Inpatients in substance use treatment with co-occurring psychiatric disorders: a prospective cohort study of characteristics and relapse predictors

Table 1 Sample characteristics of patients with COD compared to patients without COD.

Variables	Total (N = 611)		COD (n = 289)		Without COD (n = 322)		COD versus without COD
Variables	N^a	% or mean (SD)	n	% or mean (SD)	n	% or mean (SD)	p-value	Effect size
Age (years) at intake	610	30.0 (13.9)	288	32.9 (11.6)	322	42.7 (14.1)	< 0.000	0.759
Female	175	28.7	91	31.6	84	26.2	0.139	0.060
Education level low	189	32.6	109	39.5	80	26.3	< 0.001	0.140
Unstable housing arrangement¹	210	35.1	117	41.2	93	29.6	0.003	0.121
Motivation	610	4.25 (0.77)	289	4.26 (0.78)	321	4.24 (0.75)	0.663	0.027
Mental distress baseline	611	2.15 (0.71)	289	2.30 (0.69)	322	2.02 (0.70)	< 0.001	0.403
Mental distress follow-up²	425	1.93 (0.74)	206	2.07 (0.73)	219	1.79 (0.72)	< 0.001	0.386
Improved mental distress³	425	0.22 (0.79)	206	0.22 (0.79)	219	0.23 (0.80)	0.852	0.013
Length stay (days)	611	93.8 (79.6)	289	112.2 (92.2)	322	77.2 (61.9)	< 0.001	0.446
SUD diagnoses ⁴	585	95.7
- Alcohol use F10	345	59.0	132	48.4	213	68.3	< 0.001	0.202
- Opioid use F11	111	19.0	64	23.4	47	15.1	0.010	0.107
- Cannabis use F12	216	36.9	128	46.9	88	28.2	< 0.001	0.193
- Sedative use F13	170	29.1	99	36.3	71	22.8	< 0.001	0.148
- Stimulant use F15	188	32.1	105	38.5	83	26.6	0.002	0.127
- Alcohol use only	229	39.1	62	22.7	167	53.5	< 0.001	0.315
- Polysubstance use⁵	134	21.9	77	28.2	57	18.3	0.004	0.118

¹Reference: Owned or rented residence. ² For both patients with and without COD, the mean score of mental distress was lower at follow up (t = 41.15, df = 205, p < 0.001, and t = 36.66, df = 218, p < 0.001, respectively). ³ Calculated by subtracting the mean follow-up score from the mean baseline score. ⁴ More than one diagnosis could be registered for each patient. The most common two-substance combination was cannabis use and stimulant use (n = 113).⁵ Three or more SUD diagnoses
Notes: Percentages in valid percent. Categorical variables presented as valid percentages (%); continuous variables presented as mean (SD). Effect size was measured using Cohen’s d and Cramer’s V, as appropriate

ISSN: 1471-244X