• I A Scott, H Harden, and M Coory.
• Department of Internal Medicine, Princess Alexandra Hospital, Brisbane, QLD. ian_scott@health.qld.gov.au
• Med. J. Aust. 2001 Feb 5; 174 (3): 130-6.

ObjectiveTo assess the evidence that higher rates of coronary angiography (CA) and revascularisation (RV) in the subacute phase of acute myocardial infarction (AMI) improve patient outcomes.Data SourcesMEDLINE 1990 - December 1999, Current Contents 1990-1999, Cochrane Library (Issue 4, 1999), HealthSTAR 1990-1999, selected websites and bibliographies of retrieved articles.Study Selection And Data ExtractionStudies selected were (1) randomised trials comparing outcomes of "invasive" versus "conservative" use of CA and RV following AMI; (2) observational studies with formal methods comparing outcomes of high versus low rates of use of these procedures; and (3) clinical practice guidelines (CPGs), expert panel statements and decision analyses which met critical appraisal criteria, and which specified procedural indications. Outcome measures were rates of mortality, re-infarction and limiting or unstable angina.Data Synthesis56 articles were identified; 24 met inclusion criteria. Pooled data from nine RCTs of "invasive" (CA rate 96%; RV rate 66%) versus "conservative" (CA rate 28%; RV rate 19%) strategies showed no significant differences in mortality or re-infarction rates. Pooled results from 12 observational studies showed no mortality differences, but an excess reinfarction rate (8.0% vs 6.4%; P<0.001) in high- versus low-rate populations. Evidence of survival benefit from procedural intervention was strongest for patients with recurrent ischaemia combined with left ventricular dysfunction.ConclusionsIn the subacute phase of AMI, rates of CA and RV in excess of 30% and 20%, respectively, may not confer additional benefit in preventing death or re-infarction. However, variability between studies in design, patient selection, and extent of cross-over from medical to procedural groups, as well as limited data on symptom status, limits generalisability of results.

14 keywords...

hide…