Table 1.

Author, date, country	Patient group	Study type	Outcomes	Key results	Study weaknesses
McGlone et al, 1998, UK	102 children (1–7 years). Divided into either midazolam (0.5 mg/kg nasal) or ketamine (2.5 mg/kg IM)	Prospective “allocation”	Behaviour before, during and after procedure. Vomiting before and after discharge. Time to discharge. Parental and nursing satisfaction	Significantly less restraint required in ketamine group (p<0.01). Ketamine caused more vomiting during recovery (p = 0.012). Midazolam children discharge 7 min earlier. Ketamine significantly preferred (p = 0.018)	? randomised. Two different routes of administration, uncertain final bioavailability of intranasal route. Small numbers. Atropine added to ketamine group
Everitt et al, 2002, Australasia	54 Australasian EDs. Ketamine 12%, midazolam 77%	Survey of agents used and ED physicians. 45 of 54 surveyed departments responded	Linear analogue depiction of perceived efficacy of sedation	IV ketamine 14% better sedation than midazolam	One respondent per ED. Open to responder bias. Not patient focused. No standard doses. Not limited to ketamine and midazolam alone
Roback et al, 2005, USA	2500 consecutive children (median age 6.7 years) receiving IV or IM procedural sedation. Ketamine 59%, midazolam/ketamine 12%, midazolam 10.4%	Prospective database. Retrospectively analysed	Respiratory complications.Apnoea/ larygospasm/ desaturation	Ketamine 6.1%, ketamine/midazolam 10%, midazolam 5.8%	Not randomised, not blinded. Multiple drug combinations. ?standardised doses. Glycopyrolate was given to all who had ketamine. ?mandatory reporting
			Vomiting	Ketamine 10.1%, ketamine/midazolam 5.4%, midazolam 0.8%
Sacchetti et al 2007, USA	226 children aged <13 years. Ketamine 60%, midazolam 28%. Across 14 community EDs	Prospective database	Adverse events	Nil recorded for both drugs	Database “self‐reported”, ?reliability. Low rate of respiratory events reported

ED, emergency department; IM, intramuscular; IV, intravenous.