Screening for cognitive impairment among individuals aged 60 years or over: scoping review

BACKGROUND: Growth in aging of the population has led to increasing numbers of elderly people presenting cognitive impairment and evolution to dementia. There is still no consensus within primary care on the best strategy for screening for cognitive impairment among elderly people. Standardization of a simple but reasonably accurate instrument for a brief cognitive test, in primary care environments, would enable healthcare professionals to identify individuals who require a more in-depth assessment of cognition. OBJECTIVES: To investigate the instruments used by healthcare professionals in studies conducted worldwide and ascertain the most suitable instruments for screening for cognitive impairment among individuals aged 60 years or over, in the Brazilian population. DESIGN AND SETTING: Scoping review developed at Pontifícia Universidade Católica de São Paulo, Brazil. METHOD: A systematic search of the literature was conducted for primary studies using instruments to screen for cognitive impairment among individuals aged 60 years or over, in the MEDLINE, EMBASE, Cochrane Central and LILACS databases. RESULTS: A total of 983 articles were identified by two independent reviewers, from which 49 were selected for full-text reading, based on the criteria defined for this review. From this, 16 articles adhering to the theme of screening for cognitive impairment among the elderly were selected for in-depth analysis. CONCLUSION: The Mini-Mental State Examination was the instrument most cited in these studies. The Pfeffer Functional Activities Questionnaire and the Verbal Fluency Test (semantic category) present characteristics favoring further studies, for testing as screening instruments for cognitive impairment among elderly people in Brazil.

bring more information than when used separately. The functional scales of questionnaires are less influenced by the interviewee's age, education level or other sociocultural factors. 7 So far, there is no consensus regarding the best strategy within primary care for screening for cognitive impairment among elderly patients. However, several brief instruments for screening for cognitive impairment have been recommended. 8,9 No specific drug therapy for treating mild cognitive impairment (MCI) is currently approved. Nonetheless, it was recommended through the FINGER (Finnish Intervention Study to Prevent Cognitive Decline and Disability) study that healthy lifestyle factors such as leisure activities, social interaction, cognitive stimulation, Mediterranean diet and regular physical activity, both for elderly people in general and for those with MCI, should be encouraged as possible protectors against neurodegenerative diseases of aging. 10 Most individuals and their caregivers would rather know about a diagnosis of dementia as early as possible. This knowledge allows such individuals to make decisions regarding future plans while they still have the ability to do so. 11,12 In Brazil, around 75% of the population receives its medical care through the public healthcare system (Sistema Único de Saúde, SUS). In this, care is centered on general practitioners, who play an increasingly important role in screening for cognitive impairment among elderly people, which is often neglected within primary care. Moreover, many primary care providers have difficulty in diagnosing dementia accurately. Particularly at the mild stage, dementia is poorly recognized. 13,14 Thus, instruments for cognitive screening that are quick to apply but relatively accurate are needed, so that healthcare professionals working within primary care can identify individuals who may require a more in-depth evaluation of cognition, at an early stage, and refer them to secondary care. 15 The present study consisted of a scoping review, in which instruments for screening for cognitive impairment that have been used in studies in the literature, as applicable to individuals aged 60 years or over, were assessed.

OBJECTIVES
To investigate the cognitive screening instruments used by healthcare professionals in studies conducted worldwide and ascertain which of these are most suitable for use in screening for cognitive impairment among individuals aged 60 years or over, in the Brazilian population.

Design
This study consisted of a scoping review of the literature. It was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology. 16

Search strategy
The searches were conducted in June 2020 in following databases: The same search strategies were used in all databases. The search was refined by specifying randomized clinical trial (RCT) and the elderly age group, or studies that included individuals aged 60 years or over, depending on the filter for searching the information sources for articles, as described in Table 1. No limit was placed on the date of publication or the languages of these documents. For cases in which an update was found, the latest version was considered.

Criteria for inclusion in the scoping review
We only included studies that met the following criteria: randomized clinical trials (RCTs) that had been duly registered or observational studies with random sampling; individuals aged 60 years or over who had been recruited from the general population or from primary healthcare attendees, for random sampling, with absence of any reports of presence of pathological conditions or previous treatments; and application of instruments for screening for cognitive impairment and their implications and results. We considered any outcomes that had been assessed and reported by the original authors.

Selection of studies
The selection process was performed by two authors (MRLR, PRPF), who independently screened all titles and abstracts that had been found through the electronic search. These authors checked their eligibility in relation to the inclusion criteria.
Any disagreements in the selection process were resolved through reaching a consensus or by consulting a third author (JEM).
To assess the methodological quality of the studies included, the

Selection of articles
We found 983 articles in the first stage of article selection, but 244 articles were excluded due to duplication in the research databases.
Thus, 739 articles were retained for assessment of eligibility. In the next phase, articles that did not have the research topic in the title or abstract were excluded. Thus, a further 690 articles were excluded and 49 were selected for assessment of eligibility. Of these, only 16 articles met the objectives of this scoping review (Figure 1). Tables 2 and 3 provide details on the studies included, so that readers can make their own judgments about the research in these studies.

Results from blinded randomized clinical trials (RCTs)
The RCTs ( Table 2) were conducted on a total population of 10,445 people, with a weighted average age of 77.49 years. The educational level was only recorded in the study by Fowler et al. 18 In these studies, four instruments were used, which were all cognitive assessment tests: memory impairment screening (MIS), Mini-Cog, DemTect and clock drawing test 19 (CDT).
The RCT by Fowler et al. 18 did not detect any differences in healthcare, quality of life or harm from symptoms of depression and anxiety among individuals who were screened as positive for dementia, through application of MIS or Mini-Cog. 18 Reiner et al. 20 compared positive results from cognitive screening using DemTect with the results obtained through the CDT. 19 They suggested that the CDT 19 was not a suitable instrument for detection of probable dementia within primary care. 20 ("Disfunção cognitiva" OR "Comprometimento Cognitivo" OR "Comprometimento Cognitivo Leve" OR "Declínio Cognitivo" OR "Deficiências Cognitivas" OR "Deterioração Mental" OR "Distúrbio Neurocognitivo Leve" OR "Transtorno Neurocognitivo Leve") AND ("Programas de rastreamento" OR "Exame Coletivo" OR "Identificação Sistemática" OR rastreamento OR screening OR "Triagem de Massa") AND ( db:("LILACS") AND limit:("aged"))

Results from observational studies (OS)
The sample size in the 14 observational studies (OS) ranged from 50 to 15,051. It was in the range of up to 100 in one article; 101 to 1,000 in seven articles; 1,001 to 10,000 in five articles and more than 10,000 in one article. The total population of the OSs was 35,010 individuals ( Table 3). The participants' cognitive status was classified as follows: cognitively normal (CN); cognitive impairment with no dementia (CIND); mild cognitive impairment (MCI); and dementia in its respective clinical stages of evolution.
Among the observational studies, 19 instruments (14 cognitive assessment tests and five functional assessment scales) were used to screen the cognition of individuals aged 60 years or over (Tables 4 and 5).

DISCUSSION
The criteria used for analysis in this scoping review, on the instruments that might be best suited for use in the Brazilian population, were the following: quick application, validation for use in primary care locations or in the community; adequate psychometric properties; ease of application by members of the healthcare team; the least possible influence from the subject's educational and cultural level; and whether elderly people's interest in the evaluation was aroused. The sensitivity and specificity of screening instruments for cognitive impairment among the elderly were also considered.
Use of indiscriminate screening, i.e. for the entire elderly population, irrespective of any cognitive complaints, has been controversial. This is not only because of the need for adaptations to instruments, for them to be applied (given the lack of standardization), 21 but also because positive results could lead to harm such as anxiety and depression among individuals without any proven dementia. Nonetheless, in the RCT conducted by Fowler et al.,18 no harm due to symptoms of depression and anxiety was found after positive screening for dementia. 18 DemTect 22 , the instrument used by Reiner et al. 20  The Mini-Cog 26 includes the CDT, 19 with its characteristics as described above, along with immediate and late evocation of three repetitions of words. In the study by Fowler et al., 18 Mini-Cog was applied together with MIS. Those authors concluded that Mini-Cog was suitable for routine screening within primary care.
However, this test has not been recognized as a good tool for cognitive screening among elderly people in the Brazilian population with less than five years of formal education. 27 The memory impairment screening (MIS) 28 33 Although MoCA 31 has the disadvantage of taking longer to apply than MMSE 34 and presents limitations with regard to the capacity for illiterate individuals to perform the proposed tasks, it is a tool that provides a superior overall assessment in the early stages of cognitive decline. 38 Burkart et al. 39 compared the selective reminding procedure 40 (SRP) with MMSE 34 and concluded that the SRP was not recommendable for cognitive screening for dementia. 40 The Fuld Object Memory Evaluation (FOME) 41 assesses memory and learning through the SRP and can be applied to elderly people with a low level of formal education. It uses late evocation after distraction and is applied through a semantic VFT. 36 In Brazil, the only studies found involved a professional trained in psychology as the evaluator of this test. 42,43 The MMSE 34 was the instrument most cited and used in this scoping review, thus corroborating other findings reported in the literature. 21,44 It has been validated for application both in the community and in primary care in many countries, with the aim of increasing the recognition of cognitive impairment. It has been accepted both by patients and by interviewers, even without assessment of executive function. 29 Despite being widely used in Brazil, MMSE 34 needs adjustments to its cutoff scores, which are variable, because it can be influenced by age and level of formal education. 45 It is a screening tool that can be applied rapidly, and it addresses the main cognitive domains with high specificity and sensitivity for dementia. A wide variety of healthcare professionals have the capacity to use it. 46 The criteria used in the MMSE 34 make it highly capable of screening for moderate and severe cognitive impairment.
However, its ability to signal milder or earlier degrees of cognitive decline is significantly lower. It is not suitable for screening for the initial phases of dementia and can lead to higher rates of false negative results, since it does not evaluate executive function. 47        34 given that it is very effective in evaluating executive function and language ability, mainly due to its semantic approach, which seems to require a high level of thought process. The articles selected for the present review showed certain limitations. These included the rate of losses and the short follow-up period for the patients in the RCTs. 18,20 There was also selection bias in the subsample categories, when tests at different times of assessment were compared. Furthermore, there was no reassessment of participants with a negative result from screening for cognitive impairment. 53 Evaluation of a sample of patients from primary care and not from the community in general was criticized in some studies, 18,20,21 but this met the inclusion criteria of this scoping review. In addition, given that cognitive impairment can begin many years before dementia syndromes are diagnosed, 58  It also necessary to create new cognitive screening instruments for future studies, with the characteristics common to the MMSE, 34 VFT 36 and PFAQ, 37 such as ease of application, in order to obtain standardized results. General practitioners within primary care services can then apply such instruments to elderly people, in order to be able to refer them for wide-ranging and timely evaluation in specialized services, when necessary.
In the context of aging of the population, it is important that professionals should screen for cognitive impairment, 59,60 as a routine procedure within primary healthcare. Through this, preventive interventions can be provided in order to avoid or minimize the negative effects of dementia on elderly people's health.