Validation of key behaviourally based mental health diagnoses in administrative data: suicide attempt, alcohol abuse, illicit drug abuse and tobacco use

Hyungjin Myra Kim; Eric G Smith; Claire M Stano; Dara Ganoczy; Kara Zivin; Heather Walters; Marcia Valenstein

doi:10.1186/1472-6963-12-18

Table 3 Alcohol dependence, illicit drug dependence and tobacco use diagnoses: comparison of administrative diagnoses against chart notation during the past 12 months prior to and including the index date

From: Validation of key behaviourally based mental health diagnoses in administrative data: suicide attempt, alcohol abuse, illicit drug abuse and tobacco use

	*Alcohol Problems, Abuse/Dependence*
	*Unweighted Estimates, %* ^b *(N in fraction)*
*Sample*	*Agreement*	*Sensitivity*	*Specificity*	*PPV*	*NPV*
Suicide Death	90 (331/368)	74 (90/121)	98 (241/247)	94 (90/96)	89 (241/272)
Control	90 (327/362)	54 (28/52)	96 (299/310)	72 (28/39)	93 (299/323)
AD New User	90 (515/571)	63 (73/116)	97 (442/455)	85 (73/86)	91 (442/485)
Combined^a	90 (658/730)	68 (118/173)	97 (540/557)	87 (118/135)	91 (540/595)
	*Weighted Estimates, %* ^b *(95% confidence interval)*
Suicide Death	90 (87, 94)	78 (70, 87)	97 (95, 100)	94 (88, 100)	89 (84, 93)
Control	90 (85, 95)	55 (34, 75)	97 (95, 99)	80 (67, 94)	91 (86, 96)
AD New User	91 (87, 94)	59 (46, 73)	98 (96, 100)	86 (75, 97)	91 (88, 95)
Combined^a	90 (85, 95)	55 (35, 75)	97 (95, 99)	80 (68, 93)	91 (86, 96)
	*Illicit Drug Use/Abuse/Dependence*
	*Unweighted Estimates, %* ^b *(N in fraction)*
	*Agreement*	*Sensitivity*	*Specificity*	*PPV*	*NPV*
Suicide Death	90 (331/368)	74 (26/35)	92 (305/333)	48 (26/54)	97 (305/314)
Control	96 (348/362)	58 (11/19)	98 (337/343)	65 (11/17)	98 (337/345)
AD New User	93 (533/571)	70 (38/54)	96 (495/517)	63 (38/60)	97 (495/511)
Combined^a	93 (679/730)	69 (37/54)	95 (642/676)	52 (37/71)	97 (642/659)
	*Weighted Estimates, %* ^b *(95% confidence interval)*
Suicide Death	90 (86, 94)	78 (62, 94)	92 (88, 95)	51 (35, 67)	97 (95 99)
Control	97 (94, 99)	61 (30, 91)	99 (97, 100)	70 (43, 96)	98 (96, 100)
AD New User	91 (85, 96)	54 (24, 85)	95 (92, 98)	58 (37, 79)	94 (89, 100)
Combined^a	97 (94, 99)	61 (33, 89)	99 (97, 100)	69 (45, 93)	98 (96, 100)
	*Tobacco Use*
	*Unweighted Estimates, %* ^b *(N in fraction)*
	*Agreement*	*Sensitivity*	*Specificity*	*PPV*	*NPV*
Suicide Death	73 (267/368)	42 (70/167)	98 (197/201)	95 (70/74)	67 (197/294)
Control	76 (274/362)	36 (46/128)	97 (228/234)	89 (46/52)	74 (228/310)
AD New User	74 (424/571)	40 (95/237)	99 (329/334)	95 (95/100)	70 (329/471)
Combined^a	74 (541/730)	39 (116/295)	98 (425/435)	92 (116/126)	70 (425/604)
	*Weighted Estimates, %* ^b *(95% confidence interval)*
Suicide Death	70 (64, 76)	42 (33, 52)	97 (94, 100)	94 (87, 100)	63 (56, 70)
Control	76 (70, 82)	39 (27, 50)	98 (96, 100)	91 (83, 99)	73 (66, 80)
AD New User	77 (72, 82)	44 (34, 54)	99 (97, 100)	95 (90, 100)	73 (67, 79)
Combined^a	76 (70, 82)	39 (27, 50)	98 (96, 100)	91 (84, 99)	73 (66, 80)

Note: For alcohol dependence, kappa values are 0.76 for suicide cases, 0.57 for controls and 0.66 for AD new users. For illicit drug dependence, kappa values are 0.53 for suicide cases, 0.55 for controls and 0.63 for AD new users.
Abbreviation: NA is not available, AD is antidepressant; PPV is positive predictive value of an administrative diagnosis; NPV is the negative predictive value of an administrative diagnosis.
^a Combined sample of suicide deaths and controls
^b Rounded to a whole number

Back to article page