TY - JOUR T1 - Journal scan JF - Emergency Medicine Journal JO - Emerg Med J SP - 135 LP - 137 DO - 10.1136/emj.18.2.135 VL - 18 IS - 2 AU - Jonathan Wyatt AU - Rakesh Khanna Y1 - 2001/03/01 UR - http://emj.bmj.com/content/18/2/135.abstract N2 - Edited by Jonathan Wyatt; this Scan Coordinated By Rakesh Khanna S R Mehta, J W Eikelboom, S Yusuf. Lancet2000;356:449–53 Objectives—To determine whether there is an increased risk of intracranial haemorrhage (ICH) if thrombolytic agents are administered as boluses in patients with acute myocardial infarction. Method—A meta-analysis of randomised trials comparing bolus and infusion thrombolytic therapy. Data pertaining to ICH and also non-haemorrhagic stroke, reinfarction rates and 30 day mortality were also subject to meta-analysis. Results—A total of seven trials involving more than 100 000 patients were identified for inclusion in the meta-analysis. Bolus treatment was found to be associated with a greater risk of ICH (0.8% v 0.6%; odds ratio 1.25 (95% CI 1.08, 1.45); p=0.003). The authors calculated that this translated to an absolute excess of two ICHs per 1000 patients treated. This increased risk was not offset by any decrease in 30 day mortality, reinfarction or non-haemorrhagic stoke, which did not significantly differ between the two techniques. Conclusions—Although bolus administration may be more convenient, this must be balanced against an increased risk of ICH. Given that the convenience of bolus administration of thrombolytic therapy in hospital practice is of limited value, there is little justification in using this technique. Critique—The authors identified 43 randomised controlled trials comparing bolus with standard infusion of thrombolytic therapy by electronic and manual searches. After using a structured exclusion process seven trials were identified for inclusion in the meta-analysis. However, the trials analysed were not entirely homogenous in terms of the original methodologies used. Two trials compared the same bolus and infusion agents, whereas the other five compared different agents. Four of the trials studied were double blinded and three were open labelled. Despite these differences in study design, all of the trials did study very similar population groups. The lines of the confidence limits for the … ER -