Reliability of Third Ventricle Assessment by Transcranial Ultrasound: A Computational Model of the Effect of Insonation Angle.

INTRODUCTION: Transcranial sonography (TCS) is a bedside examination which is currently used in multiple neurocritical care settings. Third ventricle ultrasound is usually a simple technique, though a large insonation angle could lead to an overestimation of third ventricular diameter. The aim of this study was to use a mathematical model to evaluate the impact of probe inclination on the false positive rate when using TCS to evaluate third ventricle enlargement. METHODS: Using R software, we simulated a pool of 100,000 fictitious patients with a normal third ventricle size (diameter from 0 to 9 mm) in daily follow-up for ventricle enlargement for 30 consecutive days using TCS. Each day, a different, random insonation angle (α) was generated and a corresponding measured diameter calculated as: measured diameter=real diameter/cos α. If the measured diameter was >9.0 mm, the simulation registered a "misdiagnosis" episode and the simulation loop was interrupted; otherwise, the simulation continued to its thirtieth iteration. RESULTS: Of the 100,000 "patient" simulations, 30,905 (30.9%) had an erroneous TCS diagnosis of ventricular enlargement. Angles of insonation >35 degrees contributed to 79.3% of the total misdiagnoses of ventricular enlargement (false positive rate, 3.71%), whereas misdiagnosis was rare when the insonation angle was ≤15 degrees (1.30% of the total misdiagnoses; false positive rate, 0.06%). CONCLUSION: Using probe inclinations <15 degrees, erroneous diagnosis of third ventricular enlargement was rare. Our results suggest that TCS has a low rate of false positives when the angle of insonation is minimized.

Optimizing post anesthesia care unit admission after elective craniotomy for brain tumors: a cohort study.

BACKGROUND: Postoperative admission to intensive care unit (ICU) after craniotomy for brain tumor was the routine in the past years. However, there is little evidence supporting this dogma and doubts have been casted by many authors in the last years. Our aim was to identify risk factors for ICU admission after elective brain tumor surgery in order to propose an individualized admission to ICU tailored on patient needs. METHODS: We conducted a retrospective cohort study including all patients undergoing elective surgery for brain tumor in a neurosurgical post anesthesia care unit of a university hospital over a period of 6 years. In order to identify and validate risk factors for ICU admission, we split the final cohort of patients in a training cohort (two/third of the cohort) and the validation cohort (one/third of the cohort) using a random sequence. Using univariate and multivariate logistic regression, we created a scoring system in the training cohort and tested it with the validation cohort. Moreover, we perform a sensitivity analysis on the overall population. RESULTS: A total of 420 patients were eligible for this study. ASA-PS, tumor volume, and surgery length entered the scoring system. Sensitivity analysis on the overall population for the scoring system had an AUC of 0.774 (95% CI 0.668-0.880, the best threshold at 12.5) CONCLUSIONS: We created a tool based on ASA-PS, length of surgery, and tumor volume to evaluate the risk for ICU admission after supratentorial tumor resection. Prospective studies are deemed necessary to validate our tool.