Prediction of Undesired Obstruction in External Ventricular Drains.

HIV severity will be estimated using an adaptation of the existing Graeb's score: the Graeb-EVD score. EVD obstruction is defined by an abnormally reduced drainage flow, requiring intervention to restore normal function (cases of EVD malposition will be excluded).The predictive performance of the Graeb-EVD score will be evaluated by the area under the receiver operating characteristic curve (AUC-ROC). The AUC-ROC ranges from 0 to 1. The predictive ability will be considered acceptable if associated with an AUC-ROC greater than 0.70, good if greater than 0.80, and excellent if greater than 0.90.

The researcher will study the risk factors for the occurrence of at least one episode of unplanned EVD obstruction (dependent variable in a logistic regression model) by forcing the inclusion of HIV into this model as an explanatory variable. If the 95% confidence interval of the adjusted odds ratio associated with HIV is greater than 1, HIV will be considered an independent risk factor for unplanned EVD obstruction.

A p-value of <0.05 in the pairwise comparison of areas under the receiver operating characteristics curves will be used to assert the superiority of the performance of one score over another.

HIV severity will be estimated using an adaptation of the existing Graeb's score: the Graeb-EVD score. EVD obstruction is defined by an abnormally reduced drainage flow, requiring intervention to restore normal function (cases of EVD malposition will be excluded). The predictive performance of the Graeb-EVD score will be evaluated by the area under the receiver operating characteristic curve (AUC-ROC). The AUC-ROC ranges from 0 to 1. The predictive ability will be considered acceptable if associated with an AUC-ROC greater than 0.70, good if greater than 0.80, and excellent if greater than 0.90.

The predictive performance of the tested scores will be evaluated by the AUC-ROC. The predictive capacity will be considered acceptable if associated with an AUC-ROC greater than 0.70, good if greater than 0.80, and excellent if greater than 0.90.

The predictive performance of the tested scores will be evaluated by the AUC-ROC. The predictive capacity will be considered acceptable if associated with an AUC-ROC greater than 0.70, good if greater than 0.80, and excellent if greater than 0.90.

Functional outcome will be considered unfavorable if the modified Rankin scale (ranging from 0 to 6) is 2 or higher, and favorable otherwise. The predictive performance of the tested scores will be evaluated by the AUC-ROC. A score will be considered useful if it is associated with an AUC-ROC significantly greater than 0.50.

Functional outcome will be considered unfavorable if the modified Rankin scale is 2 or higher, and favorable otherwise. A multivariate analysis involving risk factors for poor functional outcome (dependent variable) and incorporating the occurrence of at least one episode of unplanned EVD obstruction (explanatory variable) will be conducted. If the 95% confidence interval does not include the value 1, then the variable will be retained as an independent risk factor for the occurrence of the dependent variable. As an explanatory variable, we will test, in separate models, the occurrence of at least one episode of unplanned obstruction and the occurrence of at least two episodes.

Neuromeningeal infection is defined as the administration of antimicrobial therapy for this specific indication for at least 10 days. The predictive performance will be assessed through the area under the receiver operating characteristics curve.

A multivariate analysis involving risk factors for the need for internalization of the ventricular shunt (dependent variable) and incorporating the occurrence of at least one episode of unplanned EVD obstruction (explanatory variable) will be conducted. If the 95% confidence interval does not include the value 1, then the variable will be retained as an independent risk factor for the occurrence of the dependent variable.