To scan or not to scan? Examining the controversial issue of performing neuroimaging in adolescent patients presenting to a tertiary psychiatric inpatient unit

BACKGROUND: Imaging techniques such as computerised tomography (CT), magnetic resonance imaging (MRI) and single photon emission computed tomography (SPECT) scans are used in various clinical and diagnostic neuropsychiatric assessments. However, these investigations may be costlier when compared to their clinical utility. AIM: To examine the clinical utility of neuroimaging in an acute adolescent psychiatric inpatient population admitted to Tygerberg Hospital between January 2012 and December 2013. SETTING: The study was conducted at a tertiary level adolescent psychiatric inpatient unit at Tygerberg Tertiary Hospital, Parow, Cape Town, Western Cape, South Africa. METHOD: A retrospective chart review was conducted to gather data from 125 inpatient adolescents who had neuroimaging performed during admission. Clinical information was obtained from folders and collated with neuroimaging data. The Pearson Chi-squared test was used to test for correlations between clinical variables and the outcomes (abnormalities) of CT scans. There were too few MRI or SPECT scans to warrant statistical testing for these modalities. RESULTS: Out of the total CT scans performed (n = 120), 11 (9.2%) were clinically significant or pathological. Five cases (4.2% of all CT scans) resulted in a change in diagnosis and management. There was no association between clinical variables and clinically relevant CT abnormalities (n = 11). There were three MRI abnormalities (30%), with two resulting in changes in management. Single photon emission computed tomography scans revealed abnormalities in all 10 cases. CONCLUSION: Routine neuroimaging in this population of psychiatric adolescents has high clinical utility. However, the decision to conduct structural neuroimaging should be guided by good clinical assessment. Single photon emission computed tomography scanning is useful for detecting underlying neurophysiological abnormalities in patients presenting with psychiatric and behavioural symptoms to potentially aid diagnosis and for interventional purposes.