Correlates of outcomes in routine clinical care in patients suffering from severe mental illnesses

Background

Mental disorders rank among the top ten global contributors to disability, with estimates suggesting an annual global cost of $4.7 trillion. For individuals, mental disorders often result in profound functional impairments, significant distress, and markedly reduced life expectancy. In mental health research, randomized controlled trials (RCTs) are the gold standard for assessing treatment efficacy. However, there has been growing recognition of the need to complement RCTs with real-world data studies, which analyze outcomes in routine clinical settings. The Scandinavian countries are uniquely positioned for such research due to extensive public health registries and the ability to link data via individual civic registration numbers. This thesis investigates associations between aspects of care delivery and key clinical outcomes in populations with severe mental disorders treated in routine healthcare settings.

Methods and results

Paper 1 explored the relationship between aspects of court-mandated forensic psychiatric care and body mass index (BMI) in a cohort of 3,389 individuals treated between 2009 and 2020. Data were retrieved from the Swedish Forensic Psychiatric Register, including BMI, demographic, clinical, and treatment-specific variables. Linear and spline multilevel regression analyses revealed a 1.1% annual increase in BMI during the first four years of care (p < .001, R2 = . 005), with no significant changes thereafter. Factors associated with BMI increases included antipsychotic treatment, female sex, younger age at admission, outpatient care, and access to an external support person.

Paper 2 examined the relationship between inpatient care duration and relapse into care in individuals with eating disorders (ICD-10 F50.X). Using open-access data from the Swedish Patient Registry (1998-2020), variables analyzed included annual inpatient episodes, unique patients, and total inpatient days. Mean length of stay was calculated as total inpatient days divided by care episodes. Robust linear regression demonstrated an inverse correlation between mean length of stay and relapse in both adolescents (adjusted R2 = 0.5216, p < . 001) and adults (adjusted R2 = 0.5090, p < . 00001).

Paper 3 assessed the association between regional bipolar diagnosis rates and adolescent suicide mortality (15-19 year olds) in Sweden (2008-2021). Data from the Swedish National Board of Health and Welfare included bipolar diagnosis rates, suicide rates, lithium dispensation rates, psychiatric care affiliation rates, and inpatient/outpatient visit proportions. Generalized mixed regression revealed an inverse association between bipolar diagnosis rates and male adolescent suicide mortality ( = - 0.00429; 95% CI, -0.0081 to -0.0004; p = . 03), independent of lithium dispensation and psychiatric care affiliation rates. This association persisted in beta-binomial models and after adjusting for schizophrenia and depression diagnosis rates.

Paper 4 investigated the relationship between utilization of advanced psychiatric treatments (lithium, clozapine, electroconvulsive therapy) and adolescent excess suicide mortality (AESM). Open-access data (2016-2020) from the Swedish National Board of Health and Welfare and the Swedish ECT Registry were analyzed. AESM was calculated as the difference between adolescent (15-19 years) and young adult (20-24 years) regional suicide rates. Min-max normalization was applied to treatment variables and AESM to reduce outliers. Aggregated utilization of all three treatment modalities was inversely associated with AESM in adolescents overall (B = - 0.613, p = . 0003) and in males (B = - 0.404, p = . 009) but not in females (p = . 197). Post-hoc analyses identified specific associations between lower AESM and utilization of electroconvulsive therapy and lithium.

Conclusions

The findings underscore the potential value of evidence-based and standardized approaches in psychiatric care. Study 1 suggested the need for targeted interventions to address metabolic health, particularly among women, in forensic psychiatric care, while Study 2 provided ecological support for a weight stabilization phase in eating disorder inpatient care as a possible strategy to reduce readmission risk. Study 3 highlighted an inverse association between bipolar disorder diagnosis rates and suicide rates in male adolescents, suggesting that early and accurate diagnosis could play a role in suicide prevention, though no similar association was observed for females. Study 4 indicated that advanced treatments, such as lithium and ECT, might contribute to reduced suicide mortality in adolescents. Across all studies, significant variability in clinical practices and outcomes was observed, pointing to possible systemic gaps in care. While these findings highlight areas for further investigation and potential clinical focus, the correlational nature of the research precludes definitive causal conclusions. Rigorous future research, encompassing both controlled trials and well-designed real-world studies, is essential to clarify these relationships and guide evidence-based improvements in care.

List of scientific papers

I. Andersson, P., von Schreeb, A., Johansson, L., Sturidsson, K., Wetterborg, D., & Sorjonen, K. (2024). Changes in Body Mass Index During Mandatory Forensic Psychiatric Care: Findings from a Long- Term (2009-2020) Cohort Study Based on Swedish Registry Data. International Journal of Forensic Mental Health. 23(2), 106-116.
https://doi.org/10.1080/14999013.2023.2214381

II. Andersson, P., Jamshidi, E., Ekman, C., Tedroff, K., Björkander, J., Sjögren, M., Lundberg, J., Jokinen, J., & Desai Boström, A. E. (2023). Mapping length of inpatient treatment duration and year-wise relapse rates in eating disordered populations in a well-defined Western-European healthcare region across 1998- 2020. International Journal of Methods in Psychiatric Research. 32(4), e1960-n/a. https://doi.org/10.1002/mpr.1960

III. Andersson, P., Jokinen, J., Jarbin, H., Lundberg, J., & Desai Boström, A. E. (2023). Association of Bipolar Disorder Diagnosis With Suicide Mortality Rates in Adolescents in Sweden. JAMA Psychiatry (Chicago, IlI.). 80(8), 796-802. https://doi.org/10.1001/jamapsychiatry.2023.1390

IV. Desai Boström, A. E., Andersson, P., Rask-Andersen, M., Jarbin, H., Lundberg, J., & Jokinen, J. (2023). Regional clozapine, ECT and lithium usage inversely associated with excess suicide rates in male adolescents. Nature Communications. 14(1), 1281-1281. https://doi.org/10.1038/s41467-023-36973-4