Speed of clinical response to external lumbar drainage as a predictor of outcome of ventriculoperitoneal shunt in normal pressure hydrocephalus: A prospective cohort study

Background: External lumbar drainage (ELD) is definitely a known predictor of ventriculoperitoneal shunt (VPS) outcome in normal pressure hydrocephalus (NPH). However, the speed of clinical response following ELD has not been studied well enough to predict shunt outcome. This study aims to assess the predictive value of ELD response for VP shunt outcome. Methods: A prospective observational study of 67 patients diagnosed with NPH based on clinical and radiological criteria. All patients underwent ELD, and the time to improvement in gait after ELD was calculated. Those who showed improvement within 72 h were defined as early responders, those who showed improvement after 72 h as delayed responders, and those who did not show improvement as nonresponders. Early and delayed responders were given VPS surgery. Postsurgery results were assessed clinically on follow-up. Both multivariate and univariate penalized Firth regression was conducted to determine the predictive value of the speed of clinical response on the outcome of VPS. A receiver operating characteristic (ROC) curve analysis was done to assess the discriminative power of the model. A Hosmer–Lemeshow test was done to see the calibration. Results: Clinical response following ELD was a significant predictor of VP shunt outcome ( P < 0.001). Univariate Firth regression analysis demonstrated that delayed responders had significantly higher odds of poor shunt outcome as compared with early responders (odds ratio = 96.2). ROC analysis demonstrated good discriminative ability for the shunt outcome. Hosmer–Lemeshow showed good calibration ( P = 0.832). Conclusion: Time to gait improvement following ELD may aid patient selection and surgical decision-making among patients undergoing VPS after ELD responsiveness.