London, United Kingdom 2013 Poster Session Blue Cerebrovasc Dis 2013; 35 (suppl 3)1-854 821 898 Meta-analysis and reviews CT and carotid stenosis in the prediction of recurrent stroke F.M. Chappell1, J.M. Wardlaw2 University of Edinburgh, Edinburgh, UNITED KINGDOM1, University of Edinburgh, Edin-burgh, UNITED KINGDOM2 Background: Improved prediction of recurrent stroke will improve delivery of secondary prevention treatment. Many factors linked to recurrent stroke have been studied. We examined the addition of brain CT or carotid imaging to clinical prediction models. Methods: We used systematic review methods to find, appraise, data extract, and meta-analyse stud-ies of any design with data on use of clinical items plus brain CT or carotid imaging to predict re-current stroke. We used a citation search method to find studies due to the variety of study designs, a critical appraisal tool based on those for prognostic studies, and extracted demographic, design, and imaging data. We meta-analysed hazard ratios (HR) where data were available. We added patient data from a large cohort study. Results: 1124 studies were screened by title/abstract. 121 full-text papers were assessed. 22 stud-ies met inclusion criteria, including 1 meta-analysis, with CT data for primary studies not available elsewhere. We used its primary studies in our review, with its CT estimates. There were 20 primary CT studies and 11 carotid studies but most were not meta-analysable due to variable effect size mea-sures. Different clinical items were used. We used 3 CT and 4 carotid studies and the cohort study (total 7 studies and 7246 patients) to meta-analyse HRs. Where given, the definition of ‘positive CT’ was a probably relevant infarct. 2 studies dichotomised stenosis at 50%, 1 at 70% (all NASCET), and 1 used nearest ECST 10%. Definition of recurrent stroke did not include severity. Recurrent stroke was significantly and independently associated with a probable relevant infarct on CT (HR 1.67, 95% CI 1.05 to 2.68, I2=69%), and carotid stenosis (HR 2.51, 95% CI 1.39 to 4.53, I2=82%). Conclusion: Severe carotid stenosis increases recurrent stroke risk 2.5-fold and CT 1.7-fold , inde-pendent of clinical items; but estimates have heterogeneity. 899 Meta-analysis and reviews Serum uric acid level and cognitive function; a systematic review and meta-analysis. A. Khan1, T. J. Quinn2, J. Dawson3 Institute of Cardiovascular & Medical Sciences, Glasgow, UNITED KINGDOM1, Institute of Cardiovascular & Medical Sciences, Glasgow, UNITED KINGDOM2, Institute of Cardiovascular & Medical Sciences, Glagsow, UNITED KINGDOM3 Background Previous studies have reported possible association between serum uric acid and cognitive function. We performed a systematic review to collate evidence on association of serum uric acid and cogni-tive decline or dementia. Methods We followed MOOSE guidance for review of observational studies. We created a sensitive search strategy, using previously validated search strings. Two independent reviewers searched six interna-tional, cross-disciplinary electronic databases (all inception to December 2012). We included studies that described association between uric acid level and a measure of cognitive function or clinical de-mentia diagnosis. Data and quality assessment were performed using standardised, validated forms. Results From 4,811 titles, 18 papers (n=11,194 participants) met our selection criteria. There was substantial clinical heterogeneity. Studies included patients with dementia (n=6 studies), Parkinson’s disease (n=5 studies) or unselected groups. Three studies were graded low risk of bias. Out of the 18 studies, 9 showed that serum uric acid levels were lower in cases of dementia or that higher levels are associated with better cognitive function. Conversely, 5 revealed an association between increased serum uric acid level and worse cognitive function or that serum uric acid levels were higher in cases of dementia. Four papers found no such relationships. For the 7 case control studies (n=1416 with dementia, n=4663), which compared uric acid level in patients with and without dementia, the standardised mean difference on meta-analysis (random ef-fects model) was -0.33 (95%CI:-0.59 to -0.07). Conclusion Meta-analysis showed that serum uric acid levels were lower in cases of dementia compared to con-trols. Narrative review however showed data were conflicting. Conclusions are limited by clinical heterogeneity and risk of bias in studies. An adequately powered study looking at uric acid inter-vention and robust cognitive outcomes is required.
Karger_ESC London_2013
To see the actual publication please follow the link above