Abstract 1435: The Utility of the Montreal Cognitive Assessment as a Screening Measure of Mild Cognitive Impairment in Patients With Heart Failure
Background: Subtle cognitive impairments occur frequently in patients with heart failure (HF), yet sensitive screening measures suitable for use in clinical practice are limited.
Objective: To compare the ability of the Montreal Cognitive Assessment (MoCA) and the Mini Mental State Exam (MMSE) to detect subtle but clinically relevant cognitive dysfunctions in patients with HF.
Methods: Both screening measures were administered to 93 HF patients during hospitalisation. A history of neuro-cognitive problems was an exclusion criterion to the study. Based on previously published results, patients were coded as having mild cognitive impairment (MCI) if they scored <27 on MMSE (possible range 18 –30 based on inclusion criteria) or <26 on MoCA (range 0 –30). Chi-square analysis was used to explore differences in MCI as assessed by the MoCA and MMSE. Receiver operator characteristics (ROC) were examined to determine the sensitivity of MoCA in classifying patients with MCI.
Results: Sixty-eight (73%) patients were coded as having MCI: 31 had scores below the threshold on both MoCA and MMSE. Another 36 had scores <26 on MoCA but >27 on MMSE. Patients with MCI were more likely to have errors on 3 cognitive domains assessed by both MoCA and MMSE: language (LR<0.01); memory recall (LR<0.05); orientation to time and place (LR<0.01). As assessed on the MoCA only, patients with MCI had a 72% increase in the likelihood of executive function errors (OR 1.72; 95% CI 1.12 to 2.62, LR <0.01), 52% increase in visuo-spatial errors (OR 1.52, 95% CI 1.21 to 2.07, LR <0.01) and 42% increase in concentration task errors (OR 1.42, 95% CI 1.17 to 1.71, LR <0.01). Nineteen patients with MCI had errors on the working memory task of MoCA but none had errors on the equivalent task on MMSE. The area under the curve in the ROC analysis was good (.817, p<0.001). Score <26 on MoCA had 97% sensitivity to detect MCI and produced a negative score in cognitively unimpaired patients 29% of the time.
Conclusion: The MoCA, a brief screening measure, may be more sensitive than MMSE at detecting subtle but clinically relevant cognitive problems in patients with HF.