Construction and Application of a Model for Predicting the Risk of Delirium in Postoperative Patients With Type a Aortic Dissection

A nomogram-based model to predict postoperative transient neurological dysfunctions in patients receiving acute type A aortic dissection surgery

The purposes of this study were to develop and validate a nomogram for predicting postoperative transient neurological dysfunctions (TND) in patients with acute type A aortic dissection (AAAD) who underwent modified triple-branched stent graft implantation. This retrospective study developed a nomogram-based model in a consecutive cohort of 146 patients. Patient characteristics, preoperative clinical indices, and operative data were analyzed. Univariate and multivariable analyses were applied to identify the most useful predictive variables for constructing the nomogram. Discrimination and the calibration of the model was evaluated through the receiver operating characteristic curve (ROC), the Hosmer-Lemeshow goodness-of-fit test and the decision curve analysis (DCA). At the same time, to identify and compare long-term cumulative survival rate, Kaplan-Meier survival curve was plotted. The incidence rate of postoperative TND observed in our cohort were 40.9%. Supra-aortic dissection with or without thrombosis, creatinine >115 μmol and albumin <39.7 g/L, selective antegrade cerebral perfusion (SACP) time >7 min and total operation time >303 min, were confirmed as independent predictors that enhanced the likelihood of TND. Internal validation showed good discrimination of the model with under the ROC curve (AUC) of 0.818 and good calibration (Hosmer-Lemeshow test, p > .05). DCA revealed that the nomogram was clinically useful. In the long-term survival there was no significant difference between patients with or without TND history. The results showed the predict model based on readily available predictors has sufficient validity to identify TND risk in this population, that maybe useful for clinical decision-making.

Machine learning for prediction of delirium in patients with extensive burns after surgery

AIMS

Machine learning-based identification of key variables and prediction of postoperative delirium in patients with extensive burns.

METHODS

Five hundred and eighteen patients with extensive burns who underwent surgery were included and randomly divided into a training set, a validation set, and a testing set. Multifactorial logistic regression analysis was used to screen for significant variables. Nine prediction models were constructed in the training and validation sets (80% of dataset). The testing set (20% of dataset) was used to further evaluate the model. The area under the receiver operating curve (AUROC) was used to compare model performance. SHapley Additive exPlanations (SHAP) was used to interpret the best one and to externally validate it in another large tertiary hospital.

RESULTS

Seven variables were used in the development of nine prediction models: physical restraint, diabetes, sex, preoperative hemoglobin, acute physiological and chronic health assessment, time in the Burn Intensive Care Unit and total body surface area. Random Forest (RF) outperformed the other eight models in terms of predictive performance (ROC:84.00%) When external validation was performed, RF performed well (accuracy: 77.12%, sensitivity: 67.74% and specificity: 80.46%).

CONCLUSION

The first machine learning-based delirium prediction model for patients with extensive burns was successfully developed and validated. High-risk patients for delirium can be effectively identified and targeted interventions can be made to reduce the incidence of delirium.

Systematic review and meta-analysis of the risk factors for postoperative delirium in patients with acute type A aortic dissection

Background

Delirium is a common postoperative complication of acute type an aortic dissection, which is a serious threat to the patient's life after operation. However, there are many risk factors for delirium and there are different conclusions. The aim of this study was to systematically analyze the risk factors for postoperative delirium in patients with acute type a aortic dissection by means of meta-analysis.

Methods

Literature related to the risk factors of postoperative delirium in patients with acute type A aortic dissection was searched via the China National Knowledge Infrastructure (CNKI), cqvip.com (VIP), WanFang, PubMed, Willey Library, Embase, and Web of Science databases. Two persons independently conducted literature screening, data extraction and literature quality evaluation according to the inclusion and exclusion criteria. The quality of literature was evaluated with Newcastle-Ottawa Scale (NOS). R 4.2.1 was used to compare the risk factors for meta-analysis.

Results

After screening, 12 articles were included with a total of 2,511 cases, and 4 articles were at medium risk of bias and 8 articles were at low risk of bias. The meta-analysis results showed that patients in the delirium group had a higher probability of postoperative hypoxemia [odds ratio (OR) =1.65, 95% confidence interval (CI): 1.28-2.13, P<0.01], longer postoperative duration of ventilator assistance (OR =3.05, 95% CI: 2.47-3.77, P<0.01), higher incidence of renal insufficiency (OR =1.86, 95% CI: 1.33-2.58, P<0.01), lower hemoglobin levels (OR =0.33, 95% CI: 0.23-0.48, P<0.01), longer postoperative stay duration in the intensive care unit (ICU) (OR =2.25, 95% CI: 2.13-2.37, P<0.01), longer duration of hospitalization (OR =2.74, 95% CI: 2.37-3.16, P<0.01), and higher postoperative Acute Physiology and Chronic Health Evaluation II (APACHE II) scores (OR =1.01, 95% CI: 0.90-1.12, P=0.92).

Conclusions

Post-op aortic dissection in patients with acute type A diabetes should be monitored for oxygen and blood levels. When patients had prolonged mechanical ventilation, renal insufficiency, decreased hemoglobin, and prolonged ICU stay, timely intervention is needed to prevent the high-risk factors of postoperative delirium.

Effects of inflammation and oxidative stress on postoperative delirium in cardiac surgery

The past decade has witnessed unprecedented medical progress, which has translated into cardiac surgery being increasingly common and safe. However, complications such as postoperative delirium remain a major concern. Although the pathophysiological changes of delirium after cardiac surgery remain poorly understood, it is widely thought that inflammation and oxidative stress may be potential triggers of delirium. The development of delirium following cardiac surgery is associated with perioperative risk factors. Multiple interventions are being explored to prevent and treat delirium. Therefore, research on the potential role of biomarkers in delirium as well as identification of perioperative risk factors and pharmacological interventions are necessary to mitigate the development of delirium.

Automated machine learning-based model for the prediction of delirium in patients after surgery for degenerative spinal disease

OBJECTIVE

This study used machine learning algorithms to identify critical variables and predict postoperative delirium (POD) in patients with degenerative spinal disease.

METHODS

We included 663 patients who underwent surgery for degenerative spinal disease and received general anesthesia. The LASSO method was used to screen essential features associated with POD. Clinical characteristics, preoperative laboratory parameters, and intraoperative variables were reviewed and were used to construct nine machine learning models including a training set and validation set (80% of participants), and were then evaluated in the rest of the study sample (20% of participants). The area under the receiver-operating characteristic curve (AUROC) and Brier scores were used to compare the prediction performances of different models. The eXtreme Gradient Boosting algorithms (XGBOOST) model was used to predict POD. The SHapley Additive exPlanations (SHAP) package was used to interpret the XGBOOST model. Data of 49 patients were prospectively collected for model validation.

RESULTS

The XGBOOST model outperformed the other classifier models in the training set (area under the curve [AUC]: 92.8%, 95% confidence interval [CI]: 90.7%-95.0%), validation set (AUC: 87.0%, 95% CI: 80.7%-93.3%). This model also achieved the lowest Brier Score. Twelve vital variables, including age, serum albumin, the admission-to-surgery time interval, C-reactive protein level, hypertension, intraoperative blood loss, intraoperative minimum blood pressure, cardiovascular-cerebrovascular disease, smoking, alcohol consumption, pulmonary disease, and admission-intraoperative maximum blood pressure difference, were selected. The XGBOOST model performed well in the prospective cohort (accuracy: 85.71%).

CONCLUSION

A machine learning model and a web predictor for delirium after surgery for the degenerative spinal disease were successfully developed to demonstrate the extent of POD risk during the perioperative period, which could guide appropriate preventive measures for high-risk patients.

Development and validation of a nomogram to predict postoperative delirium in type B aortic dissection patients underwent thoracic endovascular aortic repair

Objective

Postoperative delirium (POD) is a common postoperative complication after cardiovascular surgery with adverse outcomes. No prediction tools are currently available for assessing POD in the type B aortic dissection (TBAD) population. The purposes of this study were to develop and validate a nomogram for predicting POD among TBAD patients who underwent thoracic endovascular aortic repair (TEVAR).

Methods

The retrospective cohort included 631 eligible TBAD patients who underwent TEVAR from January 2019 to July 2021. 434 patients included before 2021 were in the develop set; 197 others were in the independent validation set. Least absolute shrinkage and selection operator (LASSO) and logistic regression were applied to identify the most useful predictive variables for constructing the nomogram. Discrimination and the agreement of the model was assessed with the area under the receiver operating characteristic curve (AUC), Brier score and the Hosmer-Lemeshow goodness-of-fit test. The results were validated using a bootstrap resampling and the validation set.

Results

The incidence rate of POD observed in the development and validation cohort were 15.0% and 14.2%, respectively. Seven independent risk factors, including age ≥60 years, syncope or coma, postoperative blood transfusion, atelectasis, estimated glomerular filtration rate (eGFR) &lt;80 ml/min/1.73 m2, albumin &lt;30 g/L, and neutrophil to lymphocyte ratio, were included in the nomogram. The model showed a good discrimination with an AUC of 0.819 (95% CI, 0.762–0.876) in the developed set, and adjusted to 0.797 (95% CI, 0.735–0.849) and 0.791 (95% CI, 0.700–0.881) in the internal validation set and the external validation, respectively. Favorable calibration of the nomogram was confirmed in both the development and validation cohorts.

Conclusion

The nomogram based on seven readily available predictors has sufficient validity to identify POD risk in this population. This tool may facilitate targeted initiation of POD preventive intervention for healthcare providers.