作者: Ivo W Tremont-Lukats , Bernardo O Ratilal , Terri Armstrong , Mark R Gilbert
DOI: 10.1002/14651858.CD004424.PUB2
关键词: Medicine 、 Brain tumor 、 Psychiatry 、 Absolute risk reduction 、 Relative risk 、 Pediatrics 、 Meta-analysis 、 Cochrane Library 、 Number needed to harm 、 Adverse effect 、 Clinical trial
摘要: Background Seizures can present at any time before or after diagnosis of a brain tumor. The risk seizures varies by tumor type and its location in the brain. For long we believed that preventing with antiepileptic drugs (seizure prophylaxis) was effective necessary, but supporting evidence little mixed. Such basis for previous reviews to conclude seizure prophylaxis ineffective people tumors. Objectives To estimate effectiveness tumors, adverse event rates identified clinical trials. Search methods A search strategy included free-text MeSH terms LILACS, EMBASE, PubMed, CENTRAL, Cochrane Library (1966 2007). Selection criteria Controlled trials random allocation, blinded unblinded, placebo observation control groups. Data collection analysis We screened articles, extracted data, rated validity each trial assess bias. Our primary outcome occurrence first seizure. secondary events. We pooled aggregate data into random-effects model meta-analysis using relative (RR). events, also number needed harm (NNH) absolute increase compute NNH. Main results There no difference between treatment interventions groups participants tumors. an higher those on than not (NNH 3; RR 6.10, 95% CI 1.10 34.63; P = 0.046). Authors' conclusions The is neutral, neither nor against prophylaxis, These conclusions apply only phenytoin, phenobarbital, divalproex sodium. decision start drug ultimately guided assessment individual factors careful discussion patients.