Representativeness of trial populations: an example from a trial of adenotonsillectomy in children.

OBJECTIVE: To compare demographic and disease-specific characteristics of children included in a trial of the effectiveness of adenotonsillectomy with those of children who, although eligible, were not randomized.

MATERIAL AND METHODS: Characteristics were compared between (i) randomized children (n = 270); (ii) eligible children whose parents immediately decided not to participate (n = 243); and (iii) eligible children whose parents were informed about the trial by a member of the study team but did not give their informed consent for participation in the trial (n = 406).

RESULTS: Statistically significant but relatively small differences were observed between the groups. Notably. breathing difficulties at night and abnormal appearance of the tonsils were more prevalent among eligible children whose parents immediately refused participation in the trial than in the other two groups. In contrast, cervical lymphadenopathy and previous ENT surgery were less prevalent among eligible children whose parents immediately refused participation in the trial.

CONCLUSION: Few and relatively small differences were observed between randomized and eligible, non-randomized children. The trial population appears to be representative of the relevant patient domain, i.e. children currently undergoing adenotonsillectomy in The Netherlands.