¹Medical University of Vienna, Department of Psychiatry and Psychotherapy, Vienna, Austria, ²Medical University of Vienna, Comprehensive Center for Clinical Neurosciences and Mental Health, Vienna, Austria, ³Karl Landsteiner University, Research Centre Transitional Psychiatry at the University Hospital Tulln-NOE LGA, Tulln, Austria, ⁴University Hospital St. Poelten, Department of Psychiatry and Psychotherapy, St. Poelten, Austria, ⁵Medical University of Vienna, Department of Neurology, Vienna, Austria, ⁶Outpatient Psychiatric Clinic, Leben Psyche Zentrum, Vienna, Austria

Background: Long COVID (LC) poses a considerable clinical and socioeconomic burden on affected patients and their surroundings (1-3). LC presents with diverse symptom clusters (4). One of the most prevalent is the neuropsychiatric symptom cluster. Symptoms related to this cluster are chronic fatigue, cognitive dysfunctions, symptoms related to depression and anxiety including sleep disturbances, inner tension, pain, hypersensitivity to acoustic and visual stimuli, as well as vegetative dysautonomia (5). At the time of writing no definitive treatment for LC has been approved but different treatment approaches are currently being investigated (6-8). Heterogeneous LC diagnostic criteria and trial designs, including variable symptom surveys and cognitive or functional assessments, limit cross-trial comparability and complicate evidence synthesis.

Aims & Objectives: This analysis aims to describe and quantify trial design across pharmacological interventional trials addressing neuropsychiatric LC outcomes.

Method: This is a registry-based cross-sectional analysis of ClinicalTrials.gov records registered between 01 Jan 2020 and 11 Feb 2026 using the query (“Long COVID” OR “Post-COVID-19 Syndrome” OR “Long Covid19” OR “Long COVID-19 Syndrome”). We include interventional trials in adults (≥18 years) investigating pharmacological agents and assessing neuropsychiatric outcomes (defined by prespecified neuropsychiatric outcome domains and instruments listed as primary/secondary endpoints). Extracted variables comprise design features (phase, randomization, blinding), eligibility criteria, and outcome domains/instruments. Outcomes are grouped by domain and instrument to describe heterogeneity.

Results: 558 interventional clinical trials listed in ClinicalTrials.gov have been identified. These are either completed, ongoing or planned trials. All identified trials are reviewed according to the previously described criteria to ascertain the number of trials retained for further analysis. The included trials designated outcomes, inclusion and exclusion criteria and selected variables are grouped based on similarities in study design and frequency of occurrence. Based on the broad variety of debilitating symptoms observed in LC (5) and the suggested complex pathophysiology of neuropsychiatric LC (9), a broad variety of outcome assessment tools is used in the current clinical trial landscape of neuropsychiatric LC.

Discussion & Conclusions: Pharmacological intervention trials addressing neuropsychiatric LC show substantial heterogeneity in eligibility criteria and outcome measurement, with previous initiatives establishing some core outcome domains but remaining undecided on most domains. Corresponding test instruments for some of these outcome domains have been proposed (10, 11)(12). Aligning future trials with a minimum harmonized outcome set (core domains plus recommended instruments), while retaining intervention-specific mechanistic endpoints, could improve comparability, facilitate meta-analytic synthesis, and accelerate identification of clinically meaningful, patient-centered treatment effects.