Journal Club Summary
Methodology Score: 4/5
Usefulness Score: 2.5/5
Mentzelopoulos SD, et al. JAMA. 2013;310(3):270-279.
This randomized, multi-centered, double blind, placebo controlled trial compared vasopressin, steroids and epinephrine (VSE) vs. epinephrine for in-hospital cardiac arrest and found a 18% ARR in favour of the VSE group for ROSC > 20 mins and a 8.8% ARR in favour of the VSE group for survival with a neurologically favourable outcome. JC attendees were intrigued by the results, however, important baseline characteristics were different between both groups and a multivariate analysis would have added strength to the results. Applicability to ED patients was also questioned as these were admitted patients mostly presenting with asystole or PEA. For now, this trial is best considered to be hypothesis generating rather than practice changing.
By: Dr. Lisa Fischer
(Presented Nov 2013)
Epi lesson:
Adjustment analyses for unbalanced baseline characteristics
One purpose of randomization is to create two comparison groups which are balanced in terms of demographic and important baseline clinical variables. When examining table 1, the baseline characteristics, if the reader notes a clinically significant difference in an important baseline characteristic which could influence the outcome being studied, one should search for an adjusted analysis which controls for this difference. If the results of the adjusted analysis match the unadjusted analysis, the difference is unlikely to significantly influence the result and the reader can be reassured. Without the adjustment analysis, however, a critical reader cannot tell whether the lack of balance for this characteristic significantly biases the result.
By: Dr. Lisa Calder
