Sixty-four studies met the inclusion criteria (the number of patients was unclear); one was a RCT, five were non-randomised controlled trials, and the remainder were observational studies. The evidence was considered to be of low-to-moderate quality.
For community-based management (seven studies), oral antibiotics were associated with reductions in all-cause mortality (RR 0.75, 95% CI 0.64 to 0.89; four studies), and pneumonia-related mortality (RR 0.58, 95% CI 0.41 to 0.82; three studies). No studies were identified on the effect of oral antibiotics on sepsis-related mortality.
For community-based injected antibiotics, the RCT reported a significant reduction in all-cause mortality with a perinatal care package (RR 0.66, 95% CI 0.47 to 0.93), in domiciliary settings, where referral to hospital was not possible. A non-randomised controlled study reported a significant reduction in mortality, in a similar situation (RR 0.56, 95% CI 0.41 to 0.77). An observational study of children from the RCT reported a case-fatality rate of 4.4%. One uncontrolled study reported a case-fatality rate of 3.3%.
Across 55 observational studies evaluating hospital management of sepsis, the case-fatality rate ranged from 6.7% to 67%. Two observational studies evaluated hospital management of pneumonia, and found case-fatality rates of 14.4% and 30.8%. The results of the Delphi processes were reported.