After the completion of a multicentre study, it was disputed whether becoming acquainted with a novel drug would affect dosing, and how many consecutive patients would be sufficient for this phenomenon.Methods:
A post hoc analysis of the data on 408 patients from a multicentre study on propofol was performed to reveal a possible learning effect. For study purposes, the patients were given consecutive anaesthetic serial numbers per anaesthesiologist. Patients eligible for the study with serial numbers 3-8 (113) were included in the initial group, and numbers 12-17 (89) in the end group.Results:
The patients in the end group opened their eyes (4.1 vs. 5.5 min, P<0.05), gave their date of birth (4.5 vs. 6.3 min, P<0.005) and walked sooner (27.1 vs. 49.8 min, P<0.05) than the patients in the initial group. They received a higher dose of propofol at induction (2.37 vs. 2.26 mg kg-1, P<0.05) and the last additional dose of propofol earlier (3.3 vs. 2.7 min, P<0.05).Conclusion:
This study shows that the learning effect can influence the results in a multicentre study. Learning contamination may occur if a novel drug is dosed by clinical judgement, and if the allocation of patients into groups is markedly uneven during the different stages of the study.