Table 1.

Baseline Characteristics and Multivariable Logistic Regression Models for Affordability of Mental Health Services

	Baseline Char.		Unable to afford MHS
	N	%	AOR	95% CI		p > |t|	% unable to afford MHS
1A
Race	14,655
White	12,703	88.77	1.00	-	-	-	2.57
Non-white	1952	11.23	1.27	0.84	1.94	0.260	3.14
Age	14,655
18–64	6799	51.80	1.00	-	-	-	4.6
65–74	4008	25.57	0.30	0.19	0.49	< 0.001	0.8
≥ 75	3848	22.63	0.08	0.03	0.25	< 0.001	0.22
Sex	14,655
Male	5881	42.24	1.00	-	-	-	1.54
Female	8774	57.76	1.58	1.12	2.25	0.010	3.44
Kessler score	14,655
K6 = 0	6176	42.13	1.00	-	-	-	0.32
K6 = 1–6	6203	42.71	3.50	1.74	7.02	< 0.001	1.45
K6 = 7–12	1637	10.98	15.78	8.03	31.01	< 0.001	8.34
K6 = 13–18	496	3.29	40.42	19.58	83.43	< 0.001	21.83
K6 = 19–24	143	0.89	54.58	22.09	134.88	< 0.001	28.09
Insurance status	14,655
Insured	13,912	94.88	1.00	-	-	-	1.95
Uninsured	743	5.12	3.49	2.47	4.92	< 0.001	15.45
Socioeconomic status (SES)	14,655
SES < 1.00	649	3.71	1.00	-	-	-	9.51
SES 1.00–1.99	3560	19.98	1.18	0.75	1.85	0.484	4.83
SES 2.00–3.99	4814	32.47	0.85	0.53	1.38	0.517	2.55
SES > 4.00	5632	43.84	0.51	0.29	0.89	0.018	1.12
1B
Insurance status	14,655
Insured	13,912	94.88	1.00	-	-	-	1.95
K6 ≥ 13 vs. K6 ≤ 13	-	-	7.20	4.77	10.86	< 0.001	17.91 vs. 1.33
Uninsured	743	5.12	3.49	2.47	4.92	< 0.001	15.45
K6 ≥ 13 vs. K6 < 13	-	-	8.68	4.45	16.93	< 0.001	50.49 vs. 10.15
1C
Socioeconomic status (SES)	14,655
SES < 1.00	649	3.71	1.00	-	-	-	9.51
K6 ≥ 13 vs. K6 < 13	-		6.16	2.70	14.02	< 0.001	26.67 vs. 5.93
SES 1.00–1.99	3560	19.98	1.18	0.75	1.85	0.484	4.83
K6 ≥ 13 vs. K6 < 13	-	-	6.56	3.85	11.18	< 0.001	26.39 vs. 3.07
SES 2.00–3.99	4814	32.47	0.85	0.53	1.38	0.517	2.55
K6 ≥ 13 vs. K6 < 13	-	-	12.31	6.53	23.22	< 0.001	22.24 vs. 1.74
SES ≥ 4.00	5632	43.84	0.51	0.29	0.89	0.018	1.12
K6 ≥ 13 vs. K6 < 13	-	-	9.02	3.85	21.14	< 0.001	15.12 vs. 0.88

1A represents the entire cohort. In addition to the variables shown, the model was also adjusted for highest education level attained (less than high school [referent] vs. some high school vs. high school diploma/GED vs. some college vs. college degree or greater), US geographic region (Northeast [referent] vs. Midwest vs. South vs. West), self-reported health status (excellent vs. very good vs. good vs. fair vs. poor [referent]), time from cancer diagnosis (≤ 5 years [referent] vs. ≥ 6 years), birth status (US born [referent] vs. non-US born), and comorbidity score (0 [referent] vs. 1 vs. 2 vs. 3 vs. 4 vs. 5 comorbid conditions), none of which was significant. 1B and 1C show the results of multivariable logistic regression stratified by insurance status and socioeconomic status, respectively. Each of these models was also adjusted for all of the demographic characteristics listed above