CT scans in psychiatric patients – an exploratory study at Chris Hani Baragwanath Hospital

In psychiatric research studies, CT provides important insights into the aetiology and long-term structural abnormalities in the brains of patients with schizophrenia, mood disorders, metabolic and neurological disorders of the brain. However, the place of CT in clinical psychiatry is less clear. Disorders of the central nervous system resulting from trauma, intracranial haemorrhage, tumours and vascular abnormalities, etc. may initially manifest as psychiatric symptoms. As these symptoms may be misdiagnosed as indicating a primary psychiatric disorder, there is a danger of missing potentially treatable causes if a CT scan of the brain is not performed.

In South Africa, more than 50% of illnesses presenting in tertiary institutions are due to communicable diseases and injury.A significant percentage of these patients present with predominantly psychiatric symptoms and may warrant CT.Chris Hani Baragwanath Hospital in Soweto has a psychiatric unit with facilities for 155 inpatients and regularly admits acutely ill patients who fit this profile and who warrant CT of the head as part of their assessment.Because of the high cost of this procedure and the limited resources, it is important to have evidence-based guidelines for eligibility for CT.This study was an exploratory study to provide further information on this procedure when creating guidelines.

Procedure
The study was approved by the Human Research Ethics

Criteria for eligibility
The following patients were eligible for CT scanning: (i) all those presenting with a first episode of psychosis (with or without mood features); and (ii) all psychotic patients (with or without mood features) with either features of a delirium, some focal physical or neurological signs, and/or abnormal results of special investigations.

Statistical analysis
Descriptive statistics were computed as counts and percentages, minimum and maximum, and means (standard deviations (SD), 95% confidence intervals (CIs).The chi-square (χ

Results
During this 6-month period approximately 600 patients were screened and a total of 55 were found to be eligible for X-ray CT.All 55 patients gave written consent and were included in the study.Twenty of these patients (36.4%) showed some abnormal findings, and the details are listed in Table I.
The mean age of the study population was 38.3 years (SD 16.3, 95% CI: 33.9 -42.7, minimum 18, maximum 73 years).Seven patients (35%) with abnormal CT were over the age of 60 years compared with only 1 patient in the group with a normal scan (χ 2 =17.2, p = 0.0006).There was also a significant correlation between abnormal CT and advancing age (r = 0.5, p < 0.001) (Fig. 1).

Discussion
In this study the yield of abnormal scans was significantly high (36.4%).Similar high yields have been reported in other studies: 86% of 37 patients 8 and 57% of 323 patients. 9In contrast, Agzarian et al. 10 reported that 95% of 397 patients at an acute tertiary service were within normal limits and McClellan et al. 11 reported that 88.1% of all inpatients during a 3-year period were within normal limits.Both studies with low yields reported that in the few patients with abnormal findings the main finding was cortical atrophy.We suggest that if scanning is done routinely the yield would be considerably lower; however, if as in this study the patients were screened adequately, the yield would probably be higher.
A significantly large percentage of our patients with abnormal, CTs were over the age of 60 years and there was a positive correlation between abnormal CTs and advancing age.Similar high yields of abnormal CTs in first-episode psychosis patients in this age group have been reported in other studies. 12e factor most predictive of abnormal CT in our patients was a first episode of psychosis (32% of all patients with first-episode psychosis, and 85% of all patients with abnormal scans).Previous studies 2,13,14 also recommend that all patients presenting with a first onset of psychotic illness should receive CT of the head to rule out such causes as tumours, abscesses, Huntington's disease, encephalitis, Wilson's disease, and trauma.Yet other studies 15 suggest that scanning is indicated if there is an atypical presentation or inadequate response to standard treatment in such patients.
Some researchers 11,12,16 recommend that only mentally ill patients with neurological abnormalities (namely focal signs, seizures, head injuries, etc.) and/or abnormal special investigations (EEG, blood tests etc.) 4 should be eligible for CT.This is similar to the guidelines of the Chris Hani Baragwanath unit.However, in our group of patients only a significant minority of the patients with abnormal CTs had a concurrent abnormal physical examination and/or special investigations.This suggests that if we use these criteria for scanning strictly, we will most probably miss many abnormalities.This view is supported by Colohan et al, 17 who also report a poor correlation between abnormalities on CT scan and findings on physical examination, laboratory testing, EEG and psychological testing.It is recommend that in first-episode for clarification or to complement other findings, 1 and that utilising all of these factors to guide the ordering of CTs would greatly increase the yield of the procedure for psychiatric patients, without excess medical morbidity. 5is study is limited in its generalisability because the study population were attendees of a tertiary academic hospital and represent a select group of severely ill patients.The small sample size may have statistical limitations in terms of ascertaining meaningful differences when comparing groups.Notwithstanding these limitations the findings were significant and different from published data and warrant further investigation into the role of CT scanning in mentally ill patients.

Conclusion
The study revealed that CT scans in this psychiatric population yielded a significant number of abnormalities, especially in patients with first-episode psychosis.This study also suggests that clinical abnormalities (physical and laboratory) may not be reliable predictors of abnormal CTs.It is recommended that further studies using a larger sample size be conducted to determine if guidelines should include all patients with first-episode psychosis despite the high cost and limited availability of this procedure.
Committee (HREC) of the University of the Witwatersrand.The procedure was fully explained to each patient and written consent was obtained.If the patient was eligible for CT but did not have the capacity to understand and give written consent, this was obtained from alternative sources (family or hospital superintendent).The patients' demographic data (age, race and gender), clinical details (physical, mental state) and all other special investigations conducted (laboratory, radiological and EEG) were recorded by the investigators.Thereafter the patients underwent CT and the scan was read by a radiologist who was blind to the patients' history and initial diagnosis.