Five RCTs (n=1,519) met the inclusion criteria.
There were no statistically significant differences between vasopressin and epinephrine in terms of any of the five pre-specified outcomes of interest: failure of ROSC (5 RCTs; RR 0.81, 95% CI: 0.58, 1.12), death before hospital admission (2 RCTs; RR 0.72, 95% CI: 0.38, 1.39), death within 24 hours (2 RCTs; RR 0.74, 95% CI: 0.38, 1.43), death before hospital discharge (5 RCTs; RR 0.96, 95% CI: 0.87, 1.05), or combination of the number of deaths and neurologically impaired survivors (3 RCTs; RR 1.00, 95% CI: 0.94, 1.07).
The results of the pre-specified subgroup analyses also showed no statistically significant difference between vasopressin and epinephrine. However, the presence of statistical heterogeneity was reduced, in particular for death before hospital discharge when based on initial rhythm.
The methodological quality of the included studies ranged from a Jadad score of 5 (highest quality) for 3 studies to a score of 2 for the remaining 2 studies, although one of these trials was only available as an abstract. Those scoring highly had adequate allocation concealment and blinding of the caregivers. All of the trials had similar rates of follow-up and performed an intention-to-treat analysis.
The funnel plot appeared asymmetric, indicating publication bias. However, no evidence of publication bias was found when using Egger's method for failure of ROSC (P=0.24).
The levels of inter-rater agreement at each stage were adequate to high, with kappa statistics ranging from 0.64 (agreement of quality assessment) to 0.72 (agreement for retrieval of full papers) to 1 (agreement of full papers to be included in the review).