Urgent carotid surgery in patients with crescendo transient ischaemic attacks and stroke-in-evolution: a systematic review.

OBJECTIVE To document outcomes after urgent carotid surgery in patients with crescendo transient ischaemic attacks (CTIAs) and stroke-in-evolution (SIE). METHODS A systematic review of the English-language literature using the MEDLINE and EMBASE databases "was undertaken from January 1984 through December 2007". Studies reporting the data after surgery for CTIAs and SIE were identified and stroke, stroke/death and stroke/death/major cardiac event rates calculated. The studies were divided into those reporting on CTIAs (group I) and those reporting on SIE (group II). RESULTS From the databases, 34 relevant series (915 patients) were retrieved. After excluding those reports on CTIAs, SIE and other 'urgent' indications combined together and from which separate data could not be extracted, a total of 12 series with 176 patients reporting on CTIAs (group I) and 16 series with 114 patients reporting on SIE (group II) met the inclusion criteria. All the identified studies were case series of low methodological quality suffering from selection bias. Peri-operative stroke, stroke/death and stroke/death/major cardiac event rates were 6.5% (95% confidence interval (CI): 3.4-10.4), 9.0% (95% CI: 4.3-15.1) and 10.9% (95% CI: 5.5-17.9), respectively, in group I; and 16.9% (95% CI: 9.2-26.2), 20.0% (95% CI: 12.5-28.6) and 20.8% (95% CI: 13.2-29.6), respectively, in group II. CONCLUSIONS The combined risk of neurological and cardiac complications following urgent carotid surgery for unstable neurological symptoms, such as CTIAs and SIE, is higher than that anticipated after elective surgery for stable symptoms.