Forty-two observational studies were included (over 9,000 participants). These included 30 pre/post test studies, seven studies where controls were used (four which used pre/post test and controls) and one interrupted time series study. Most studies (53 per cent) had less than 100 participants. Only seven were multi-site studies.
The designs of all the included studies were weak and insufficient information was presented in the studies about the implementation of the interventions, which made it difficult to generalise from the results.
Process interventions: Fast-track or direct admission to cardiac care unit (three studies), administering thrombolytics in the emergency department (eight studies), nurse-administered thrombolytics (six studies) and use of technology (two studies: computerised decision support; and faxing electrocardiograms) all showed improved door-to-needle times. Emergency department activation of catheterisation laboratory (two studies) and policy changes (four studies: on-site percutaneous coronary intervention without surgical backup; or protocol for transferring patients to primary percutaneous coronary intervention facility) showed reduced door-to-balloon times.
System interventions: Continuous quality improvement significantly decreased door-to-needle time time (five studies) and door-to-balloon time (two studies). Two studies evaluated clinical staff education, but only one found a significant change in door-to-needle time time. One study of audit and feedback showed a reduction in door-to-needle time, although no information was available regarding the significance of this reduction. Four studies of critical pathways all showed a significant reduction in door-to-needle time (two studies) and door-to-balloon time (two studies). Two studies assessed multifaceted interventions, but neither of these showed a reduction in door-to-needle time.