The diagnostic value of the lumbar infusion test to predict symptomatic improvement after shunting for normal pressure hydrocephalus. A meta-analysis.

Abstract

The aim of this meta-analysis is to determine the diagnostic value of the Lumbar Infusion Test (LIT) to differentiate between patients suffering from normal pressure hydrocephalus who will benefit from CSF shunting and those who will not. A systematic search was performed in Ovid MEDLINE to identify RCTs or observational studies that evaluated LIT for predicting shunt responsiveness. Sensitivity and specificity values were pooled and Bayesian meta-analysis was performed. The Resistance to outflow (Rout) sensitivity was 76.9%; 81.6% and 36.6% for a Rout cut off of respectively 12, 14 and 18 mmHg/ml/min. Specificity rose with higher Rout cut off and was 34.0%; 37.2% and 78.0% for a cut off of respectively 12, 14 and 18 mmHg/ml/min. The negative predictive value was low for each cut off and was at most 33.3%. The LIT proves poor negative predictive value and appears an ineffective tool in the prediction of non-response to a shunting procedure. Therefore, in its present form and based on its current parameters, it cannot be used as a test to exclude patients from shunt implantation.