The risk of a stroke within 7 days of a transient ischemic attack was lowest when patients were treated by specialist stroke services on an emergency basis, according to a systematic review and meta-analysis that included 10,126 TIA patients.
The review looked at 18 cohorts in 17 studies of different design in multiple countries, all assessing the risk of stroke following TIA within 7 days. The investigators searched Ovid Medline from 1950 to June 2007 and Embase between 1980 and June 2007, as well as books of abstracts from recent conferences for studies not yet published as full papers. Studies of patients with underlying pathologies were excluded.
The mean age of patients in the cohorts ranged from 61 to 73 years. The proportion of men ranged from 39% to 62% (Lancet Neurol. 2007;6:1063–72).
The pooled stroke risk across all cohorts for which 2-day data were available (a total of 15 cohorts, with 9,433 patients) was 3.1%. In the 17 cohorts for which data at 7 days were available (7,830 patients), the risk was 5.2%. However, significant heterogeneity among study design led to a large range of risk, from 0% to 12.8%.
The researchers classified the studies into the following categories, and calculated the risk for each category:
▸ Population-based studies over multiple emergency departments (EDs), with face-to-face follow-up. The three studies in this category had an overall 6.7% risk of stroke at 2 days and 10.4% risk at 7 days.
▸ Population-based studies at multiple EDs that relied on administrative follow-up (notes and diagnostic coding review) and that had important exclusions. Such exclusions included patients who were admitted to the hospital, patients with persistent symptoms at discharge from the ED, and strokes on the same day as the TIA. Three studies in this category had an overall stroke risk of 1.6% within 2 days and 3.0% within 7 days.
▸ Population-based studies at multiple EDs that used administrative follow-up but had no exclusions. Three studies in this group had an overall 2-day risk of 4.8% and a 7-day risk of 6.5%.
▸ Studies that ascertained consecutive patients attending a single ED with face-to-face or telephone follow-up. Three studies in this category showed an overall stroke risk after 2 days of 3.1% and risk at 7 days of 5.8%.
▸ Routine outpatient studies that used administrative follow-up. There was a 2-day stroke risk of 1.7% and a 7-day risk of 3.3% in the two studies in this group.
▸ Studies conducted at sites using specialist stroke services. The four studies in this group had the lowest incidence both of 2-day stroke risk, 0.6%, and 7-day risk, at 0.9%.
This analysis is not direct evidence for specialized centers, “but it is certainly supportive of specialist units,” lead investigator Dr. Matthew F. Giles, of the Oxford (England) University department of clinical neurology, said in an interview.
Rather than specialized stroke centers, “This study is supportive of specialist 'TIA units,' which should offer urgent access, rapid investigation, and immediate preventive treatment,” he explained. Four studies included in the meta-analysis involved that type of urgent access treatment.
The risk of publication bias was not totally overcome in his analysis, Dr. Giles admitted. However, he doubted this bias had any significant effect on his analysis. “Producing data on the outcome of TIA is a very labor-intensive exercise, and once a study has been completed, I very much doubt that it would go unpublished,” he said.
The researchers reported no conflicts of interest.