#7199 Indications for MR neuroimaging in first episode psychosis

<h3>Background</h3> In certain cases, psychosis may have underlying organic causes that are identifiable by neuroimaging. However, due to uncertain estimates of the prevalence of treatable lesions in psychotic populations, NICE does not recommend the routine use of neuroimaging to assess patients presenting with first episode psychosis. Nevertheless, selective imaging to exclude organic causes of psychosis based on history or examination findings which suggest possible organic aetiology is commonly undertaken. In this study, we examined the indications for structural imaging using MRI in a clinical cohort presenting with psychosis. <h3>Methods</h3> Using a database (Clinical Record Interactive Search) of anonymised patient records for the South London and Maudsley NHS Trust, we identified all records of individuals with a first primary diagnosis of a psychotic disorder between 2007 – 2021 who were referred for an MRI scan within 18 months of the diagnosis. Two psychiatrists screened these records for the indication for each MRI scan and, when indications were found, grouped them into one of ten categories. Any discrepancies between the two assessors’ classification were checked by a third assessor to resolve them. Descriptive statistics were obtained. <h3>Results</h3> Records for 1,749 patients (53.9% male, mean age 37) were included in the analysis. The number of patients receiving an MRI for a specific indication were: 204 (11.7%) for an atypical psychotic presentation, 191 (10.9%) for cognitive impairment, 103 (5.9%) for neurological features, 64 (3.7%) for head injury, 36 (2.1%) for suspected encephalitis, 31 (1.8%) for a suspected space-occupying lesion, 30 (1.7%) for headaches, and 27 (1.5%) for hyperprolactinaemia. In 693 (39.6%) of cases, the indication was routine screening for an organic cause of psychosis, with no other concerning feature mentioned. In 370 (21.2%) of cases, no indication was found. Of the 1537 individuals with available scan results, 1139 (74.1%) had a normal result, and 149 (9.7%) led to a change in management. Of the 625 scans that were done as part of routine screening, 532 (85.1%) were normal and 30 (4.8%) led to a change in management. <h3>Conclusions</h3> In this real-world clinical sample, more than one-third of MRI scans performed in patients with psychosis were carried out as part of routine screening, in contrast to NICE recommendations. Nevertheless, almost one in twenty routine scans led to a change in management. These findings highlight the importance of clinicians remaining attuned to concerning aspects of patients‘ presentations which may be associated with scan abnormalities.