Wilfred Dang BS, Presenter: Nothing to Disclose

Sabina Imran Khan MBBS, Abstract Co-Author: Nothing to Disclose

James P. Ross BSc, Abstract Co-Author: Nothing to Disclose

Santanu Chakraborty FRCR, DMRD, Abstract Co-Author: Speakers Bureau, Merck KGaA Speakers Bureau, Novartis AG Grant, Bayer AG

The low diagnostic yield associated with the usage of CT/MRI imaging in first episode psychosis (FEP) has caused a change in British guidelines to decrease redundant neuroimaging. However, imaging for FEP is still common in North America and remains a point of controversy. The purpose of this study is to assess current literature regarding the diagnostic value for FEP neuroimaging. This will help us determine the necessity of head CTs/MRIs in clinical practice for FEP patients. 

Electronic searches were performed in MEDLINE, PSYCHINFO and EMBASE in November 2013. Search criteria consisted of: “Psychotic Disorders”, “Psychosis”, “MRI”, “CT”, “Predictive Value of Tests”, “Differential Diagnosis”, “Biomedical Technological Assessment” and syntax derivatives of these search terms. After duplicates were removed, two reviewers screened 545 articles for inclusion. Studies included needed to meet the following: a) Patients must be presenting with FEP, b) Patients had a CT or MRI at the time of presentation, c) Studies must have reported abnormal or normal image results, and d) The rationale of the scan must be reported. Disputed articles between reviewers were resolved by a senior neuro-radiologist. All statistical analyses were performed using the Comprehensive Meta-Analysis Software. The event rate was computed as the number of patients with abnormal radiological findings possibly accountable for psychosis over the total number of patients that underwent imaging. Event rates for CT and MRI were summarized in separate forest plots, with the corresponding 95% confidence interval for each study.

Preliminary results from 8 abstracted studies show that, out of 1,019 CT/MRI scans, 838 scans were completely normal, whereas some abnormalities were observed in only 181 scans. Most abnormalities seen were either benign or incidental and did not have any impact on patient management. The calculated overall rate of abnormal findings that accounted for psychosis was 0.9 % (95% CI: 0.5%-1.9%). 

Given the low diagnostic yield of neuroimaging for FEP, more consideration needs to be given to ordering a head CT/MRI for FEP due to radiation and resource utilization concerns.

Neuroimaging for FEP has very little value in determining the cause for psychosis and should be re-evaluated in psychiatric guidelines.

Dang, W, Khan, S, Ross, J, Chakraborty, S, Appropriate Imaging Utilization: A Meta-analysis of Neurological Imaging in First Episode Psychosis.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14005669.html