Table 6

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study weaknesses
Rasanen J et al, 1985, Finland	40 patients with acute cardiogenic pulmonary oedema. RR >25 and Pao₂ <200 mm Hg	PRCT	Need for intubation	6/20 v 12/20 (NS)	Small numbers.
					Unblinded
	CPAP (20) v control (20)		Hospital mortality	17/20 v 14/20 deaths in hospital (NS)
Bersten A et al, 1991, Australia	39 patients with acute cardiogenic pulmonary oedema. Pao₂ <70 mm Hg and Paco₂ >45 mm Hg	PRCT	Need for intubation	0/19 v 7/20 (p<0.005)	Small numbers.
					Unblinded.
					Randomisation not concealed.
	CPAP (19) v control (20)		Hospital mortality	2/19 v 4/20 (NS)
Lin M and Chiang HT, 1991, Taiwan	55 patients with acute cardiogenic pulmonary oedema. RR >20	PRCT	Need for intubation	7/25 v 17/30 (p<0.05)
			Hospital mortality	2/25 v 4/30 (NS)
			Shunt size	Significantly improved in CPAP group
	CPAP (25) v control (30)		Pao₂	Significantly improved in CPAP group
Lin M et al, 1991, Taiwan	100 patients with a clinical diagnosis of acute cardiogenic pulmonary oedema	PRCT	Need for intubation	8/50 v 18/50 (p<0.01)	Unblinded
	CPAP (50) v control (50)		Hospital mortality	4/50 v 6/50 (NS)
Takeda S et al, 1998, Japan	22 patients with acute cardiogenic pulmonary oedema. Pao₂ <80 mm Hg.	PRCT	Need for intubation	2/11 v 8/11 (p=0.03)	Small numbers.
					Unblinded
	CPAP (11) v control (11)		Hospital mortality	1/11 v 7/11 (p=0.02)