Fourteen RCTs (n=846) and two observational studies (n=283) were included in the review. Four RCTs reported use of adequate allocation concealment, seven used an intention-to-treat analysis and none reported use of blinding.
Continuous or extended infusion was associated with neither improvement in clinical cure (OR 1.04, 95% CI 0.74 to 1.46, I2=0%; nine studies) nor improvement in mortality (OR 1.00, 95% CI 0.48 to 2.06, I2=15%; nine studies).
A funnel plot indicated a possibility of publication bias. Two observational studies indicated extended or continuous infusion was associated with improvement in clinical cure.