionThe preprocessing methodology significantly reduced the number of relevant features, simplifying computational processes, and identified critical features previously overlooked. Integrating these features and tuning the parameters, our model demonstrated strong generalization to unseen data. This highlights the potential of ML as a crucial tool in ICUs, enhancing resource allocation and providing more personalized interventions for MV patients.

Dataset Information

Prediction of extubation outcome in mechanically ventilated patients: Development and validation of the Extubation Predictive Score (ExPreS).

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets