This review compared the use of sirolimus-eluting stents (SES) and bare metal stents for coronary stenosis. The authors concluded that SES did not appear to increase the incidence of stent-related thrombosis. Given the limitations of the review, the conclusions should be treated with caution.

To investigate the risk of thrombosis associated with the use sirolimus-eluting stents (SES), compared with bare metal stents (BMS).

MEDLINE was searched from 2001 to 2004; the search terms were reported. In addition, abstracts of recent cardiology conferences were searched and the Science Citation Index was used to check references within identified articles. Authors of identified studies and experts were also contacted.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) and cohort or registry studies were eligible for inclusion. The duration of follow-up ranged from 8 to 13.5 months.

Specific interventions included in the review

Studies that compared the use of SES with BMS were eligible for inclusion. Studies were excluded if both treatment groups received drug-eluting stents. In the included studies, intraprocedural glycoprotein IIb/IIIa inhibitors were used in 6 to 60% of the participants. In addition, either clopidogrel or ticlopidine was administered for 2 to 6 months.

Participants included in the review

Studies on people with coronary artery disease who were undergoing stent implantation were eligible for inclusion. In the included studies, the mean age of the participants ranged from 60 to 64 years and 66 to 71% were men. Between 14 and 38% of the participants were current smokers, 15 to 30% were diabetic, and 28 to 48% had had a previous myocardial infarction. Details of coronary lesions were given.

Outcomes assessed in the review

To be included, studies had to assess thrombosis using angiographic or clinical data at 30 days' follow-up or later. The primary outcome of interest was stent thrombosis, defined as either angiographically documented thrombosis or a cardiac event when the study investigators clinically presumed that thrombosis had occurred. The outcomes were reported for short-term (up to 30 days) and longer term (over 30 days) follow-up.

How were decisions on the relevance of primary studies made?

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

The authors did not state that they assessed validity.

Two reviewers independently abstracted data on an intention-to-treat basis. Any discrepancies were resolved with a third reviewer. The number of thromboses and time to event were extracted from each study. The incidence of thrombosis and risk ratios (RR) with 95% confidence intervals (CIs) were calculated.

How were the studies combined?

A summary RR with 95% CI was calculated using a fixed-effect model. Where there were no events in a single arm, the automatic 'zero cell' correction method was used. The study results were also synthesised in a narrative.

How were differences between studies investigated?

Differences between the studies were presented in tabular format and discussed in the narrative.

Six studies (2,963 participants) were included: five RCTs (2,005 participants) and one registry study (958 participants).

In total (six studies), there were 8 thromboses in the SES groups and 16 in the BMS groups; an incidence of 0.5% and 1.1%, respectively. There was no significant difference in the risk of stent thrombosis with SES compared with BMS (RR 0.49, 95% CI: 0.22, 1.12, P=0.09).

The incidence of thrombosis within 30 days was 0.46% in the SES groups and 0.62% in the BMS groups (RR 0.76, 95% CI: 0.30, 1.88, P=0.55).

The incidence of thrombosis at over 30 days was 0.07% in the SES groups and 0.48% in the BMS groups (RR 0.25, 95% CI: 0.05, 1.16, P=0.08).

Compared with BMS, SES did not appear to increase the incidence of stent-related thrombosis up to 13.5 months after coronary intervention.

The inclusion criteria were clearly stated. The search was limited to one database, although the authors made attempts to find unpublished studies. It was unclear whether any language restrictions were applied and it is possible that some studies were missed. The method for selecting studies for the review was not described. It is possible that decisions made at this stage could have introduced error and bias into the review. The authors made no mention of assessing the quality of the included studies.

The data from RCTs and an observational study were pooled. In addition, heterogeneity did not seem to have been assessed statistically. Visual inspection of the forest plot suggests some degree of heterogeneity between the RCTs; the results of a statistical test for heterogeneity would have been helpful. Given the potential for heterogeneity and the different study designs, pooling of the results might not have been appropriate. In view of these comments, the authors' conclusions should be interpreted with caution.

The authors did not state any implications for practice or further research.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.