Shunting for Idiopathic Normal Pressure Hydrocephalus: A Cochrane Systematic Review
Christopher Carswell1, Anastasia Gontsarova3, Methley Abigail4, Hilary Watt2, Kevin Tsang5, Pearce Ronald1
1Brain Sciences, 2Department of Primary Care and Public Health, Imperial College London, 3Neuroradiology, Imperial College Healthcare NHS Trust, 4Department of Clinical Neuropsychology, North Staffordshire Combined Healthcare NHS Trust, 5Neurosurgery, Imperial College Hospitals NHS Trust
Objective:
To understand if there is high-quality evidence to support cerebrospinal fluid (CSF)-shunting for idiopathic Normal Pressure Hydrocephalus (iNPH).
Background:
In 1965 Hakim and Adams reported three cases of iNPH who improved with CSF-shunting. Subsequent clinical studies did not aways report a similar benefit leading to uncertainty about iNPH as a clinical entity. Historical meta-analyses reporting benefit of CSF-shunting in iNPH included retrospective, non-randomised studies with limited controls and substantial doubt remains.
Design/Methods:
We performed a Cochrane Systematic Review of the effect of CSF-shunting in iNPH including only prospective, randomised and controlled studies (RCTs). Outcomes included gait speed, disability, cognitive and urinary function, quality of life and adverse events using standard Cochrane methodology.
Results:
We detected 5 RCTs and excluded 1 unpublished RCT with no available results. We excluded 1 RCT with no population variation data from quantitative analysis. The 3 remaining RCTs included 140 participants (73 immediate and 67 delayed shunting). There was no data on ethnicity. Gait speed improved significantly with shunting with a “moderate” Cohen’s effect size. Disability improved with a risk ratio of 2.08 for functional independence with a number needed to treat of 3.4. There was no significant effect of shunting on cognitive screening. Urinary dysfunction improved with shunting by an uncertain effect-size. Risk of bias was low in all outcomes quantitatively evaluated. There was ”moderate” certainty of effect size for gait, and disability. There was no data on quality of life. Adverse events occurred in 29% overall but cranial surgery post shunt was rare (2%) and there was no mortality clearly associated with shunt surgery.
Conclusions:
There is some high-quality RCT evidence to support the use of CSF-shunting in iNPH but the sample size is small more studies are needed to increase the certainty of this evidence. More work is needed to understand which patients benefit from surgery.