Clinical decision support systems in psychiatric drug treatment: a systematic review on current use and treatment impact

Background

Clinical decision support systems (CDSS) have demonstrated value in improving quality of care and guideline adherence across somatic medicine, yet their application in psychiatric pharmacotherapy remains underexplored. Psychiatric drug treatment is characterized by complexity, including polypharmacy, comorbidities, and variable treatment response, alongside inconsistent adherence to monitoring and prescribing guidelines. This systematic review synthesizes current evidence on CDSS designed to support psychotropic drug prescribing, monitoring, and follow-up in psychiatric populations, and identifies key determinants of successful implementation.

Methods

A systematic review was conducted in accordance with PRISMA guidelines and registered in PROSPERO (CRD42024462820). PubMed, EMBASE, PsycINFO, and Scopus were searched from 1998 to March 2024. Eligible studies evaluated knowledge-based CDSS interventions in psychiatric settings, focusing on pharmacological treatment and including a comparator condition. Two reviewers independently screened studies, extracted data, and assessed methodological quality using design-appropriate tools. CDSS quality and implementation characteristics were additionally evaluated using an adapted GUIDES checklist.

Results

From 11,203 records, 10 studies met inclusion criteria. Studies targeted depressive disorders (n=4), psychotic disorders (n=3), opioid use disorder (n=2), and ADHD (n=1), with sample sizes ranging from 18 to 8,581 participants. CDSS functionalities included support for treatment initiation (n=8), diagnostic assessment (n=4), adherence monitoring (n=2), shared decision-making (n=2), and population-level risk screening (n=2). Systems varied in architecture, with web-based platforms most common (45%), and limited integration into electronic health records (EHRs) (36%).

Seven studies reported care-related primary outcomes, of which most demonstrated statistically significant improvements compared to usual care. Effects were generally modest, including small reductions in symptom severity, earlier initiation of appropriate treatment, and improved adherence to monitoring guidelines. CDSSs appeared particularly effective in enhancing process-related outcomes, such as timely follow-up and guideline-concordant prescribing, rather than producing large changes in clinical endpoints.

Methodological quality was heterogeneous, with most studies rated fair to poor. Lower-quality studies tended to report larger effect sizes, raising concerns about potential bias. Substantial variability in study design, outcome measures, and CDSS functionality precluded quantitative synthesis.

Implementation of CDSS in psychiatric care is facilitated by integration into electronic health records, user-friendly (e.g., web-based) accessibility, and embedding within collaborative care models where multiple professionals share responsibility for follow-up. Clinician-related factors, particularly risk perception and perceived clinical urgency, influence adherence to CDSS recommendations. Key barriers include limited interoperability, poor workflow integration, and the narrow, non-scalable design of many systems.

Discussion

This review highlights an emerging but still limited evidence base for CDSS in psychiatric pharmacotherapy. Compared to other medical fields, where CDSS effects are more firmly established, psychiatric applications remain fragmented and insufficiently evaluated. A notable “evaluation gap” was identified: many CDSS initiatives are developed and implemented without rigorous assessment of clinical impact.

Conclusions

CDSSs show potential to improve consistency, safety, and timeliness of psychiatric pharmacological care, primarily through incremental improvements in clinical processes. However, evidence remains limited and methodologically heterogeneous. Future research should prioritize robust study designs, standardized outcome measures, and scalable, interoperable CDSS architectures. Bridging the evaluation gap will be essential to translate technological innovation into meaningful improvements in mental health care.

References

No references