Background:
Currently there is still some reluctance among emergency physicians and even neurologists to prescribe recombinant tissue plasminogen activator (rt-PA) for acute stroke. The majority of those unlikely to use it are concerned about causing harm to their patient. The object of this meta-analysis was to pool the mortality estimates at various time points following rt-PA administration. We hypothesized that there was a significant difference in mortality of patients treated with rt-PA and those treated with placebo. Methods: The databases PUBMED, EMBASE, SCOPUS, CENTRAL, and clinicaltrials.gov were searched for English-language randomized double blinded, placebo controlled trials of rt-PA. Mortality estimates were pooled and reported separately for days 7, 30, 90, and 180. Pooled relative risk (RR) and 95% confidence intervals were calculated using Dersimonian and Laird Random effects model. A priori designed sensitivity analyses included exclusion sensitivity analysis, and subgroup analysis into studies that included and excluded patients presenting within 3 hours of onset. Results: Eleven trials (N =6905) met the inclusion criteria. Total variance attributable to heterogeneity was not significant (I2 <50%) for any time point except at 30 days (I2, 53%). Publication bias was not significant on Begg’s and Egger’s tests. Mortality was not significantly different between the two groups at any time point. Exclusion sensitivity analysis revealed that exclusion of the NINDS trial rendered the 7 day and 30 day difference significant. Trials that randomized patients beyond 3 hours did not drive the mortality difference seen at any time point. Conclusion: In conclusion, compared with placebo, rt-PA does not impact mortality at any time point after administration through 6 months.
