Table 3

Studies of firearm storage practices and suicide risk

Authors	Cases	Controls	Measures of firearm storage	Odds ratio (CI)*/p value
*Odds ratios = 95% confidence interval. †NS, not significant; p values not reported.
Brent et al, 1988⁴²	27	56 suicidal psychiatric inpatients	Firearms stored loaded	NS†
			Firearms stored locked	NS†
Brent et al, 1991⁴¹	47	(a) 47 suicide attempters	Guns stored locked	NS†
		(b) 47 never suicidal psychiatric controls	Guns stored together	NS†
		(b) 47 never suicidal psychiatric controls	with ammunition	NS†
			Guns stored loaded
Brent et al, 1993³⁸	7, all with no apparent psychopathology	(a) 60 adolescent suicide victims with psychiatric disorder	Loaded gun in the home	Cases v a: p<0.01
		(b) 38 community controls with no psychiatric disorder		Cases v b: p<0.01
Brent et al, 1993⁴³	67	67 matched community dwelling controls	Loaded gun in the home	NS†
Conwell et al, 2002³⁷	86, ages 50 and over	86 matched community dwelling controls	⩾1 gun kept unlocked	9.52 (1.52 to 58.82)
Conwell et al, 2002³⁷	86, ages 50 and over	86 matched community dwelling controls	⩾1 gun kept loaded	6.41 (1.17 to 35.71)
Kellermann et al, 1992³⁶	438	438 matched community dwelling controls	Any gun kept loaded	9.2 (4.1 to 20.1)
Kellermann et al, 1992³⁶	438	438 matched community dwelling controls	All guns kept unloaded	3.3 (1.7 to 6.1)
			Any gun kept unlocked	5.6 (3.1 to 10.4)
			All guns kept locked up	2.4 (1.0 to 5.7)
Shah et al, 2000³⁹	36 firearm suicide victims	36 matched community dwelling controls	⩾1 unlocked gun in the home	p = 0.05
Shah et al, 2000³⁹	36 firearm suicide victims	36 matched community dwelling controls	⩾1 unlocked gun in the home	Crude OR 2.57 (0.98 to 6.70)