Factors predicting misclassification in brief cognitive testing identified
medwireNews: Over a third of patients tested for dementia in the primary care setting may be misclassified due to specific biases in commonly used brief cognitive tests, study findings indicate.
The findings show that factors predicting such misclassification tend to be specific to single tests and knowledge of these could aid clinical decision-making.
The three tests studied were the Mini-Mental State Examination (MMSE), the Memory Impairment Screen (MSI) and the animal naming (AN) test, and the misclassification rates for each were 21%, 16% and 14%.
This was among 824 participants, aged an average of 82 years, of whom 35.3% tested positive for dementia when fully assessed according to DSM III-R/IV criteria.
Misclassification on at least one of the three tests occurred in 301 (35.7%) patients, whereas only 14 (1.7%) were misclassified by all three. False-positive diagnoses were more common than false-negative ones, with ranges of 6.8% to 18.6% versus 2.4% to 7.7%.
The researchers highlight in Neurology: Clinical Practice that absence of informant-rated poor memory was the only factor that consistently predicted misclassification across all three tests, generally associated with false-negative results.
The absence of this information therefore needs to accounted for during testing, say David Llewellyn (University of Exeter Medical School, UK) and colleagues.
“This can be done by incorporating informant-reported cognition into the test (e.g., the General Practitioner assessment of Cognition) or by combining brief cognitive assessment results with a complementary informant-rating scale (e.g., the Informant Questionnaire on Cognitive Decline)”, the team suggests.
False-positive findings were consistently predicted by age, nursing home residency – possibly reflecting acute illness and the need for re-assessment – and non-Caucasian ethnicity.
Overall misclassification on the MMSE was significantly more likely among patients of African–American ethnicity (odds ratio [OR]=2.49) and those with heart problems (OR=1.48) and less likely among those with more years of education (OR=0.86 for every extra year in education) and informant-rated poor memory.
Patients tested using the MSI were significantly more likely to be misclassified if they were older (OR=1.06 for every year older), depressed (OR=2.00) or visually impaired (OR=1.66), and less likely if they were apolipoprotin E (APOE) ε4 carriers (OR=0.50) or if they or a carer had rated their memory as poor (OR=0.19).
And for the AN test, heart problems (OR=1.92) and informant-rated memory decline (OR=1.71) significantly increased the risk for misclassification, whereas more years in education (OR=0.91), the presence of hyperlipidaemia (OR=0.45) and informant-rated poor memory (OR=0.34) significantly decreased the risk.
“Knowledge of factors which predict misclassification and are readily available in clinical practice may improve clinical decision making by enhancing the selection and interpretation of assessments. If an assessment is known to produce biased results for a given patient group, an alternative and more appropriate assessment can be selected”, the researchers comment.
“Alternatively, stratified cut-points could be provided to adjust for known biases, for example level of education on the MMSE and AN. Failing that, clinical judgement can be used to help interpret assessment results more appropriately when the biases affecting misclassification on a given assessment are known.”
By Lucy Piper
medwireNews is an independent medical news service provided by Springer Healthcare. © 2018 Springer Healthcare part of the Springer Nature group