Article
An algorithm to assess the rehabilitation potential in chronic hydrocephalus with a minimal data set
Ein Algorithmus um das Erholungspotential bei chronischem Hydrozephalus mit wenigen Daten zu ermitteln
Search Medline for
Authors
Published: | May 30, 2008 |
---|
Outline
Text
Objective: In the daily clinical practice it is often necessary to judge whether patients with a chronic hydrocephalus (e.g. Normal pressure hydrocephalus) might have a realistic chance to benefit from treatment or not. Objective was to establish a decision tree with the need of a minimal data set to provide a reliable prognosis.
Methods: A total of 125 patients with chronic hydrocephalus (60% idiopathic Normal-Pressure Hydrocephalus) was assessed, shunted and followed-up for at least one year. At the time of clinical presentation a minimal data set comprising of a clinical grading (Kiefer-Index), a Co-Morbidity Grading (CMI), age and the duration of the anamnesis have been collected. The clinical result was dichotomized to responders and non-responders after a follow-up of 1 year after shunting. Shunt indication based on CSF hydrodynamics and was not influenced by the estudied data. Statistics: ANOVA, CHI-square-, Spearman-, Kuskal-Wallis, Wilcoxen-U-Test.
Results: The overall responder rate was 79%. All variables had a significant (age: p=0.02, anamnesis duration: p=0.04, Co-Morbidity: p<0.000, preoperative clinical state: p=0.0016) but not independent influence on outcome. A decision tree could be established which allows with an astonishing precision to predict whether patients are good shunt candidates or not. Some examples: The first to look at is the CMI: The prognosis becomes clearly worse with CMI >3 points and the rehabilitation chance is only minimal with CMI >12 points. A preoperative worse clinical state (KI >12 points) together with an older age (>70 years) and an anamnesis duration >1 years reduces the responder rate to <40%. For each given combination of data a precise prediction of the individual rehabilitation chance is possible.
Conclusions: With the established decision tree it is possible to predict the individual chance to benefit from hydrocephalus treatment at the hand of a very restricted data set which can be assessed with the first presentation of outpatients facilitating the decision whether or not further invasive measures are justified or not.