Abstract

Background: The unavailability of standardized treatment algorithms to properly select tuberculous meningitis (TBM) patients for surgery delays optimal management.

Aim and Objective: This study developed a mathematical model as an objective guide in deciding whether to perform surgical intervention or not among patients with hydrocephalus from TBM.

Materials and Methods: Using 13 different clinical and cerebrospinal fluid parameters, a saturated logistic regression model was developed and tried retrospectively on 37 previously managed cases of hydrocephalus from TBM.

Results: The model is statistically significant (P = 0.0170) with 93.2% concordance index in predicting physician’s decision on admission whether the patient will undergo surgical intervention or not. Patients are recommended for surgical intervention if g-score >0.04 with 90.48% sensitivity.

Conclusion: With its significant P-value and high concordance index, the model is a good representative of the actual decision making of physicians in the management of hydrocephalus from TBM.

Key words: Hydrocephalus; Tuberculous Meningitis; Shunt; Surgery; Mathematical Model; Logistic Regression