More or less equal? Trends in horizontal equity in mental health care utilization in Stockholm county, Sweden (2006-2022). Repeated survey-registry linked studies

Abstract

Background: Horizontal equity is defined as equal care for equal needs, regardless of socioeconomic factors. This study investigated trends in horizontal equity in mental health care (MHC) utilization in Sweden from 2006 to 2022. Monitoring equity provides valuable information for healthcare system governance (e.g., planning and resource allocation) necessary for ensuring equitable provision of services.

Methods: A total of 81,650 Stockholm residents aged 18-64, who participated in the Hälsa Stockholm surveys of 2006, 2010, 2014 or 2021, were analysed. Their subsequent use of MHC (primary, in- and outpatient specialized care, and psychotropic medication) within six months after survey response was collected from registries between 2006 and 2022. Concentration index (CI) and need-standardized CI (Horizontal inequity index, HI), summative measures of inequalities, were used in this study. HI was estimated using self-reported psychological distress (measured with the General health questionnaire 12 in 2006-2014 and Kessler 6 in 2021) as the primary need indicator, with general health status and long-term limiting illness as additional need indicators. Equivalized disposable household income was used as the ranking variable, while education status, migration status, age, and sex were included as non-need variables that we controlled for in the analyses.

Results: Lower-income individuals used MHC services more than their higher-income counterparts with comparable levels of psychological distress. These "pro-poor" inequities in the probability of MHC use increased from HI = -0.057 [95% Confidence Limits, CL: -0.079, -0.034] in 2006/2007 to HI = -0.130 [95% CL: -0.159, -0.102] in 2014/2015. By 2021/2022, the "pro-poor" inequities had decreased (HI = -0.034 [95% CL: -0.06, -0.009]), partly due to an increase in MHC use among higher-income groups but a decrease in the lowest income group. Standardizing for additional need indicators reduced the "pro-poor" inequities but maintained the observed trends. Among non-Nordic migrants, "pro-rich" inequities fell between 2006/2007 and 2014/2015 but rose in 2021/2022, with significant "pro-rich" inequities among non-European migrants in 2021/2022 (HI = 0.100 [95% CL: 0.024, 0.176]). Among patients in outpatient services, "pro-poor" inequities in visit frequency decreased over time (2006-2022).

Conclusion: We observed increasingly higher probability of MHC use among lower-income individuals than their higher-income peers with similar (measured) needs from 2006 to 2015. However, during the pandemic (2021/2022), potential access problems led to diminishing of "pro-poor" inequities in the total sample, and to "pro-rich" inequities among non-Nordic migrants. The Covid-19 disruption to the healthcare system-such as restrictions on in-person visits and the rapid transition to digital healthcare services-along with its impact on care-seeking, may explain the trend shifts.

Keywords: Concentration index; Covid-19; Healthcare reforms; Mental disorders; Migrants.

Fig. 1

Derivation of the analytic sample. Individuals participated in one of the Stockholm Public health surveys “Hälsa Stockholm” between 2006 and 2021. Their healthcare records were collected from registries 6 months from survey-response between 2006/2007 and 2021/2022

Fig. 2

Concentration Curves showing the income-related inequalities in the probability of psychological distress. Distress defined as scoring ≥ 3 on the GHQ-12 (2006–2014) or ≥ 8 on the Kessler 6 (2021). *** - P-value < 0.0001. Curves above the Line of Equality indicate higher concentration of distress in individuals with lower income, “pro-poor inequalities”. The closer the curve is to the Line of Equality, the more equal the distribution of psychological distress among individuals with varying income

Fig. 3

Concentration Curves (CC) showing the distribution of need-standardized MHC use across the income rank. Models standardized for psychological distress (as the only need-indicator: distress measured using the GHQ-12 in 2006–2014 and Kessler 6 in 2021) and controlled for education, sex, age-group and migration status. The closer the curve is to the Line of Equality, the more equal the distribution of need-standardized MHC use among individuals with varying income

Fig. 4

Horizontal inequity indices in MHC use over time, shown for the overall sample and among non-Nordic migrants. Negative values indicate higher MHC use in lower income individuals in comparison to those with higher income, “pro-poor inequities”. Point estimates and 95% confidence limits (CL) shown for models standardizing for psychological distress and models standardizing for all need indicators. Models controlled for education, sex, age-group and (migration status in the overall sample)