Ten studies (39,453 patients) were included in the review, comprising five RCTs (4,768 patients) and five observational studies (34,685 patients). Sample sizes ranged from 142 to 17,400 patients. All the RCTs reported adequate allocation concealment, blind assessment of outcomes zero or minimal losses to follow-up, and use of intention-to-treat analyses. There was no blinding of patients or caregivers in any of the studies. The quality assessment of the observational studies indicated that most of the studies had low to moderate risks of the biases evaluated; one study did not adequately report sufficient information to ascertain the risk of selection bias.
At the longest follow-up, there were fewer deaths and recurrent myocardial infarctions when angiography was performed within 48 hours of initial symptom presentation than when angiography was performed more than 48 hours after the initial event (data not shown).
Patients receiving angiography more than 48 hours after initial symptoms were significantly more likely to die or experience recurrent myocardial infarction than patients who received angiography between 24 and 48 hours after symptom onset (OR 0.64, 95% CI 0.52 to 0.79; Ι²=0%; two observational studies; one RCT) for both one-month follow-up and longest follow-up.
Patients who received angiography within 12 to 24 hours of symptom onset were significantly less likely to die or experience recurrent myocardial infarction than patients who began angiography between 24 to 48 hours after symptom onset (OR 0.85, 95% CI 0.75 to 0.97; Ι²=50%; two RCTs, three observational studies) combined at both one month follow-up and longest follow-up.
There were no differences at one month follow-up, longest follow-up or overall for angiography commencing less than 12 hours after initial symptom presentation and angiography performed 12 to 24 hours after initial symptom presentation.
Rates of death or recurrent myocardial infarction were lower in patients treated with angiography from three to 48 hours after initial symptom presentation than for patients in which angiography commenced within three hours of symptom presentation at one-month follow-up (OR 0.42, 95% CI 0.25 to 0.69, Ι²=0%, two randomised trials).
Analyses of the RCTs comparing deaths or recurrent myocardial infarction found no statistically significant differences between different periods (12 to 24 hours compared with 48 hours and less than 12 hours and 12 to 24 hours after symptom presentation) of initiation of angiography.
There was no evidence of publication bias identified for the results in the review.