Fourteen RCTs (673 participants) were included. One trial had 199 participants others ranged from 15 to 90 participants. In eight studies follow-up was between three days to eight weeks, and in the remaining six from four months to two years.
Only three trials reported an adequate method of allocation concealment, seven were described as blinded and only two met all validity criteria.
Tests did not show evidence of publication bias.
Compared to control, dobutamine had no statistically significant effect on mortality (OR 1.47, 95% CI 0.98 to 2.21; Ι²=4.5%; 13 trials). In subgroup analyses, results were similar to the main analysis for in-patients and for out-patients. Meta-regression indicated no effect on results of any aspect of study validity; sensitivity analysis showed no evidence of undue influence of any one study.
Where reported, symptomatic changes varied, were insufficient for pooling and were reported in tables. Some studies also reported on adverse events in control and dobutamine groups.