Evidence ‘Insufficient’ for Cognitive-Impairment Screening

Abstract

Citing an ongoing lack of data about the benefits and harms of screening, the U.S. Preventive Services Task Force has left unchanged the recommendations of its 2003 guidelines on cognitive-impairment screening in older adults, according to an update published recently. “Evidence on the effect of screening and early detection of mild to moderate dementia on decision making, planning, or other important patient outcomes is a critical gap in the evidence,” Virginia A. Moyer, MD, said in a report on behalf of the USPSTF. Other research needs include further study of the harms of screening, new interventions that address the changing needs of patients and families, and interventions that affect the long-term clinical direction of mild-to-moderate dementia. In its review, the USPSTF evaluated 55 studies on instruments that screen for cognitive impairment, of which 46 provided evidence on the sensitivity of dementia screening and 27 provided evidence on mild cognitive impairment. Screening tests included a variety of tasks to assess at least one cognitive function, such as memory, attention, language, and visuospatial/executive functioning. The USPSTF looked at studies that used the Mini-Mental State Examination (MMSE), Clock Drawing Test, verbal fluency tests, Informant Questionnaire on Cognitive Decline in the Elderly, Memory Impairment Screen, Mini-Cog Test, Abbreviated Mental Test, and Short Portable Mental Status Questionnaire. The MMSE was the most-evaluated screening tool, with 25 published studies in people 69-95 years old. Mean prevalence of dementia ranged from 1.2% to 38%. The pooled sensitivity from 14 studies for the most commonly reported cut points was 88.3% (95% confidence interval, 81.3%-92.9%), and specificity was 86.2% (CI, 81.8%-89.7%). Other screening tools “were studied in far fewer studies (four to seven studies each), had limited reproducibility in primary care relevant populations, and had unknown optimum cut points,” Dr. Moyer wrote. In addition, no trial compared clinical and decision-making outcomes in screened and unscreened patients, the report said. And no study probed the consequences of false-positive or false-negative screening results, such as psychological harms, unnecessary diagnostic testing, or stigmatization. Types of dementia in older adults include Alzheimer's disease, vascular dementia, frontotemporal dementia, dementia with Lewy bodies, Parkinson's disease with dementia, and mixed-cause dementia. The USPSTF distinguishes between dementia and mild cognitive impairment, which is defined as less severe and not considerably interfering with day-to-day activities. The prevalence of dementia is estimated to be 5% in adults aged 71-79 years of age, 24% in those aged 80-89 years, and 37% in those aged 90 years and older. The prevalence of mild cognitive impairment is more uncertain, and estimates range from 3% to 42% in adults aged 65 years and older. This report differs from the 2003 recommendations because it considers screening and treatment for mild cognitive impairment, as well as dementia, and it includes additional information about the performance of screening instruments. “The overall evidence is insufficient to make a recommendation on screening,” Dr. Moyer said. Editor's NoteWhether the harms of screening for cognitive impairment outweigh the benefits in the general population, everybody gets tested in our nursing homes. The Brief Inventory of Mental Status is performed on every resident no less frequently than quarterly. So we have a screening test and a surveillance test right there in the chart. The BIMS certainly can be helpful to us as clinicians in several ways, so don't forget that it's in there, along with the PHQ-9 depression inventory.—Karl Steinberg, MD, CMD Editor in Chief Whether the harms of screening for cognitive impairment outweigh the benefits in the general population, everybody gets tested in our nursing homes. The Brief Inventory of Mental Status is performed on every resident no less frequently than quarterly. So we have a screening test and a surveillance test right there in the chart. The BIMS certainly can be helpful to us as clinicians in several ways, so don't forget that it's in there, along with the PHQ-9 depression inventory. —Karl Steinberg, MD, CMD Editor in Chief