Abstract / Summary
Electroconvulsive therapy (ECT) is effective for severe mental disorders, but cognitive adverse effects are heterogeneous and poorly captured by general cognitive screening tools. We evaluated the longitudinal clinical utility of the Electroconvulsive Therapy–Cognition Inventory (ECT-CI), an ECT-specific bedside monitoring tool. In this prospective real-world cohort, 136 adults undergoing acute ECT completed the ECT-CI and Montreal Cognitive Assessment (MoCA) at baseline, within 48 h after the acute course, and at 3-month follow-up. Clinician-adjudicated cognitive impairment, assessed by blinded psychiatrists using a DSM-5-TR-based framework, was the reference standard. Longitudinal score change, domain-level change, discrimination against the clinician-adjudicated reference standard, prognostic discrimination for 3-month clinician-adjudicated impairment, ECT-CI score trajectories, and exploratory inflammatory interactions with cumulative ECT charge were examined. Clinician-adjudicated cognitive impairment increased from 12.5% at baseline to 33.1% post-course and 50.7% at 3 months.ECT-CI scores decreased from baseline to the post-course assessment and partially rebounded at follow-up, with larger observed score changes than the MoCA. ECT-CI required less administration time, showed domain-level score changes in memory, executive function, and caregiver-rated cognition, and showed higher discrimination against the clinician-adjudicated reference standard than MoCA at all assessment points. Baseline ECT-CI showed greater prognostic discrimination than baseline MoCA for clinician-adjudicated cognitive impairment at the 3-month follow-up. Growth mixture modelling identified three ECT-CI score trajectories: high-stable score, acute score decline–partial rebound, and low-baseline–marked acute decline. Lower baseline MoCA score and schizophrenia diagnosis were associated with membership in the low-baseline–marked acute decline trajectory. Exploratory analyses suggested that systemic immune-inflammatory burden characterised unfavourable trajectories and that baseline CRP interacted with cumulative ECT charge. The ECT-CI showed substantial longitudinal score variation and strong clinical and prognostic discrimination in patients undergoing ECT. Exploratory findings suggest that baseline inflammatory burden may modify the association between cumulative ECT stimulation exposure and clinician-adjudicated cognitive impairment. ECT-CI showed larger longitudinal score changes than the MoCA across the peri-ECT assessment period. ECT-CI demonstrated superior discrimination against the clinician-adjudicated reference standard and greater baseline prognostic discrimination for clinician-adjudicated cognitive impairment at 3 months. Growth mixture modelling identified three distinct post-ECT ECT-CI score trajectories, supporting risk-stratified cognitive monitoring. Exploratory analyses suggested that systemic inflammatory burden may modify the association between cumulative ECT charge and clinician-adjudicated cognitive impairment.