Sarah Kayser, Christian Plewnia, Andreas J Fallgatter
In this preliminary retrospective evaluation, the CGI-STIM was feasible to apply to routine clinical documentation and showed strong interrater reliability. These findings do not establish validity, prospective reliability, sensitivity to change, or clinical utility. Future prospective studies in larger ECT samples are required to evaluate these measurement properties and the potential utility of the CGI-STIM for routine outcome monitoring.
OBJECTIVES: Electroconvulsive therapy (ECT) is highly effective for severe and treatment-resistant mental disorders, but concerns about cognitive side effects and enduring stigma continue to hinder its broader acceptance. Although treatment response and cognitive tolerability are both clinically important, their systematic assessment and documentation remain challenging in routine clinical practice. To address this gap, we developed the Clinical Global Impression-Stimulation Outcome Monitoring Tool (CGI-STIM), a pragmatic clinician-rated tool that integrates both domains into a single clinically interpretable outcome classification.
METHODS: We retrospectively analyzed 21 ECT series in 15 patients. Two independent clinicians applied the CGI-STIM using routine clinical documentation to classify treatment outcomes based on clinical response and cognitive tolerability. Feasibility and interrater agreement were evaluated.
RESULTS: The CGI-STIM was successfully applied to all 21 ECT series. Outcome classification was feasible in all cases using routinely available clinical information. Average completion time was approximately 5 min. Interrater agreement was excellent (quadratically weighted Cohen's κ = 0.84). The resulting classifications reflected varying combinations of clinical improvement and cognitive tolerability, illustrating the instrument's ability to capture multidimensional treatment outcomes.
CONCLUSIONS: In this preliminary retrospective evaluation, the CGI-STIM was feasible to apply to routine clinical documentation and showed strong interrater reliability. These findings do not establish validity, prospective reliability, sensitivity to change, or clinical utility. Future prospective studies in larger ECT samples are required to evaluate these measurement properties and the potential utility of the CGI-STIM for routine outcome monitoring.