CRD summary

Brief screening instruments for cognitive impairment can adequately detect dementia. Whether interventions for patients or caregivers have a clinically significant benefit is still unclear. These conclusions are generally appropriate and reflect the data. However, given the lack of studies for most cognitive screening instruments, the conclusions on diagnostic performance may be overstated.

Authors' objectives

To review the diagnostic accuracy of brief cognitive screening instruments and the benefits and harms of interventions for early cognitive impairment in older adults.

Searching

Systematic reviews published since 2001 were identified by searching five databases including MEDLINE and The Cochrane Library (including DARE). Primary studies in English from these reviews were obtained. The reviews were updated by searching MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL) and PsycINFO to December 2012. Search terms were not described. Expert suggestions, reference lists of reviews and trial registries were used.

Study selection

For diagnostic accuracy, eligible studies were any study (except case-control studies) that assessed a brief screening instrument for cognitive impairment (taking no more than 20 minutes to perform) in community dwelling older adults who were not selected for cognitive impairment.

For intervention studies, any pharmacological, non-pharmacological or decision-making intervention in older adults with mild cognitive impairment or moderate dementia (excluding Parkinson dementia) was eligible. For efficacy, only randomised trials were considered; for safety and harms, large (more than 1,000 patients) cohort and case-control studies were included.

Only studies of fair to good quality applicable to primary care in the United States were included. Eligible outcomes were broadly defined and covered any level of decision-making: diagnostic accuracy or health and safety.

Two reviewers independently assessed studies for inclusion. Disagreements were resolved by consensus or a third person.

Assessment of study quality

Two reviewers independently assessed quality using a range of tools including: United States Preventive Services Task Force (USPSTF) criteria, National Institute for Health and Care Excellence (NICE) checklists, AMSTAR (for reviews), Newcastle-Ottawa (for observational studies) and QUADAS (for diagnostic studies).

Data extraction

For diagnostic accuracy studies, sensitivity and specificity were extracted with their 95% confidence intervals. For pharmacological efficacy trials, mean differences in Alzheimer’s Disease Assessment Scale Cognitive Subscale (ADAS-cog) with 95% confidence intervals were extracted. For harms or pharmacological interventions, relative risks of adverse events were extracted. For non-pharmacological interventions, standardised mean differences in effectiveness (Hedges g) were extracted.

One reviewer extracted data, which were checked by a second.

Methods of synthesis

All studies were combined in meta-analyses. For diagnostic accuracy studies, the bivariate model was used to jointly model sensitivity and specificity. Heterogeneity was assessed using Ι² and Cochran’s Q. Meta-regression was used to investigate heterogeneity. Publication bias tests were performed.

Results of the review

Diagnostic accuracy

Fifty-five studies were included. Four were of high quality and the rest were fair quality. A meta-analysis was performed only for the Mini Mental State Exam (MMSE) (14 studies, 10,185 participants). This had acceptable test performance with 88.3% sensitivity (95% CI 81.3% to 92.9%) and 86.2% specificity (95% CI 81.8% to 89.7%).

Other tests were deemed also to have acceptable diagnostic performance (results presented) but only four tests had five or more studies: Clock Drawing Test, Memory Impairment Screen, Verbal fluency and IQCODE (Informant Questionnaire on Cognitive Decline in the Elderly).

Pharmacologic interventions

There were 48 fair to good quality trials of acetylcholinesterase inhibitors (AChEIs). AChEIs produced small improvements in ADAS-cog scores that may not be clinically meaningful. Donepezil reduced scores by -2.03 (95% CI -2.68 to -1.38), galantamine by -2.25 (95% CI -2.94 to -1.55) and rivastigmine by -3.06 (95% CI -4.48 to -1.65).

The 10 fair to good quality trials of memantine showed a pooled reduction in ADAS-cog score of -1.36 (95% CI -2.02 to -0.7). Several other interventions were considered including aspirin dietary supplements: none showed any significant benefit.

AChEIs increased the risk of withdrawal due to adverse events (donepezil RR 1.79, 95% CI 1.50 to 2.13; results for other drugs were reported) but memantine did not have an increased risk of withdrawal.

Non-pharmacologic interventions

There were 32 trials (4,668 participants) of non-pharmacologic interventions aimed at the patient. Cognitive stimulation (with or without cognitive training) appeared to slightly improve cognition in patients with mild cognitive impairment (SMD -0.59, 95% CI -0.93 to -0.25; six studies). Ten trials of exercise showed no benefit on cognition outcome or patient depression. Other interventions either had inconsistent results or were not widely reported.

Caregiver interventions

There were 59 trials (8,932 participants) in this area. Psychoeducational interventions (including group, individual telephone and care management interventions) showed a slight improvement on caregiver burden (SMD -0.23, 95% CI -0.35 to -0.12) and caregiver depression (SMD -0.21 95% CI -0.30 to -0.13).

Authors' conclusions

Brief screening instruments for cognitive impairment can adequately detect dementia but there was no evidence that this improved decision making in older adults. Whether interventions for patients or caregivers have a clinically significant effect is still unclear.

CRD commentary

This well-conducted review addressed a very broad research question using appropriate inclusion criteria. A suitable search was performed. The search included searches for unpublished material but only studies published in English were included. Action was taken to reduce reviewer error and bias. Study quality was assessed and generally found to be fair at best. The review included a very large number of studies but very few studies were available for many of the cognitive tests and interventions. Where feasible, studies were combined in meta-analyses. The authors noted that there was often diversity in the studies and inconsistencies in their outcomes and that data on important health outcomes were sparsely reported.

The authors’ conclusions were generally appropriate and reflected the data. However, given the lack of studies for most cognitive screening instruments, their conclusions on diagnostic performance may be overstated.

Implications of the review for practice and research

Practice: The authors made no specific recommendations for medical practice.

Research: The authors suggested that more research was needed on which criteria should lead to performing a cognitive assessment and on improving diagnostic criteria for mild cognitive impairment.

Funding

Agency for Healthcare Research and Quality, USA.

Bibliographic details

Lin JS, O'Connor E, Rossom RC, Perdue LA, Eckstrom E. Screening for cognitive impairment in older adults: a systematic review for the US Preventive Services Task Force. Annals of Internal Medicine 2013; 159(9): 601-612. [PubMed: 24145578]

Original Paper URL

http://annals.org/article.aspx?articleid=1760977

Indexing Status

Subject indexing assigned by NLM

MeSH

Aged; Caregivers; Cholinesterase Inhibitors /therapeutic use; Decision Making; Dementia /diagnosis /drug therapy /therapy; Dopamine Agents /therapeutic use; Early Diagnosis; Humans; Mass Screening /methods /standards; Memantine /therapeutic use; Psychological Tests; Risk Assessment

AccessionNumber

12013061329

Database entry date

25/10/2013

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.