Meta-analysis: Mortality is reduced when fibrinolytic therapy is started soon after the onset of MI symptoms
ACP J Club. 1997 Mar-Apr;126:31. doi:10.7326/ACPJC-1997-126-2-031
Boersma E, Maas AC, Deckers JW, Simoons ML. Early thrombolytic treatment in acute myocardial infarction: reappraisal of the golden hour. Lancet. 1996 Sep 21;348:771-5.
To determine the relation between benefit from fibrinolytics and treatment delay (time between onset of symptoms and treatment) for mortality from myocardial infarction (MI).
Studies published from 1983 to 1993 were identified with MEDLINE.
Randomized controlled trials were selected if they studied > 100 patients, if fibrinolytic therapy was compared with placebo or control medication, and if data on time from onset of symptoms to randomization were available.
Data were extracted on treatment, time from symptom onset to treatment, and short-term mortality (up to 35 d). Patients were divided into 6 time categories from symptom onset to randomization (see the Table). Absolute and relative mortality were calculated for each time category, and linear and nonlinear regression analyses were done to determine the relation between benefit and treatment delay.
22 trials (50 246 patients) were included in the regression analysis; data from 11 trials are analyzed in the Table. 5762 patients were randomly assigned within 2 hours of symptom onset, and 10 435 were assigned within 2 to 3 hours of onset. The number of patients needed to be treated with fibrinolytics (NNT) (rather than placebo or control medications) to save 1 additional life within 35 days are listed in the Table.
The proportional mortality reduction was highest in patients treated within 1 hour (48%) and was statistically significantly higher in patients treated within 2 hours (44%) than in those treated later (20%) (P = 0.001 for comparison of the 6 groups). The regression analysis found a weak coefficient with time as a variable in the analysis (0.06) but a large coefficient with the inverse of time in the analysis (29.3) in a steep drop-off of absolute risk within the first 2 hours.
Mortality from myocardial infarction is reduced when fibrinolytic therapy is started soon after symptom onset. The greatest reduction in mortality is for treatment within 2 hours. After 12 hours, the benefit of fibrinolytic therapy is no longer apparent.
Source of funding: Not stated.
For article reprint: Prof. M.L. Simoons, Thoraxcenter Bd 434, Erasmus University and University Hospital Rotterdam-Dijkzigt, Dr Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. FAX 31-10-4362841.
|0 to 1||15||11 to 26|
|≥ 1 to 2||27||18 to 50|
|≥ 2 to 3||38||27 to 71|
|≥ 3 to 6||34||25 to 52|
|≥ 6 to 12||56||34 to 192|
|≥ 12 to 24||No significant differences|
*Numbers calculated from data in article.
Thrombolysis for acute MI is widely accepted, but questions remain. The Fibrinolytic Therapy Trialists' (FTT) (1) collaborative meta-analysis of large trials of thrombolysis identified subgroups of patients, including those who presented 7 to 12 hours after symptom onset, who should benefit from thrombolysis. Mortality reduction was strongly related to time of presentation, with an NNT of 29 in those presenting at < 1.5 hours. Boersma and colleagues pooled the FTT trials, 2 smaller studies with data on time to randomization, and 11 additional trials of > 100 patients. A regression analysis showed a nonlinear relation of treatment benefit to time, with an NNT of 33 for each hour of delay in the first 2 hours and a much more gradual drop-off thereafter.
Patients who were treated early in the smaller trials had a disproportionately high treatment advantage compared with patients in the larger trials, possibly suggesting publication bias. Because of this possibility and uncertainty in the estimates, Boersma and colleagues' meta-analysis may not represent sufficient evidence to support the costly interventions in emergency services designed to improve time to treatment. For example, field cellular-transmitted electrocardiograms and thrombolysis before arrival at the hospital save 1 hour and may reduce mortality by 16% (2). Other interventions, such as emergency-department administration of fibrinolytics and accelerated ("fast-track") triage, can save substantial time. Patient delays, however, account for > 50% of the total delay in most studies; only 6% of the FTT patients presented in < 1.5 hours.
The most striking difference between patients who present early or late is that early presenters are twice as likely to attribute symptoms to their heart (3). Physicians may be able to decrease presentation time by educating their patients at risk for MI about typical and atypical presentations. Prompt emergency access and transport may reduce unnecessary delay and save lives.
Craig Redfern, DO
Portland Providence Medical CenterPortland, Oregon, USA