COGNISTAT AND THE MCI INDEX
Cognistat is a widely used cognitive screening examination with well-established sensitivity to brain dysfunction. I was the lead developer of this test in 1979, and it has demonstrated its clinical utility in numerous research studies since that time. It assesses functioning in ten distinct cognitive areas: Attention, Registration, Comprehension, Repetition, Naming, Calculations, Memory, Constructions, Reasoning and Judgment. Cognistat provides a profile of scores that reveals a diagnostically useful pattern of cognitive strengths and weaknesses. It did not, however, offer a single score indicator of generalized cognitive decline that could be used to monitor progressive changes in mild cognitive impairment or dementia until the recent development of the MCI Index.
The MCI Index: Previously published formulations regarding MCI either place too much emphasis on cognitive testing or too little. This derives from a misunderstanding of the role of cognitive assessment in diagnostic work. Test scores on cognitive tests are not definitive because test scores can be impacted by a variety of factors that have nothing to do with progressive cognitive decline. Personal factors like intelligence level, learning problems, language and cultural differences, age, and amount of education can greatly affect cognitive test performances. Transient state factors like anxiety, depression, fatigue, pain and medication effects can undermine test performances at any given point in time. Although there is no simple way to determine the contribution of any particular factor to cognitive test performances, clinicians must still investigate these factors and incorporate them into their diagnostic formulation.
Because cognitive testing alone cannot provide a medical diagnosis, some clinicians believe that they cannot depend upon test scores diagnostically, and they therefore dismiss or devalue cognitive testing. Even formal, comprehensive neuropsychological testing is often devalued in this way. Nevertheless, cognitive testing is typically the most sensitive early indicator of progressive cognitive decline. A large percentage of patients who develop progressive cognitive decline only show changes on cognitive testing in the early stages of their illness. Some patients continue to show increasing cognitive impairment on testing for years prior to having difficulty with undemanding and over learned ADL’s or showing evidence of positive biomarkers. The clinician who devalues changes on cognitive testing loses the most sensitive indicator of both MCI and dementia.
The MCI Index was designed to provide clinicians with guidance regarding the risk or probability that cognitive test results from Cognistat indicate either MCI or dementia. This allows cognitive testing to play a critical role in a clinician’s diagnostic formulation without offering a specific diagnosis based on test scores alone. The MCI Index links increasing impairment on specific Cognistat subtest to the increasing likelihood of MCI or dementia.
The MCI Index: The MCI Index consists of a seven point scale that reflects both the degree of cognitive impairment and the increasing likelihood of MCI or a dementia syndrome. The first four MCI Index values from 0 to 3 address the probability of the presence of MCI. They range from giving no indication of MCI to strongly suggesting an MCI diagnosis. As the degree of cognitive impairment increases, the index values address the probability of a possible dementia diagnosis. Index values from 4 to 6 range from raising the question of dementia to strongly suggesting a dementia syndrome. The seven Index levels are as follows:
MCI Index = 0: no significant cognitive impairment
MCI Index = 1: raises the question of MCI
MCI Index = 2: suggests MCI
MCI Index = 3: strongly suggests MCI
MCI Index = 4: raises the question of a dementia syndrome
MCI Index = 5: suggests a dementia syndrome
MCI Index = 6: strongly suggests a dementia syndrome
We designed the MCI Index in response to the issues raised by the eight questions discussed above. We did not attempt to define a cognition construct apart from the ten Cognistat subtests used to generate the profile of scores. After considering several subtests known to be affected by cognitive decline, we determined that only the Construction and Memory subtests were sufficiently sensitive to and reliably involved in MCI and dementia. We have used these two subtests in conjunction with each other to determine the seven levels of the MCI Index.
The MCI Index levels were determined based upon clinical rather than statistical criteria. The “mild” impairment level was not set at 1.0 or 1.5 standard deviations below the mean. We used Cognistat score combinations that are typically associated with mild cognitive impairment according to clinical experience in setting the MCI Index levels. Instead of attempting to establish boundary criteria for a diagnosis of normal, mild impairment or a dementia, we chose to establish guidelines for setting an appropriate level of clinical concern. At a fairly minimal level of impairment, we believe that it is necessary to raise the question of a possible MCI diagnosis. As impairment increases, the likelihood of an MCI diagnosis increases accordingly. A greater probability is indicated by the designation, “suggesting MCI,” and an even greater probability by the designation, “strongly suggesting MCI.” As impairment increases further, it “raises the question” of a dementia. Greater impairment “suggests” a dementia, and yet more impairment “strongly suggests” a dementia.
None of the levels in the MCI Index can be used to actually make a diagnosis. Both the MCI and dementia diagnoses depend upon clinical judgment that incorporates the historical and the clinical context of the individual patient in addition to the results of cognitive testing. The clinical context must include the presence of any previously diagnosed medical or psychiatric conditions as well as the possible impact of prescribed medications. The ability to carry out the basic activities of daily living (ADL’s) is another factor that helps establish the clinical context. ADL’s, however, tend to be insensitive to MCI and to the early stages of dementia as well. Thus, they are seldom helpful in guiding clinicians in making an initial diagnosis of MCI or dementia. The MCI Index scores, however, provide a range of guidance for clinicians regarding cognitive testing. They suggest the degree of concern warranted in considering either the MCI or dementia diagnoses.
We framed the guidance offered by the MCI Index in terms of probabilities in order to relieve the clinician from the need to define a diagnosis based on cognitive testing alone. We want clinicians to recognize the importance of cognitive test results to their diagnostic considerations, and we provide guidance regarding the level of concern appropriate to the scores obtained in an individual case. The level of concern warranted can be stated in probabilistic form with each increasing point in the MCI Index indicating a greater degree of concern. Since MCI Index scores also reflect increasing cognitive impairment, the diagnostic concerns regarding MCI expressed at MCI Index scores of 1, 2 and 3 transition into concerns regarding a possible dementia diagnosis at MCI Index scores of 4, 5 and 6.