Forty-nine children having day-stay surgical procedures were randomly assigned to receive oral midazolam 0.75 mg.kg-1 or placebo in a double blind fashion. The child's level of anxiety was assessed before premedication using parental, child and observer scales. The child and observer anxiety scores were repeated in the anaesthetic room. Most children presented for anaesthesia in a calm state, irrespective of whether they had received midazolam. Parents tended to overestimate their child's level of anxiety. Observer anxiety scores reliably predicted behaviour during induction of anaesthesia in the absence of a sedative. Observer scores decreased in the midazolam group (P < 0.02), but not in the placebo group, children below six years having the greatest decrease with midazolam. The median time to discharge from hospital was delayed by 30 min in the midazolam group (P < 0.01). Children do not require routine sedative premedication for day case procedures, but oral midazolam is useful in producing calm behaviour in those children with high observer anxiety scores.

A woman at 32 weeks' gestation with eclampsia was given 120 mg diazepam shortly before emergency caesarean section. The infant had persistent apnoea and required respiratory support. Spontaneous respiration began after intravenous flumazenil infusion was started. Diazepam and its active metabolites were assayed during and after 5 d of treatment with flumazenil.