Elsevier

The Lancet Neurology

Volume 6, Issue 12, December 2007, Pages 1063-1072
The Lancet Neurology

Fast track — Articles
Risk of stroke early after transient ischaemic attack: a systematic review and meta-analysis

https://doi.org/10.1016/S1474-4422(07)70274-0Get rights and content

Summary

Background

Stroke is often preceded by transient ischaemic attack (TIA), but studies of stroke risk after TIA are logistically difficult and have yielded conflicting results. However, reliable estimation of this risk is necessary for planning effective service provision, clinical trials, and public education. We therefore did a systematic review of all studies of stroke risk early after TIA.

Methods

All studies of stroke risk within 7 days of TIA were identified by use of electronic databases and by hand searches of reference lists, relevant journals, and conference abstracts. Stroke risks at 2 days and 7 days after TIA were calculated overall and analyses for heterogeneity were done, if possible, after categorisation by study method, setting, population, and treatment.

Findings

18 independent cohorts were included, which reported stroke risk in 10 126 TIA patients. The pooled stroke risk was 5·2% (95% CI 3·9–6·5) at 7 days, but there was substantial heterogeneity between studies (p<0·0001), with risks ranging from 0% to 12·8%. However, the risks reported in individual studies over different durations of follow-up were highly correlated (0–7 days vs 8–90 days, r=0·89, p<0·0001), and the heterogeneity between studies was almost fully explained by study method, setting, and treatment. The lowest risks were seen in studies of emergency treatment in specialist stroke services (0·9% [95% CI 0·0–1·9], four studies) and the highest risks in population-based studies without urgent treatment (11·0% [8·6–13·5], three studies). Results were similar for stroke risk at 2 days.

Interpretation

The reported early risks of stroke after TIA were highly heterogeneous, but this could be largely accounted for by differences in study method, setting, and treatment, with lowest risks in studies of emergency treatment in specialist stroke services.

Introduction

Up to 23% of strokes are preceded by transient ischaemic attack (TIA),1 and there is substantial research interest in improving prevention during the short window between TIA and stroke.2 However, the prospective estimation of the early risk of stroke is logistically difficult and studies have yielded inconsistent results,3 with 7-day risks ranging from 0% to 12·8%.4, 5 Nevertheless, reliable estimation of this risk is essential to allow us to understand the potential absolute benefits of early treatment and the likely cost-effectiveness of different methods of service provision, to inform future clinical trial design, and to justify investment in public education.

The measured risk will depend on several factors. First, study method is important, particularly the delay from TIA to study entry, the inclusion or exclusion of patients having a stroke during this interval, and the thoroughness of follow-up used to identify subsequent stroke.6 Second, the clinical setting (population, emergency department, clinic, or specialist stroke service) and study population may influence the measured risk. Third, the intensity and timing of secondary preventive treatment may reduce stroke risk after TIA.4, 7, 8 Fourth, the stroke risk after TIA has been shown to vary according to characteristics of the patient (age, blood pressure, diabetes),9, 10 clinical features of the event (focal weakness, speech deficit, and symptom duration),9, 10, 11 and underlying aetiology.12 Thus, differences in study method, setting and population, treatment, and case mix must all be considered when interpreting heterogeneity in measured risk between studies.

We therefore systematically reviewed studies of the risk of stroke within 7 days after a TIA, to estimate stroke risk overall and to determine the influence of study method, setting, population, treatment, and case mix.

Section snippets

Methods

We aimed to identify all studies reporting the risk of stroke within 7 days of TIA, irrespective of the study design, setting, or language, in accordance with the MOOSE guidelines13 for meta-analysis of observational studies in epidemiology. Ovid Medline (1950 to June, 2007) and Embase (1980 to June, 2007) were searched by use of both the medical subject heading (MESH) terms and text words: [TIA OR amaurosis fugax OR transient isch(a)emic attack] AND [outcome OR prognosis OR follow-up OR cohort

Results

The search of electronic databases yielded 11 350 publications. After initial screening, 69 reports were identified for full-text review. 15 additional reports were identified by searching relevant reference lists and abstract books from recent conferences. No further studies were identified by hand searches of the three journals from which most eligible studies were identified electronically (Stroke, The Lancet, and Cerebrovascular Diseases). Of the resulting 84 reports that were reviewed in

Discussion

This review showed an overall risk of stroke at 2 days after TIA of 3·1% (95% CI 2·0–4·1) in 15 cohorts including 9433 patients, and 5·2% (3·9–6·5) at 7 days in 17 cohorts including 7830 subjects, but with significant heterogeneity (p<0·0001) between studies for both results. This degree of heterogeneity, the high correlation between stroke risks at 0–7 days and 8–90 days in individual studies, and the reduction in heterogeneity by stratification of studies by common methods and setting suggest

References (37)

  • MF Giles et al.

    Prediction and prevention of stroke after transient ischemic attack in the short and long term

    Expert Rev Neurother

    (2006)
  • M Correia et al.

    Transient ischemic attacks in rural and urban northern Portugal: incidence and short-term prognosis

    Stroke

    (2006)
  • PM Rothwell

    Incidence, risk factors and prognosis of stroke and transient ischaemic attack: the need for high-quality large-scale epidemiological studies

    Cerebrovasc Dis

    (2003)
  • MD Hill et al.

    The high risk of stroke immediately after transient ischemic attack. A population-based study

    Neurology

    (2004)
  • JK Lovett et al.

    Early risk of recurrence by subtype of ischemic stroke in population-based incidence studies

    Neurology

    (2004)
  • DF Stroup et al.

    Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group

    JAMA

    (2000)
  • P McCullagh et al.

    Generalised linear models

    (1979)
  • S Ricci et al.

    A community-based study of incidence, risk factors and outcome of transient ischaemic attacks in Umbria, Italy: the SEPIVAC study

    J Neurol

    (1991)
  • Cited by (0)

    View full text