Knowledge of treatment group does not bias assessment of time to seizure in an animal model of cocaine poisoning.

Abstract

Blinded outcome assessment decreases bias in human clinical trials. The necessity of blinded outcome assessment on animal studies is unknown. The authors determined the effect of knowledge of treatment group on assessment of time to seizure in an animal model of cocaine poisoning. Four subjects observed 20 animal experiments where all animals were administered a high dose of cocaine and placebo. For each experiment, two of the observers were told the animal had been treated with placebo and two were told the animal had been treated with a medication expected to delay the onset of seizures. Each observer recorded the time from cocaine administration to onset of seizure. The median time to seizure was compared between observers told the animal received placebo and those told the animal received active treatment. Seizures were reported by all subjects in 12 animals and by no subjects in five animals, and there was disagreement in three animals. The reported median time to seizure was similar for observers told that the animals were treated with placebo and those told they were treated with study medication. It is feasible to determine whether unblinded assessments are biased in an animal study. Knowledge of treatment group did not bias the assessment of time to seizure in this animal model.