The 2005 Cochrane review of randomisedtrials comparing carotid endarterectomy with carotid angioplasty and stenting concluded that the evidence did not support a "widespread change in practice from recommending carotid endarterectomy as the treatment of choice" in patients with carotid disease.In today's Lancet, SPACE becomes the eighth randomised trial to be published2 and its sample size of 1200 will double the number of patients in the Cochrane database. So will SPACE lead to a change in recommendations after the next review?The methods in SPACE were different from those of its predecessors. In the traditional randomised trial, the null hypothesis assumes no difference between two interventions. The investigators then hope to reject the null hypothesis and show a clinically significant difference (a superiority trial). SPACE, however, aimed to show non-inferiority-ie, carotid angioplasty and stenting was no worse than carotid endarterectomy.3 In this type of study, a predetermined margin of non-inferiority is defined from the outset whose value will reflect the power calculation, the likely magnitude of risk to be encountered, and an allowance for whether a slightly increased procedural risk after carotid angioplasty and stenting might be offset by other important benefits (eg, no incision, no cranial nerve injury). In SPACE, the predefined limit for non-inferiority was that the upper confidence interval (CI) for the actual difference in the primary endpoint should be less than 2.5%. With intention-to-treat analysis, the actual difference was 0.51% (90% CI -1.89 to 2-91). In a per-protocol analysis, the actual difference was 1.32% (-110 to 3.76). The figure illustrates these findings in the context of the margin of non-inferiority. If carotid angioplasty and stenting was non-inferior (ie, as good as carotid endarterectomy), the upper CI would be less than 2.5 and within the shaded box. In both methods of analysis, however, the upper CI is more than 2.5, which shows that non-inferiority was not achieved. However, because the CI crosses zero, the difference is non-significant or uncertain. In other words, surgeons will conclude that carotid angioplasty and stenting was inferior to carotid endarterectomy, although interventionists will conclude that there was no significant difference.
展开▼