Prognosis of transient ischaemic attack in modern health care settings - a systematic review and meta-analysis

Background: Transient ischaemic attacks (TIAs) are common and place patients at risk of subsequent stroke. The 2007 EXPRESS and SOS-TIA studies, demonstrated the efficacy of rapid initiation of treatment. We hypothesised that, with these findings having informed subsequent TIA management protocols, the prognosis of TIA in contemporary (2008 and later) patient cohorts would be more favourable than in historical cohorts. Methods: A systematic review and meta-analysis of cohort studies of TIA (published 2008-2015). Primary outcome for the systematic review was stroke and secondary outcomes were mortality, TIA and myocardial infarction. For statistical analysis, the pooled cumulative risks of stroke recurrence were estimated from study specific estimates at 2, 7, 30 and 90 days post-TIA, using a multivariate Bayesian model. Results: We included 47 studies in the systematic review and 40 cohort studies (33 prospective and 7 retrospective) in the meta-analysis. In the systematic review (191,202 patients), stroke at 2-days was reported in 13/47 (27.7%) of studies, at 7-days in 20/47 (42.6%), at 30-days in 12/47 (25.5%) and at 90-days in 33/47 (70.2%). Studies included in the meta-analysis recruited 68,563 patients in total. The meta-analysis showed a cumulative risk of stroke of 1.2% (95% Credible Interval (CI) 0.6-2.2), 3.4% (95% CI 2.0-5.5), 5.0% (95% CI 2.9-8.9) and 7.4% (95% CI 4.3-12.4) at 2, 7, 30 and 90 days post-TIA, respectively. Conclusion: In modern health care settings, the prognosis of TIA patients is more favourable than in historical cohorts.