Current issues of ACP Journal Club are published in Annals of Internal Medicine


Mini-Mental State Examination and the Informant Questionnaire on Cognitive Decline were efficient screening tests for dementia

ACP J Club. 1997 Jan-Feb;126:20. doi:10.7326/ACPJC-1997-126-1-020

Source Citation

Mulligan R, Mackinnon A, Jorm AF, Giannakopoulos P, Michel JP. A comparison of alternative methods of screening for dementia in clinical settings. Arch Neurol. 1996 Jun; 53: 532-6.



To compare 3 screening approaches for dementia in a clinical environment: cognitive testing, informant report, and neurovisual assessment.


A blinded comparison of the results from each of the 3 screening tests with the diagnosis of dementia based on the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R).


A university geriatric hospital and memory clinic in Geneva, Switzerland.


76 patients (mean age 82 y, 79% women) who were admitted to the hospital or were attending an outpatient clinic. Patients were excluded if they had no informant; were not fluent in French; or had severe sensorimotor deficit, delirium, or acute physical conditions.

Description of tests and diagnostic standard

The French-language versions of the Mini-Mental State Examination (MMSE), the Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE), and the Clinical Antisaccadic Eye Movement Test (AEMT) were used according to instructions provided by the test developers. The MMSE was administered by a geriatric registrar, the IQCODE was given to informants by a nurse, and the AEMT was administered by a neuropsychologist. Each test was administered by personnel blinded to the results of the other tests and to the results of the diagnostic standard (diagnoses of dementia and depression were made by senior psychiatrists using a checklist of DSM-III-R criteria).

Main outcome measures

Sensitivity, specificity, and likelihood ratios.

Main results

33 patients were diagnosed with dementia and 11 with depression; 2 of the latter group also had dementia. The sensitivity and specificity of the MMSE at the 26/27 cut-off point were 91% and 63%, respectively. The corresponding figures for the IQCODE at the cut-off point of 3.27/3.30 were 100% and 42%, respectively. {The likelihood ratio for a positive test (+LR) with the MMSE at the 26/27 cut-off point was 2.46 (10.86 at the 23/24 cut-off) and the likelihood ratio for a negative test (-LR) was 0.14 (0.26 at the 23/24 cut-off). The +LR with the IQCODE at the 3.27/3.30 cut-off point was 1.72 (2.53 at the < 3.6/≥ 3.6 cut-off) and the -LR was 0 (0.34 at the < 3.6/≥ 3.6 cut-off.}* The AEMT did not perform as well as the other 2 tests.


The Mini-Mental State Examination and the Informant Questionnaire on Cognitive Decline in the Elderly were efficient methods of screening for dementia in clinical settings and comparably discriminated between patients with and without dementia.

Source of funding: Not stated.

For article reprint: Dr. R. Mulligan, Consultation Mémoire, Hôpitaux Universitaires de Genève, 6 rue du XXXI décembre, 1207 Geneva, Switzerland. FAX 41-22-718-45-99.

*Numbers calculated from data in article.


The prevalence of dementia is 3% by 70 years of age and doubles every 5.1 years thereafter. Because clinicians often miss mild-to-moderate dementia, routine screening with valid and reliable instruments is recommended for patients at risk. This can also be the first line of assessment for cognitive decline. The study by Mulligan and colleagues compared 3 validated, brief screening tools. They show that the less-known, informant-based IQCODE (1) is as effective as the familiar MMSE and support incorporation of the former into the clinician's armamentarium. The IQCODE eliminates the biases of education, pre- morbid ability, and sensory and language barriers but is affected by the extent of the informant's contact. The AEMT needs further study and is not recommended.

The accuracy of the tests may be skewed by the presence and severity of dementia in the study population. The present study, which used patients from a memory clinic and geriatric hospital, showed performance characteristics for the MMSE and IQCODE that differed from those found by Jorm and colleagues (2) in a sample derived from inpatients, a geriatric day hospital, and outpatient clinics. The greatest challenge in diagnosing cognitive impairment is interpreting scores near the cut-off point. In this situation, the IQCODE may complement the MMSE by assessing temporal changes in daily cognitive functioning, thus reducing the need for costly neuropsychologic testing. Clinicians should use judgment in interpreting test results. For example, a "normal" score of 27 on the MMSE carries a different clinical meaning when it results from the failure of short-term recall than when a patient who is new to the clinic does not know the date, county, or building where he or she is. When speed is essential, MMSE items for recall and orientation appear to have comparable sensitivity and specificity to the full MMSE (3), but positive findings should be followed by a more complete examination.

Calvin H. Hirsch, MD
University of California at DavisSacramento, California, USA


1. Jorm AF, Jacomb PA. Psychol Med. 1989;19:1015-22.

2. Jorm AF, Scott R, Cullen JS, MacKinnon AJ. Psychol Med. 1991;21:785-90.

3. Galasko D, Klauber MR, Hofstetter CR, et al. Arch Neurol. 1990;47:49-52.