Table 5

Author, date and country	Patient group	Study type	Outcomes	Key results	Study weaknesses
Smith DA et al, 1992, USA	124 patients presenting to an ED with a heroin overdose	Observational	Time to decision	20 min	Treatments given were neither standardised nor randomised so analysis of outcome could not be performed in relation to mode of treatment.
			Further treatment after discharge	None	Follow up was poor so it is possible that patients who sought further treatment or who died elsewhere would have been missed.
Osterwalder JJ, 1995, Switzerland	192 patients attending an ED with clinical suspicion of opioid od	Observational	Time to decision	15 min	No attempt was made to compare the outcomes of different treatment modes The period of observation in the ED was not recorded.
			Reattendance if discharged	1 patient died
Watson WA et al, 1998, USA	84 patients attending an ED who had been given naloxone for a presumed opioid od	Observational	Subsequent recurrence of opioid toxicity	Patients who have taken a longacting opioid are more likely to experience a recurrence of toxicity	No follow up of patients was attempted after admission to hospital/discharge from the ED to assess the incidence of late complications.
					The period of observation in the ED was not recorded.
Vilke GM et al, 1999, USA	317 patients with a clinical suspicion of opioid od who refused to be transported to the ED after being given naloxone by the paramedics	Observational	Death	No patients treated with naloxone died	Variable doses and routes of administration of naloxone were used.
			Reattendence of the ambulance within 12 hours	Nil	No follow up of patients was attempted to ascertain if they received subsequent treatment or died in another area or attended the ED by other means of transport.
Christenson J et al, 2000, Canada	573 patients attending an ED with clinical evidence of opioid intoxication who had been given naloxone either in the prehospital setting or ED	Observational	Clinical prediction rule to predict safe discharge	Patients can be safely discharged one hour after administration of naloxone if they have normal mobility, Spo₂ >92%, respiratory rate 10–20/ min, heart rate 50–100/min , temperature 35–37.5°C, GCS 15/15	The rule has not been validated yet.
					The pattern of drug misuse in Vancouver is different from other cities, so there are concerns about whether these results can be applied to different populations (for example, those that misuse a higher proportion of longer acting agents).