In comparing outcomes between drug-eluting stents (DES) and coronary artery bypass grafting for unprotected left main coronary disease, the meta-analysis by Jiang et al included 16 studies in which major adverse cardiac and cerebrovascular events (MACCEs) were assessed with a follow-up period of ≤2 years. We reexamined the same 16 studies (and their results on the basis of this follow-up length) to determine whether any temporal trend could be detected in the risk ratio (RR) for 3 main outcome measures (MACCEs including target vessel revascularization [TVR], MACCEs excluding TVR, and TVR). Temporal trends were investigated using standard meta-regression methods ; as usual, RR values were log transformed. We reextracted the event frequencies from the primary studies because Jiang et al did not present crude event rates. Two of us (A.M. and D.M.) separately carried out this extraction; differences were resolved by consensus, but priority was assigned to outcomes determined after a follow-up period of about 2 years. In analyzing temporal trends, individual studies were represented by the midpoint of their enrollment interval.
Figure 1 shows the results of our meta-regression. The 2 end points of MACCEs including TVR and MACCEs excluding TVR showed substantial stability in their RRs over time (with a slight increase from 2002 to 2007, particularly for the second end point, but significance was absent in both cases). Likewise, the temporal analysis of RR for TVR found that although DES were clearly associated with increased risk for TVR, the RR for this end point showed only a modest decrease over time, which however remained far from statistical significance.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
