Emergency Psychiatry: Rapid Screening for Cognitive Impairment in the Psychiatric Emergency Service: I. Cognitive Screening Batteries

Perhaps the most commonly used cognitive screen in psychiatric emergency services is the Mini Mental State Examination (MMSE) (3). The MMSE is useful because it is easy to administer and assesses a wide range of cognitive functions in verbal modalities and some visual modalities.

For example, a recent factor analytic study found that the MMSE tapped frontal, memory, and spatial domains of cognitive functioning among psychiatric patients (4). The MMSE also identifies individuals who have a high probability of moderate to severe global cognitive impairment, especially elderly patients (5).

However, the MMSE also has important limitations. Although the instrument is sensitive in the detection of severe dementing conditions, it has been found to be less sensitive in the detection of milder forms of dementia and cognitive dysfunction among elderly patients (5). In addition, the MMSE greatly underestimates cognitive impairment among psychiatric patients (6). Also, detection specificity of the MMSE (and of other screening tools) has been found to be related to patients' ethnicity and educational level.

For example, in one study MMSE detection of dementia was found to be significantly more accurate among white patients than among black patients (7). Another group of investigators found that the use of a functional screening instrument that taps performance on activities of daily living—the Direct Assessment of Functional Status—was as sensitive in the detection of dementia and was less biased by patients' level of formal education than cognitive screening instruments such as the MMSE (8).

However, supplementing the MMSE with other brief assessment strategies may enhance the detection of dementia and other neuropsychiatric conditions. Mackinnon and Mulligan (9) combined the MMSE with an informant interview about the cognitive status of 106 patients with suspected memory impairment or dementia. Logistic regression analysis showed that combining informant interview information with patients' MMSE results enhanced the accuracy of discriminating patients who had dementia from those who did not have dementia.

The High Sensitivity Cognitive Screen (HSCS) (10) is a 20-minute interview-based test designed to quickly identify patients who show neuropsychologic impairment. Like the MMSE, the HSCS examines a wide range of functions across various cognitive domains, including memory, language, attention and concentration, visual and motor skills, spatial perception, and self-regulation and executive functioning. The HSCS has been found to be sensitive in the detection of subtle cognitive impairment among geriatric and psychiatric patients with HIV infection (10). This test has demonstrated adequate reliability and validity compared with standardized neuropsychologic tests (10).

The Neurobehavioral Cognitive Status Exam (NCSE) (11) is a screening instrument that was originally developed to briefly assess dementia and disorders of the central nervous system. The NCSE has multiple subtests that tap a variety of cognitive functioning domains—for example, attention and memory—to determine whether formal neuropsychologic testing is warranted in a particular cognitive domain. The NCSE subtests were validated by comparisons with frequently used and well-standardized neuropsychologic tests that assess similar constructs. Although the NCSE subtests have been shown to be highly correlated with abilities measured by neuropsychologic tests, the instrument's utility as a screen for dementia is limited (11). Specifically, the NCSE was found to be poor at discriminating individuals with clinically significant cognitive impairment from those without such impairment and to have a low false-negative rate and a high false-positive rate for detecting domain-specific cognitive impairments among patients with organic brain dysfunction (11).

Other investigators have also had mixed results with the NCSE. Fields and colleagues (12) compared the utility of the NCSE with that of the MMSE as a cognitive screen for dementia in a geriatric inpatient sample. The NCSE was found to be more sensitive in detecting dementia than the MMSE but had poorer specificity and predictive properties. Studying a nongeriatric adult population, Blostein and associates (13) found that the NCSE was a useful cognitive screening tool for identifying deficits associated with mild traumatic brain injury. Memory deficits were most strongly associated with positive cognitive screens.

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (14,15) is a screening instrument designed for detecting a variety of cognitive impairments associated with schizophrenic illness. The RBANS provides both a total cognitive impairment score and explicit cognitive functioning index scores—measures of language, visual functions, memory, and attention. The screen was standardized on a U.S. Census-matched adult population and is sensitive in the detection of deficits commonly associated with acute schizophrenic illness (14). The instrument is relatively brief to administer (25 to 30 minutes) and has demonstrated excellent test-retest reliability and convergent validity (14,15). For example, it has been found to correlate highly with established neuropsychologic tests that examine deficits commonly associated with schizophrenia, including memory and attention dysfunction (15). The screen was not found to correlate with symptoms but was found to correlate with employment outcome (15). Overall, the RBANS has been shown to have great utility as a cognitive screen for detecting cognitive deficits among patients with schizophrenia and appears to be highly appropriate for use in the psychiatric emergency service.

Unlike performance-based screening instruments, a variety of psychiatric rating scales contain items from statistically derived factors that assess an individual's cognitive status and also may be appropriate for evaluating patients' cognitive status in the psychiatric emergency service (16). For example, on the Positive and Negative Symptom Scale (PANSS) (17), cognitive symptom items to be rated include abstract reasoning, attention, stereotyped thinking, and disorientation. Harvey and colleagues (18) found mixed results regarding the validity of PANSS-based cognitive symptom ratings. Comparison of patients' ratings on five different PANSS cognitive symptom factor solutions to their performance on a wide variety of neuropsychologic assessment tasks indicated that cognitive performance was poorly correlated with clinicians' cognitive symptom rating of the same psychological constructs—for example, memory and distractibility.

Unlike the wide-ranging cognitive factors derived from the PANSS, the Schedule for the Assessment of Negative Symptoms (SANS) (19) limits ratings of cognitive symptoms to attentional dysfunction. SANS cognitive symptom items include social inattentiveness during mental status testing, subjective complaints of inattentiveness, and a global rating of inattentiveness. Factor-analytic studies have generally supported the coherence of the SANS attention subscale (20). In addition, supporting the validity of symptom-based ratings of inattention, a recent study found that the SANS attention subscale correlated well with neuropsychologic performance measures of attentional functioning among persons with schizophrenia (16). Despite significant methodologic problems in using cognitive symptom-based ratings, these scales may be of limited use as initial screens for detecting impairment.

Overall, the use of cognitive screening instruments may facilitate differential diagnoses and reduce the reliance on clinical judgment in the psychiatric emergency service. The cognitive screening instruments we have highlighted in this column assess a range of functioning across many cognitive domains. In a second column we will review individual cognitive tests that can be used in the psychiatric emergency service to detect deficits in a suspected area of dysfunction.

Dr. Serper is affiliated with the department of psychology at Hofstra University in Hempstead, New York, and with the department of psychiatry at New York University School of Medicine in New York City. Dr. Allen is with the department of psychiatry at the University of Colorado School of Medicine in Denver. Send correspondence to Dr. Serper at the Department of Psychology, 222 Hauser Hall, Hofstra University, Hempstead, New York 11549-1350 (e-mail, [email protected] or [email protected]). Douglas H. Hughes, M.D., is editor of this column.