33 Income inequalities beyond access to mental health care Depressive disorders were the most frequent primary diagnosis (234,126 patients [24.6%]), followed by anxiety disorders (167,467 [17.6%]), personality disorders (79,776 [8.4%]), disorders first diagnosed in childhood (75,689 [8.0%]), and substance use disorders (71,668 [7.5%]). The most common initial GAF category was 6 (score 51–60, 435,731 [45.8%]), followed by 5 (score 41–50, 246,730 [25.9%]), and 7 (score 61–70, 156,056 [1G.4%]). 79,946 (8.4%) of patients had an initial GAF of 4 or lower, whereas 33,067 (3.5%) had an initial GAF of 8 or greater. Patients received on average 1,357 treatment minutes (SD 1,702), with a right skewed distribution (Figure A2). By the end of the initial treatment record, 585,634 (G1.5%) of the patients remained stable in their functioning, 337,317 (35.4%) had a functional improvement, and 28,579 (3.0%) showed functional deterioration (Figure A3). The most common reasons for record closure were mutual agreement (419,550 [44.1%]) and administrative reasons (341,795 [35.9%]), and 301,718 (31.7%) of the patients had an additional treatment record. Table 2 shows that patients with lower income started treatment with greater disease severity compared with the quintile above. On average, the baseline GAF score for patients in the lowest income quintile was 5.545. In contrast, patients in the highest quintile had on average a baseline GAF score of 5.898, which is 0.353 points higher than those in the lowest quintile (95% CI 0.347–0.360). The comparison between the model controlling for demographic factors and the fully adjusted model shows that the income gradient is robust to adjusting for primary diagnosis. Average treatment minutes decreased with income when not adjusting for need. Patients in the lowest- income quintile received on average 1,490 min of treatment. The reduction in treatment time ranged from 4.1% (95% CI 3.5–4.6) for the second lowest quintile to 12.2% (11.5–12.8) for the highest quintile. Notably, once we controlled for severity and diagnosis, patients in the lowest income quintile had the lowest treatment intensity, albeit with small differences. Coefficients show that treatment minutes were 2.5% (2.0–3.1) in the second quintile, 2.8% (2.2–3.4) in the third quintile, 2.6% (2.0–3.2) in the fourth quintile, and 1.8% (1.1–2.4) in the fifth quintile higher than the minutes received by the lowest income quintile patients (Table 2). This reversal suggests that the greater need of patients with low income explained the higher treatment minutes received, whereas patients with high income received slightly more treatment time once need was accounted for. The likelihood of functional improvement increased substantially across income quintiles, with barely any changes when adjusting for need and treatment time. The likelihood of having functional improvement for patients in the highest income quintile was 17.3 pp greater than for those in the lowest quintile (95% CI 17.0–17.6; Table 2). This finding represents a more than 50% increase compared with the 25.8% of patients in the lowest quintile who had a functional improvement. Alternative definitions of the outcome corroborated that patients with high income were more likely to experience improved functioning by the end of the first treatment record (Table A5). 2
RkJQdWJsaXNoZXIy MjY0ODMw