Background: Pre-medication has been used to protect against early adverse reactions (EAR) following antivenom administration after snakebite. Studies have evaluated its efficacy with variable results.
Objective: The aim of the study was to conduct a systematic review and metaanalysis of published data to assess the effect of pre-medication on the risk of EAR.
Methods: We conducted a search of MEDLINE, the Cochrane Database and various search engines/websites, searched handbooks, book chapters and peer-reviewed articles relating to clinical snakebite, and consulted experts in this field. The search was on published literature up to September 2010. A meta-analysis was conducted of all randomized and non-randomized studies of EAR following antivenom in snakebite that utilized either adrenaline (epinephrine)-containing or non-adrenaline (antihistamines, corticosteroids)-containing pre-medications compared with control groups. We performed either random- or fixed-effects analysis based on the presence of heterogeneity as assessed with two tests, including the I(su2) statistic, and performed restricted analyses on data derived from randomized or non-randomized studies. Sensitivity analysis investigating the influence of single studies on overall estimates was conducted and we determined publication bias where detected in both of the two tests used for its assessment.
Results: Three randomized and four non-randomized studies were selected for inclusion in this study. When all ten comparisons from the seven selected studies were combined (with a total of 434 subjects in the pre-medication groups and 399 subjects in the control groups), the overall summary risk ratio (RR) for EAR was 0.70 (95% CI 0.50, 0.99; p = 0.041; I2 = 66.5%). When analysis was restricted to only studies employing adrenaline-containing premedication, the combined summary RR was 0.32 (95% CI 0.18, 0.58; p < 0.0001; I2 = 9.5%). Results were not statistically significant when analyses were restricted to studies employing non-adrenaline-containing pre-medications, or cohort or randomized controlled designs. Analysis was limited by heterogeneity, paucity and quality of data.
Conclusions: Findings are consistent with a substantial beneficial effect of adrenaline pre-medication, but a marginal benefit with the combination of pre-medications used against EAR could not be excluded. Future studies are recommended and they should explore possible synergism of broader combinations of drugs and effects of mode of antivenom administration in large randomized controlled trials. Meanwhile, highly purified antivenoms with less risk of EAR should be made available in the rural tropics.