How to select patients with normal pressure hydrocephalus for shunting

The objective was to compare predictive values of clinical and CT findings, co-existing cerebrovascular disease (CVD) and CSF outflow resistance (Rcsf) for outcome of shunting in NPH. A group of 95 NPH patients was shunted and followed for one year. Gait disturbance and dementia were quantified by an NPH scale (NPHS) and handicap by the modified Rankin scale (MRS). Improvement was defined as a change of at least 15% in NPHS and one grade in MRS at last follow-up. Clinical and CT findings at entry were classified as typical or not typical for NPH. CVD was defined as a history of stroke or CT-scans showing infarcts or moderate to severe white matter hypodense lesions. Clinical and CT findings typical for NPH, absence of CVD and Rcsf > 18 mmHg/ml/min were positive tests and the reciprocal outcomes negative tests. Typical clinical and CT findings were found in 69% and 68%, CVD (history of stroke n = 14, infarcts on CT n = 13, leucoaraiosis n = 32) in 47% and Rcsf > 18 in 38% of patients. The ratio of patients classified as improved in both scales was significantly greater for those with positive than negative tests. Mean improvement differed the most between patients with and without CVD. Using logistic regression analysis Rcsf > 18 was the only significant predictor of improvement in NPHS (OR 4.4, 95% CI 1.3-16.7) and typical CT findings in MRS (OR 5.6, 95% CI 1.8-17.9). We conclude that CVD is an important predictor of poor outcome. The best strategy is to shunt NPH patients if Rcsf is > 18 mmHg/ml/min or, when Rcsf is lower, if CT findings are typical for NPH and there is no or limited CVD.