An observational study of CT scanning in psychiatric patients

Transcription

1 Mwaffaq Elheis FRCR, Abhaya Gupta MRCP, Kalpana Pansari MRCPsych, Mobinulla Syed FRCS An observational study of CT scanning in psychiatric patients Opinion is divided on the proper use of brain scans in psychiatry. In this study, the authors reviewed CT scanning in common psychiatric conditions in a district hospital. In their study about 64 per cent of brain scans showed some abnormality. CT scanning influenced patient diagnosis, prognosis and treatment. In addition, almost 3 per cent of patient scans identified potential reversible, previously unknown, intracranial pathology. Certain disorders of the brain may present initially or solely with psychiatric signs and symptoms. 1 The possibility that some of these may be reversible has led to the use of brain imaging in psychiatric practice. A number of published studies have addressed the issue of proper use of brain computerised tomo - graphy (CT) imaging in psych - iatry. 2-9 The recommendations of these studies varied, from advocating imaging scans as a screening procedure for all patients, risking low yield, to restricting the recommendation to scan only those patients with clear focal neurological abnormalities on examination, accepting the risk of missing a rare diagnosis. Hence, there is a wide divergence of opinion about the proper utilisation of brain scans in psychiatric conditions. Moreover, CT scans are now available in most UK hospitals. They are a relatively cheap, quick, sensitive imaging test for the majority of brain lesions, are easily accessible, and are increasingly being requested for psychiatric patients. CT is a diagnostic procedure that if used unselectively may result in the discovery of incidental findings that may have important implications. In contrast, new more sensitive brain imaging techniques have become available such as: magnetic resonance imaging (MRI), positron emission tomography Primary psychiatric Number of Normal scan Abnormal scan diagnosis patients number (%) number (%) Dementia 38 5 (13%) 33 (87%) Mood disorder (46%) 16 (54%) Schizophrenia 14 9 (64%) 5 (36%) Behaviour disorder 10 2 (20%) 8 (80%) Personality disorder 7 4 (57%) 3 (43%) Anxiety disorder 3 3 (100%) 0 (0%) Total (36%) 65 (64%) Table 1. Frequency and percentage of normal and abnormal scans in relation to psychiatric diagnostic categories (PET) and single photon emission computed tomography (SPECT), which although being more sensitive for a variety of conditions are costly, less easy to perform and are not available widely. Our study was carried out to determine the current practice of CT usage among psychiatric patients in a district hospital in the UK. We wanted to study the indications for usage, the results of CT scans repor ted and factors that may influence positive results. We assessed whether the scan results influenced the diagnosis and management of these patients. Method We included consecutive patients referred for CT brain scanning for primar y psychiatric diagnosis between Januar y and December 2003 in the Depar tment of Radiology at West Wales Hospital, Carmarthen. The indications for referral for CT brain were based upon the clinical judgement of the psychiatrist. CT scans were performed using the most recently available scanner equipment in the Depar tment. The scan repor ts were read by a radiologist to decide whether the scan was within normal limits, or whether it was abnormal, and the type of abnormality was recorded. We analysed the medical records of all patients and collected data for specific parameters including age, sex, primar y psychiatric diagnosis, duration of symptoms, significant past medical histor y, presence or absence of confusion, abnormal neurological signs and previous history of head injury or 24 Progress in Neurology and Psychiatry

2 Primary psychiatric Frontal Generalised brain Brain GBA and Infarct GBA Other diagnosis atrophy atrophy (GBA) ischaemia ischaemia and infarct Dementia (normal pressure hydrocephalus) Mood disorder (previous haematoma) Schizophrenia Behaviour disorder Personality disorder 3 Anxiety disorder Total Table 2. Frequency of abnormal brain scan results for different psychiatric diagnostic categories seizure. All cognitive and neurological examinations were performed as part of the routine clinical evaluation by the treating psychiatrist. We collected data on CT brain scan results (normal or abnormal) and the type of scan abnormality repor ted for these patients. Following the CT scan we analysed patient records to assess the impact of CT scan results (normal or abnormal) on clinical management. Statistical analysis The data on patient characteristics and CT results were analysed using Statistical Package for the Social Sciences (SPSS) software version for Windows. We calculated the frequency and percentage of abnormal scan results among different psychiatric diagnostic categories. The occurrence and significance of neurological signs, seizure and head injury was correlated with abnormal brain scan results. We determined the percentage of patients for whom scans influenced clinician management. Results One hundred and two patients were included in the study. Their ages ranged between 33 and 96 years (mean 70.5 years). Fifty-two patients were male. In 85 per cent of patients, the scans were ordered by psychiatric teams to rule out structural lesions because of abnormal signs and/or limited symptoms. The duration of psychiatric symptoms ranged between one month and 24 years. Table 1 shows the primary psychiatric diagnoses and the percentage of abnormal and normal CT brain scans reported. Most patients referred for CT scans had dementia as their primary diagnosis. All brain scans were normal for anxiety disorders, and other psychiatric conditions showed variable numbers of abnormal scan results. The highest percentage of abnormal scans were found in patients with dementia (87 per cent). Among a total of 102 patients, 37 (36 per cent) scans were normal and 65 (64 per cent) showed some abnormality. Table 2 shows the categories of scan abnormality noted for each of the different psychiatric diagnoses. Generalised brain atrophy (GBA) was the most common brain abnormality repor ted, in 24 patients. Among dementia patients, nearly 66 per cent of abnormal scans showed GBA. Three patients with mood disorders showed evidence of haematoma with signs of resolution (subdural haematoma was known before the CT scan). Three patients with dementia showed normal pressure hydrocephalus on scan as a previously unknown diagnosis. Scans of patients with anxiety disorder were all normal. The other psychiatric conditions showed variable reported abnormalities. Table 3 shows the observed association between the presence of neurodeficit, history of seizure, history of head injury for each of the psychiatric diagnostic categories. Table 4 shows the impact of CT brain scans on the clinical management of dif ferent psychiatric conditions. 26 Progress in Neurology and Psychiatry

3 Discussion This hospital-based study revealed abnormalities on scans in 65 (64 per cent) patients. The most common reason for requesting the scan was to rule out any undetected structural brain lesion. The frequency of abnormal brain scans found in psychiatric patients reported in previous studies has been highly variable, ranging from 6.8 per cent to 53 per cent. 3,5,8,10 Such a wide variation in the frequency of abnormal brain scans among psychiatric patients in different studies has been attributed to a variety of reasons, such as variable selection of patients referred by psychiatrists, differences in interpretation of scans by radiologists, different patient populations studied, eg inpatient vs outpatient, varying duration of illnesses and different age groups. No previous study has reported such a high percentage of abnormal brain scans as found in our study, which could be due to the inclusion of a higher percentage of elderly patients (average age 70 years) compared with earlier studies, 3,5 as with advanced age, more CT brain abnormalities can be expected. The majority of patients in our study had dementia, which is a common condition in the elderly population. Moreover, the patients in our study were pre-selected by psychiatrists as needing CT brain scans on clinical grounds, and this group would be expected to have a higher frequency of CT abnormalities than a randomly selected sample of psychiatric patients. There is also an increasing expectation from patients and families to undergo tests such as CT brain scans. This could have prompted psychiatrists to refer more patients for scans in a relatively elderly population. Another reason for higher positive reporting of scans could be the notion that errors of omission are more serious Psychiatric diagnoses Neurodeficit H/o seizure H/o head injury Dementia Mood disorder Schizophrenia Behaviour disorder Personality disorder Anxiety disorder Total Table 3. Observed association between neurodeficit, history of (H/o) seizure and head injury for different psychiatric conditions (number of patients) than errors of commission. This attitude could have lead to an heroic search for positive test results. 11 As expected, all patients with anxiety disorders were relatively younger and had normal scans. Most abnormal scans were reported in patients with dementia. There was a wide variation in the percentage of abnormal scans reported for other psychiatric conditions. Several studies have described the various morphological abnormalities in brain CT associated with psychiatric disorders. 7 Progressive frontal atrophy 12 and lateral ventricular enlargement 13 have been observed in the brains of schizophrenic patients in CT scans. In dementia, it is widely accepted that the primary role of structural imaging, such as CT brain scanning, is not to identify an abnormality but to help differentiate potentially modifiable causes of dementia such as brain tumour, subdural haematoma, normal pressure hydrocephalus and vascular dementia. 14 In the Women s Longitudinal Health Study in Gothenburg, brain atrophy on CT brain was not correlated with depression in the population. 15 In our study, generalised brain atrophy (GBA) was the most common abnormality, reported in nearly a quarter of psychiatric patients. Moreover, as expected, 22 out of 33 patients with dementia showed GBA. Cerebral infarct was noted only in association with dementia. Nearly one-third of patients with mood disorders showed GBA, the true significance of which cannot be accurately determined in this study. The presence of features such as confusion, focal neurological signs, previous history of head injury or seizure in our study were not sufficiently precise to rule in or rule out an abnormal brain scan result. In our study, a large percentage of patients had neurodeficit (focal neurological deficit and/or confusional state), previous history of seizure or head injury. Presence of these features could have prompted the psychiatrists to refer the patient for a brain scan, accounting for more abnormal scan reports. In the literature, a study reported that abnormal scans correlated with focal findings on neurological examination. 3 Another study showed that when a history of neurological disorder or neurological/organic mental signs was absent then brain scans were normal in 75 per cent of cases but when both were positive scans were abnormal in 74 per cent of cases and when both history/examination and electroencephalography (EEG) was abnormal the scan were abnormal in 92 per cent of cases. 5 In our study not all patients had had an EEG, hence we could not compare these results. Progress in Neurology and Psychiatry 27

4 28 Impact Psychiatric diagnosis Number of Total patients (%) patients 1. Helpful in supporting GBA in dementia 14 (36.8%) 38 clinical diagnosis GBA + infarct in vascular dementia 8 (30.7%) 26 Frontal atrophy in schizophrenia 2 (14.2%) New unanticipated NPH in dementia 3 (7.8%) 38 finding requiring Clinical Alzheimer s disease but 5 (41.6%) 12 intervention infarct on CT scan requiring control of vascular risk factors 3. Evaluating the In patients with mood disorders 3 (10%) 30 progress evaluating the progress of previous haematoma 4. Impact on initiating Commencement of cholinesterase 5 (41.6%) 12 new management inhibitors for Alzheimer s disease Cerebral infarct in vascular 8 (30.7%) 26 dementia: commencement of vascular preventative therapies CT scan showing ischaemia in 3 (10%) 30 mood disorders: commencement of therapies to control BP 5. Normal scans helped to rule out any unsuspected brain lesion in various psychiatric conditions Key: GBA, generalised brain atrophy; NPH, normal pressure hydrocephalus Table 4. Frequency and percentage of CT brain results that influenced clinical care Rather than scanning all patients with psychiatric conditions, investigators have proposed using neurological abnormalities, histor y of seizure, history of head injury or age above 40 years as reasons to order a CT scan. 16 A study 17 has identified and emphasised two predictors as indications for brain scanning: dementia of less than one year duration with headache, focal signs or papilloedema; and mild dementia of acute or recent onset of less than one year. Another study found that none of the clinical predictors had a significant relationship to the actual influence of CT scans on diagnosis Progress in Neurology and Psychiatry and management of dementia patients attending a memor y clinic. 18 Other studies have also shown that neither the neurological nor cognitive examinations were sensitive enough to predict overall normal and abnormal CT scans. 6,8 These observations highlight the need for the development of clinical predictors of neuroimaging that could influence subsequent patient management. In our study, the impact of brain scans on diagnosis was mostly a negative one, ie scans helped to rule out any unsuspected diagnosis in various psychiatric conditions. In more than half the cases of vascular dementia and Alzheimer s disease, the scans supported the clinical impressions. Three patients with dementia had unsuspected normal pressure hydrocephalus, which required referral to neurosurgeons. Some patients with clinical Alzheimer s disease revealed cerebral infarcts, which required medical inter vention. Three patients with mood disorders had a previous history of haematoma and repeat reassuring CT brain scan results were helpful in evaluating the progress of the existing condition. Five out of 12 Alzheimer s patients were star ted on cholinesterase inhibitor drugs after CT brain findings were available, and nearly one in four vascular dementia patients required medical intervention to control vascular risk factors. Adequate control of blood pressure was prompted following CT brain result of brain ischaemia in three patients with mood disorders. Therefore, we observed in our study that brain scan results in 43 out of 102 patients had some impact on clinical care. Summary In summary (see Table 4), our study showed that abnormal results influenced diagnosis in 11/102 (10.7 per cent), prognosis in 3/102 (2.9 per cent) and treatment in 24/102 (23.5 per cent) patients. The greatest yield in clinical care was in patients with dementia. We found potentially reversible intracranial pathology in 3/102 (2.9 per cent) patients referred for brain scans. Different studies have reported variable results in determining the impact of brain scans on diagnosis and management of clinical conditions. One study reported that diagnosis was changed in 17 per cent of the psychiatric patient population with a mean age of 60 years. 9 Another study showed 1.2 per cent

5 patients with new or unanticipated diagnosis on CT scans. 5 A study with a pathologically confirmed cohort revealed that structural neuro imaging can help to identify vascular dementia or vascular components of Alzheimer s disease by increasing the sensitivity of clinical evaluation from 6 to 59 per cent and management of the vascular component may slow cognitive decline. 14 Another study reported that following a CT scan, diagnosis, management and prognosis of psychiatric conditions was influenced in 11.7 per cent of patients over a 37-month period. 6 In a series of 128 patients with chronic schizophrenia, CT scans identified two patients with subdural haematoma and one with meningioma and these findings were previously unexpected clinically. 2 Potentially reversible lesions on CT brain have been identified in 1.6 per cent to 4.9 per cent of psychiatric cases. 3,8 Various indications justifying CT brain scans in psychiatric patients have been proposed. 19,20 One study 5 suggested the following as definite indications for CT brain scanning in psychiatric patients: Positive history of previous head injury, stroke or other neurological disease Abnormal neurological sign or organic mental sign such as confusion or cognitive decline First psychotic break or personality change after age 50 years. Scanning is less rewarding for investigation of mental disorder associated with alcohol or substance abuse, mental disorders without other neuropsychiatric abnormalities or for elucidation of abnormal EEGs. Further studies are needed to identify characteristics of psychiatric patients most likely to benefit from CT or those who can be excluded from neuroimaging safely. Strengths and limitations of the study We collected data on patients referred by psychiatrists for CT brain scans for evaluation of psychiatric illnesses. The psychiatric diagnosis and selection of patients was at the discretion of the psychiatrist referring the patient and different psychiatric teams were involved from a neighbouring psychiatric hospital. At the time of CT brain scanning the duration of psychiatric illness among patients showed a wide range from one month to 24 years. Our study population has included a wide age range and therefore heterogenous CT scan findings could be expected. However, the common factor in our study population was the presence of a psychiatric clinical diagnosis confirmed by a psychiatrist. The study was conducted prospectively, thereby minimising the chances of missing data. A single radiologist repor ted all the CT scans thereby minimising the chances of inter-obser ver variation. Although we did not have an equal number of patients in each psychiatric diagnostic group, the pattern of representation reflects the existing clinical practice of CT scan referrals. This referral pattern is likely to be similar in other district hospitals in the UK. Therefore, results of this study could be generalisable across other UK hospitals. Although technical developments over the last few years have led to the availability of improved brain imaging techniques such as MRI, PET, SPECT, etc we only included CT scans in this study. The advantages of using CT scans over MRI imaging techniques are that: CT scanning is quicker There are fewer problems with movement artifact It is readily available 30 Progress in Neurology and Psychiatry

6 Key points Some abnormalities were reported in CT scans of up to 64 per cent patients A large percentage of scan results affected clinical diagnosis, prognosis and management Most scans were helpful in ruling out unsuspected lesions, and Patients with dementia had the greatest number of abnormal scan results (attributed to older age) It is not contraindicated with implants and pacemakers. The disadvantages of CT compared with MRI include: Exposure to radiation Lower soft-tissue contrast The risk of anaphylaxis from iodinated contrast agents Renal failure. By design, this was a short observational study and the abnormal positive findings cannot be compared with a normal control group. However, with imaging techniques that involve the use of ionising radiation such as CT scanning, the greatest limitations on interpreting the results are finding adequate control groups and performing repeated scans to gauge the relationship between abnormalities observed and the natural history of disease or effects of treatment. MRI has the advantage in this regard, as it does not involve ionising radiation and control groups from an unbiased population can be easily studied. 21 Whether the increased sensitivity of MRI leads to better outcomes for psychiatric patients remains to be determined. 22 Conclusion Our study confirms that CT brain scanning in common psychiatric conditions continues to provide valuable information on diagnosis and prognosis, and is also helpful in management. The main value appears to be in ruling out unsuspected diagnostic possibilities, but the scans have additional value in clinical management. The majority of our study population included dementia patients. Considering the social and financial burden associated with dementia, CT scans could continue to provide useful information for clinical management, and the scans would justify their usage. With the increasing demand for cholinesterase inhibitor treatments for dementia patients, the availability of CT scanners across most UK hospitals could also be justified. The high prevalence of CT abnormalities in psychiatric conditions other than dementia found in our study needs further evaluation to determine the relative efficacy of utilisation of this diagnostic resource. Despite the high level of positive results, we believe that routine CT scanning of all psychiatric patients is unnecessar y. As brain scanning is possibly one of the more expensive diagnostic tests, ordered by psychiatrists, it is desirable to make the most cost-effective use of this technique. Larger studies in the future would be helpful in identifying suitable clinical and other criteria to refine the indications for CT brain scanning in psychiatr y. There is a need to develop clinical predictors as indicators for neuroimaging that could influence patient management. Dr Gupta is a Consultant Physician in Carmarthen, Dr Elheis is a Consultant Radiologist in the Faculty of Medicine, Jordan University of Science and Technology, Dr Pansari is a Consultant Psychiatrist in Haverfordwest and Dr Syed is a Senior House Officer in Surgery in Swansea References 1. Oxman TE. Use of CAT in neuroradiological diagnosis in psychiatry. Compr Psychiatry 1979;20: Owens DGC, Johnstone EC, Bydder GM, et al. Unsuspected organic disease in chronic schizophrenia demonstrated by CT. J Neurol Neurosurg Psychiatry1980;43: Larsen EB, Mack LA, Watts B, et al. CT in patients with psychiatric illnesses: advantage of a rule in approach. Ann Intern Med 1981;95: Evans NJR. Cranial CT in clinical psychiatry. 100 consecutive cases. Compr Psychiatry 1982;23: Hollister LE, Boutros N. Clinical use of CT and MR scans in psychiatric patients. J Psychiatr Neurosci 1991;16(4): Colohan H, Callaghan EO, Larkin C, et al. An evaluation of cranial CT scanning in clinical psychiatry. Ir J Med Sci 1989;158(7): Hollister LE, Shah NN. Structural brain scanning in psychiatric patients: a further look. J Clin Psychiatry 1996;57(6): Roberts JKA, Lishman WA. The use of CAT head scanner in clinical psychiatry. Br J Psychiatry 1984;145: Beresford TP, Blow FC, Hall REW, et al. CT scanning in psychiatric inpatients: clinical yield. Psychosomatics 1986;27: Moles JK, Franchina JJ, Sforza PP. Increasing the clinical yield of computerised tomography for psychiatric patients. Gen Hosp Psychiatry 1998;20: Woolf SH, Kamerow DB. Commentary. Testing for uncommon conditions. The heroic search for positive test results. Arch Intern Med 1990;150: Madsen AL, Karle A, Rubin P, et al. Progressive atrophy of frontal lobes in first episode schizophrenia. Acta Psychiatr Scand 1999;100: Chua SE, McKenna PJ. Schizophrenia a brain disease? Br J Psychiatry 1995;166: Chui H, Qian Z. Evaluation of dementia: a systematic study of usefulness of The American Academy of Neurology s practice parameters. Neurology 1997;49: Palsion S, Larson L, Tengelin E, et al.the prevalence of depression in relation to cerebral atrophy and cognitive performance in 70 and 74 year old women in Gothenburg. The Womens Health Study. Psychol Med 2001;31(1): Tsai I, Tsuang MT. How can we avoid unnecessary CT scanning for psychiatry patients? J Clin Psychiatry 1981;42: Martin DC, Miller J, Kapoor W, et al. Clinical prediction rules for computed tomographic scanning in senile dementia. Arch Intern Med 1987;147: Condefer KA, Haworth J, Wilcock G. Prediction rules for CT in the dementia assessment: do they predict clinical utility of CT? Int J Geriatr Psychiatry 2003;18(4): Rosenburg CE, Anderson DC, Mahowald MW, et al. CT and EEG in patients without focal neurological findings. Arch Neurol 1982;39: Weinberger DR. Brain disease and psychiatric illness: when should a psychiatrist order a CT scan? Am J Psychiatry 1984;141: Cohen BM, Buonanno F, Keck Jr, et al. Comparison of MRI and CT scan in a group of psychiatric patients. Am J Psychiatry 1988;145: Kent DL, Haynor DR, Longstreth WT. The clinical efficacy of MRI in neuroimaging. Ann Intern Med 1994;120: Progress in Neurology and Psychiatry 31