CT Scans of First-Break Psychotic Patients in Good General Health

Some clinicians have argued that all patients presenting with a first onset of psychotic illness should receive a computed tomography (CT) scan of the head to rule out such causes as tumors, abscesses, Huntington's disease, encephalitis, Wilson's disease, and trauma (1-3).

In a study by Gewirtz and colleagues (1) of patients with new psychotic illness, the incidence of CT findings other than atrophy was 6.6 percent. However, only two of 168 patients (1.19 percent) had abnormalities that had actual implications for patient management. One of the patients had a colloid cyst that was judged to be symptomatic and was removed. The other had an arteriovenous malformation that was small and not easily accessible and was managed conservatively.

Rather than scanning all first-break psychotic patients, some investigators propose using neurologic abnormalities (4—6), a history of seizure (4,5), a history of head injury (5), or age above 40 years (5) as reasons to order a CT scan. A prospective study examining CT scans and organic versus functional psychosis found that CT scan abnormalities did not correlate with diagnoses of organic brain syndrome, casting further doubt on the value of routine CT scanning in this group (7).

To help advance this debate, this study examined a specific group of new-onset psychotic patients—those who are young and in good general health. A working hypothesis was that psychotic illness alone is not sufficient to warrant a CT scan. This study is valuable because it looked at a relatively homogeneous group without such confounding variables as advanced age and comorbid conditions.

In this era of cost control in health care, each dollar spent must be justified. CT scans cost money; failing to detect a brain lesion by not scanning has potentially greater costs. If investigators can define subpopulations of newly psychotic patients in which the benefits of CT scanning outweigh the costs and in subpopulations in which CT scanning is unwarranted, health care dollars could be more carefully spent.

Methods

The Naval Medical Center in San Diego is a tertiary care center with 393 inpatient beds (8), which provides care to persons on active military duty, retirees, and their dependents. On average, the daily inpatient psychiatry census is 48 (8). Currently, most of these patients receive a head CT scan as a standard part of their diagnostic work-up if they are admitted with a new psychotic illness.

The psychiatry ward's log book of admissions and discharges was used to identify potential subjects. Charts were reviewed for all patients admitted to or discharged from the inpatient psychiatry service from January 1, 1992, through September 30, 1994, with a DSM-III-R diagnosis on admission of psychotic disorder not otherwise specified, schizophreniform disorder, schizophrenia, brief reactive psychosis, schizoaffective disorder, delusional disorder, bipolar disorder, or major depression. Patients were included in the study if they were evaluated for a psychotic illness, had no previous evaluations for a psychosis, and received a CT scan of the head.

Each case was categorized by the independent variables of age (18 to 30 years, 31 to 40 years, and 41 years and older), sex, diagnosis (schizophreniform disorder and schizophrenia, bipolar disorder, major depression, schizoaffective disorder, and other diagnoses), neurological examination (normal or abnormal), and a history of seizure (present or absent). The neurological examinations were performed by a staff or resident psychiatrist within 24 hours of admission. This physician also obtained all medical history, including history of seizures. Each admission diagnosis was made by either a psychiatric resident or a board-certified psychiatrist. Discharge diagnoses were made by board-certified psychiatrists using DSM-III-R criteria.

All CT scans were read by a staff neuroradiologist; often a radiology resident also read the films. Results were classified as normal, incidental, or abnormal. An incidental CT finding was one felt by the radiologist to have no clinical or diagnostic bearing. The significance of the relationship between CT scan results and the independent variables was evaluated by contingency tables using the Fisher's exact test, which is similar to but more accurate than chi square analysis.

Results

The sample consisted of 127 patients, 102 males and 25 females. Ninety-eight patients were between the ages of 17 and 30 years, 23 were between the ages of 31 and 40 years, and six were 41 years old or older. Seventy-seven patients were single, 42 were married, and eight were separated or divorced. Of the 127 patients, 73 were non-Hispanic Caucasians, 17 were Hispanic, 25 were black, three were Asian, and nine were Filipino or Pacific Islander.

The most common admission diagnosis was psychotic disorder not otherwise specified, which was given to 60 patients (47 percent). Discharge diagnoses, in descending order, were schizophreniform disorder or schizophrenia for 41 patients (33 percent), bipolar disorder for 21 patients (17 percent), major depression for 15 patients (12 percent), psychotic disorder not otherwise specified for 13 patients (10 percent), schizoaffective disorder for eight patients (6 percent), delusional disorder for six patients (5 percent), and brief reactive psychosis for four patients (3 percent); a group of 19 patients (15 percent) were discharged with other diagnoses.

Only two of the 127 patients had any abnormality on neurological examination at admission. Only five patients had ever had a seizure. Of the 127 patients, 123 had completely normal head CT scans.

Four patients had incidental findings that had no impact on diagnosis or treatment. A 20-year-old man discharged with a diagnosis of amphetamine delusional disorder had a punctate calcification in the right frontal deep white matter. The radiologist deemed it a truly incidental finding, with no further studies or workup indicated. A 21-year-old man with a discharge diagnosis of psychotic disorder not otherwise specified had an incidental finding of an arachnoid cyst in an otherwise normal scan. A 21-year-old woman with a discharge diagnosis of schizophreniform disorder had a left posterior fossa arachnoid cyst that was not considered clinically significant. A 24-year-old man with a discharge diagnosis of bipolar disorder, manic, had a suspected pineal gland tumor, which could have been an enlarged vein. A subsequent examination using magnetic resonance imaging was normal.

Of the four patients with incidental CT findings, three were Hispanics and one was a Filipino. Thus a strong association was found between incidental findings and ethnic minority status (Fisher's exact test, p=.004). Neurologic abnormalities, found in two patients, were strongly associated with increasing age; one patient in the 31- to 40-year age group and one in the 41-and-older group had neurologic abnormalities (Fisher's exact test, p=.019).

When discharge diagnoses were grouped as schizophreniform disorder and schizophrenia, bipolar disorder and major depression, psychotic disorder not otherwise specified, and all others, most patients in the schizophreniform disorder-schizophrenia group were single (never married), whereas the other diagnostic groups were nearly equally divided between patients who were single and those who were or had been married (Fisher's exact test, p=.024). No association was found between the incidental CT scan findings and neurologic abnormalities, a seizure history, or age.

Discussion

This sample of 127 first-break psychotic patients was predominantly young and otherwise healthy. The results of patients' CT scans had no impact on their diagnosis or treatment. Older patients and those who had neurologic abnormalities and a seizure history could reasonably be expected to have more abnormal CT scans. However, this study had too few of such patients to examine these subgroups.

The study was limited by its retrospective design, single location, and possibly insensitive neurological screenings. Regarding the strong association between incidental CT scan findings and ethnic minority status, one could propose some underlying cultural, socioeconomic, or genetic factors. Of note, ethnicity and incidental CT scan findings were unrelated to diagnosis.

Some clinicians have recommended obtaining a CT scan in each instance of new-onset psychosis because of potential implications for diagnosis and management (1-3). These new data reinforce the rare nature of significant CT findings among patients with new-onset psychosis. They support the idea that routine use of a head CT scan in young and otherwise healthy military patients being evaluated for a new-onset psychosis may be unnecessary. Further work and prospective studies are necessary to find more evidence in support of this hypothesis and to better define populations in which CT scanning is indicated.

Acknowledgment

This study, number S-95-058, was sponsored by the clinical investigation program of the bureau of medicine and surgery of the Department of the Navy in Washington, D.C.

Dr. Bain is affiliated with the departments of psychiatry and clinical investigation at the Naval Medical Center in San Diego. Send correspondence to Dr. Bain at the Clinical Investigation Department, Naval Medical Center, 34800 Bob Wilson Drive, San Diego, California 92134.