CRD summary

This review concluded that, compared with medical therapy, revascularisation (by percutaneous coronary intervention or coronary artery bypass graft) reduced all-cause mortality, in people with stable coronary disease. There was no difference in nonfatal myocardial infarction. The quality of the included trials was not assessed and many were published before medical developments. These conclusions should be treated with caution.

Authors' objectives

To compare the effects of coronary revascularisation, either by percutaneous coronary intervention (PCI) or by coronary artery bypass graft (CABG), with medical therapy in people with stable coronary disease.

Searching

MEDLINE and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched for articles from 1977 to January 2008. The search terms were given and bibliographies of identified papers were checked.

Study selection

Randomised controlled trials (RCTs) that compared revascularisation (PCI or CABG) with medical therapy, for those with non-acute coronary artery disease, were eligible for inclusion. Trials had to report deaths or nonfatal myocardial infarctions at a follow-up of at least one year.

In the included trials, the inclusion criteria varied among no symptoms, stable angina, exercise-induced ischaemia, disabling angina, and recent or past myocardial infarction. Participants were mainly men (53 to 100%) and generally of working age, with one exception (mean age 80); the remaining mean ages ranged from 49 to 64 years. Where stated, zero to 39% of participants had diabetes. Some trials were on single-vessel disease, but others included two- or three-vessel disease. In earlier trials stents were not used, while they were increasingly used in later trials, where most of the PCI patients had them. Follow-up was 10 years in one study and ranged from one to 4.6 years in the others (median three years). The annual mortality in the medical therapy groups ranged from zero to 16.3%.

RCTs were selected for inclusion by two authors independently and disagreements were resolved by consensus with a third author.

Assessment of study quality

The authors did not report that they assessed validity.

Data extraction

Where all cause mortality was not available, the data for cardiac deaths were used. Event rates were extracted at the follow-up endpoint for each trial. In those with three arms the control group was used as the comparator for each treatment arm. Odds ratios, with 95% confidence intervals, were calculated.

The authors did not state how many reviewers extracted the data.

Methods of synthesis

The data were analysed on an intention-to-treat basis. Pooled odds ratios, with 95% confidence intervals, were calculated using the Peto method and a random-effects model. The results for the random-effects model appear to have been reported. Heterogeneity was assessed using the Cochran Q statistic. Cumulative meta-analyses were performed, recalculating the odds ratios as each trial was added, based on its date of publication.

Sensitivity analyses were performed to assess the effect of individual trials, by excluding each one individually. Subgroup analyses were performed based on revascularisation method (PCI or CABG) and disease status (after myocardial infarction or stable coronary disease).

Publication bias was investigated in a funnel plot and the fail-safe N method.

Results of the review

Twenty-eight RCTs (n=13,222 participants, reported as 13,121, which excluded a trial with no mortalities that was not meta-analysed) were included. Two trials had two treatment arms resulting in 30 comparisons. Of these, 18 were PCI (n=8,052; 7,951 analysed), eight were CABG (n=3,098), and four were mixed (either PCI or CABG; n=2,072). Study size ranged from 41 to 2,287 participants; publication year ranged from 1977 to 2007. There was no evidence of heterogeneity, nor publication bias.

Compared with medical therapy revascularisation reduced mortality (OR 0.74, 95% CI 0.63 to 0.88; 29 comparisons). Sensitivity analysis showed no effect from the removal of any individual trial. The cumulative odds ratio showed a significant reduction in mortality on inclusion of the third trial. Subsequent trials increased the precision, but did not change the overall findings.

Subgroup analyses: Compared with medical therapy, mortality was reduced with CABG (OR 0.62, 95% CI 0.50 to 0.77; eight comparisons) and with PCI (OR 0.82, 95% CI 0.68 to 0.99; 21 comparisons). The results for the trials that did not enrol patients who had experienced myocardial infarction were similar to the main results.

There was no difference in nonfatal myocardial infarction between medical therapy and revascularisation (OR 0.91, 95% CI 0.72 to 1.15; 28 comparisons, n=11,768).

Authors' conclusions

Compared with medical therapy alone, coronary revascularisation by CABG or PCI in conjunction with medical therapy, was associated with a reduction in mortality in people with nonacute coronary disease. There was no effect on nonfatal myocardial infarction.

CRD commentary

The inclusion criteria were clearly stated for study design, participants, interventions, and outcomes. The search was in two databases plus reference lists; there was no mention of any language and publication status restrictions. If there were restrictions, then language or publication bias could have affected the review; tests showed no evidence of publication bias. The methods of study selection were aimed at reducing reviewer error and bias, but these methods were not reported for data extraction. An assessment of the quality of the included trials was not reported, which makes it impossible to comment on their validity. The methods of synthesis were appropriate, but the participants in the control arms of two trials with two treatment arms were double counted. Potential heterogeneity was investigated. The content of the medical therapies in the included trials was not reported (for either treatment or comparison arms). The authors stated that many of the trials were performed before changes in medical treatments and it is possible that these could have affected the outcomes. The characteristics of participants in the trials mean that it might not be possible to generalise these results to individual patients.

The conclusions should be treated with caution.

Implications of the review for practice and research

Practice: The authors did not state any implications for practice.

Research: The authors stated that an adequately powered prospective trial was needed to confirm the results of their review.

Funding

None.

Bibliographic details

Jeremias A, Kaul S, Rosengart TK, Gruberg L, Brown DL. The impact of revascularization on mortality in patients with nonacute coronary artery disease. American Journal of Medicine 2009; 122(2): 152-161. [PubMed: 19185092]

Original Paper URL

http://www.amjmed.com/article/S0002-9343(08)00955-8/abstract

Indexing Status

Subject indexing assigned by NLM

MeSH

Angioplasty, Balloon, Coronary; Coronary Artery Bypass; Coronary Disease /drug therapy /mortality /therapy; Humans; Myocardial Revascularization; Randomized Controlled Trials as Topic; Survival Rate

AccessionNumber

12009103217

Database entry date

23/06/2010

Record Status

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.