Carotid bruits as a prognostic indicator of cardiovascular death and myocardial infarction: a meta-analysis

Background Although carotid bruits are deemed to be markers of generalised atherosclerosis, they are poor predictors of cerebrovascular events. We investigated whether a carotid bruit predicts myocardial infarction and cardiovascular death. Methods in this meta-analysis, we searched Medline (1966 to August, 2007) and Embase (1974 to August, 2007) with the terms "carotid" and "bruit". Bibliographies of all the retrieved articles were also searched. Articles were included if they reported the incidence of myocardial infarction or cardiovascular death in adults. Outcome variables were extracted in duplicate and included the rate of myocardial infarction and cardiovascular mortality. Quality of the articles was independently assessed with the Hayden rating scheme. Data were pooled with a random effects model. Findings Of the 22 articles included, 20 (91%) used prospective cohorts. Our analysis included 17295 patients with 62413.5 patient-years of follow-up, with a median sample size of 273 patients (range 38-4736) followed up for 4 years (2-7). The rate of myocardial infarction in patients with carotid bruits was 3.69 (95% CI 2.97-5.40) per 100 patient-years (eight studies) compared with 1.86 (0.24-3.48) per 100 patient-years in those without bruits (two studies). Yearly rates of cardiovascular death were also higher in patients with bruits (16 studies) than in those without (four studies) (2 . 85 [2.16-3.54] per 100 patient-years vs 1 . 11 [0 . 45-1.76] per 100 patient-years). in the four trials in which direct comparisons of patients with and without bruits were possible, the odds ratio for myocardial infarction was 2.15 (1.67-2.78) and for cardiovascular death 2.27 (1.49-3.49). Interpretation Auscultation for carotid bruits in patients at risk for heart disease could help select those who might benefit the most from an aggressive modification strategy for cardiovascular risk.