A prospective, single center study conducted at a tertiary hospital (Cardiology Department, Hospital Universitário de Santa Maria, Lisbon, Portugal).

All adult patients (≥18 years-old), consecutively admitted (index-admission) to the cardiology ward with acute HF (de novo or chronic decompensated HF), defined according to criteria in the Guidelines of the European Society of Cardiology (ESC),1 were potentially eligible for inclusion in a post-discharge follow-up HF-program (by protocol), after written informed consent was obtained. This structured post-discharge follow-up program has already been described elsewhere.24 Specific and validated questionnaires for assessment of QoL, and evaluation of cognitive status and depression and chronic anxiety states by application of specific and validated questionnaires at pre-specified times, are components of the program. Patients with language barriers (unable to speak and/or understand Portuguese), visual and auditory deficits, and previously diagnosed neurological or psychiatric disease were excluded from the present study.

Cognitive status, chronic anxiety and depression, were screened at hospital discharge, when the patients were stable, to reduce a possible confounder related to disease decompensation and hospitalization, and between the 6th and the 12th month after discharge.

The most recent ESC Guidelines for the diagnosis and treatment of acute and chronic HF1 stated that cognitive function can be assessed using the Mini-Mental State Examination (MMSE)25 or the Montreal Cognitive Assessment (MoCA).26 In this study the validated Portuguese version of the MoCA was used. It is a cognitive screening instrument with greater sensitivity than the MMSE to milder stages of cognitive decline.31 It is a one-page test with a maximum score of 30 points, which defines CI according to a score of <22 (cut-off validated for the Portuguese population),28,29 assessing eight cognitive domains: executive functions; visuospatial abilities; short-term memory; language; attention, concentration, and working memory; temporal and spatial orientation.26,30

The MoCA score was assessed as a continuous variable to compare evolution over time and as a dichotomous variable based on a cutoff score of 22 to distinguish patients with a normal versus an abnormal score.

Chronic anxiety and depression were assessed using the validated Portuguese version of the Hospital Anxiety and Depression Scale (HADS).27 The HADS is composed of seven questions relevant to either generalized anxiety or depression, with each item containing a four-point (0-3) response category. The possible scores range from 0 to 21 for anxiety and the same for depression. Anxiety and depression should be scored separately. A score of 0 to 7 for either subscale can be regarded as being in the normal range, a score of 8 to 10 suggests a borderline case, and a score of 11 to 21 suggests an abnormal case.32–34

The validated Portuguese version of the Kansas City Cardiomyopathy Questionnaire (KCCQ), a responsive health-related QoL measure for HF, was also used.30 The KCCQ is a 23-item, self-administered instrument that quantifies two main domains composed of multiple subdomains: symptoms (physical function and symptoms frequency, severity and recent change), and global QoL (social function, self-efficacy and knowledge and quality of life). Scores are transformed to a range of 0-100, in which higher scores reflect a better health status.35,36

The KCCQ was answered at hospital discharge (referring to the QoL status previous to hospital admission), and during follow-up at three months and between the 6th and the 12th month after discharge.

Demographic and clinical data, including laboratory values, echocardiography and electrocardiography results, and other instrumental data were collected from the patient's clinical file.

Prognostic outcomes were assessed as a composite endpoint of all-cause readmissions or mortality and individual endpoints of all-cause readmissions and HF readmissions. The association of CI with QoL was also evaluated. The occurrence of death and rehospitalization and the cause of rehospitalization were identified based on the information available in the clinical records.

Statistical analysis was performed using IBM® SPSS® Statistics 20 (Chicago, IL, United States). Categorical variables were reported in absolute numbers and percentage and continuous variables were reported as mean and standard deviation (SD) or median and interquartile range. The impact of cognitive status on endpoints was assessed using the statistical methods of Cox Regression and Kaplan-Meier survival analysis. Anxiety and depressive status were used as covariates and included together with cognitive status (evaluated by MoCA and HADS questionnaires) on multivariate analyses (adjusted for age, left ventricle ejection fraction (LVEF) and NYHA functional class at discharge). Wilcoxon's test was used to evaluate MoCA value variation during follow-up. P<0.05 was considered statistically significant.

The study was approved by the Institutional Ethics Committee and by the National Committee on Data Protection. Patient confidentiality was ensured through the anonymization of the collected data and written informed consent was obtained from all patients before enrolment. All study procedures were carried out in accordance with the ethical principles of the 2013 revised version of the Declaration of Helsinki.37

A total of 43 patients were included in the present study. The mean follow-up period was 8.2±2.1 months. Patient demographic and clinical characteristics at baseline are described in Table 1. Mean age was 67±11.3 years and 31 patients (68.9%) were male. The most frequent HF etiology was idiopathic dilated cardiomyopathy (55.8%), followed by ischemic heart disease (23%). Comorbidities (systemic hypertension, chronic kidney disease, atrial fibrillation and diabetes) were highly prevalent.

The median LVEF assessed by transthoracic echocardiography was 30.3±12.5%, and 37 (86%) patients had HF with reduced ejection fraction (HFrEF). Prior to index-hospitalization, 25 patients (58%) were on treatment with an angiotensin-converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB), 25 (58%) with a beta-blocker and 10 (23%) with a mineralocorticoid receptor antagonist. Index-hospitalization was the first HF admission for 17 patients (39.5%). Nearly half of the patients (48.8%) were in NYHA functional class III on admission and 95% were discharged in NYHA functional class I or II. The mean length of stay was 14.6±16.9 days. Neurohormonal modulation therapy was successfully initiated and/or up-titrated during the hospital stay and optimized treatment (drugs, and devices if indicated) was attempted in all patients during follow-up. The mean score on MoCA at-discharge was 22.6±4.2 and 11 patients (25.6%) had an abnormal score affecting mainly the cognitive domains of visuospatial/executive capacity, language and delayed recall (Table 2, Figure 1).

Figure 1.

Examples of patient performance in assessing visuospatial executive domain (A: Copy the cube; B: Complete the sequence; C: Draw a clock: 11:10).

(0.18MB).

A better performance on MoCA score at discharge was achieved by the male population (p=0.026) and by patients on ACEi or ARB prior to the index-hospitalization (p=0.009).

At discharge, six patients (14%) had an abnormal HADS for depression, 12 patients (27.9%) obtained a borderline value, and 25 patients (58%) had a normal score.

Regarding HADS for anxiety, 24 patients (55.8%) had a normal score at discharge, 11 patients (25.6%) had a borderline score, and 8 (19%) had an abnormal score.

Neither depression nor anxiety screened by HADS was associated with an abnormal MoCA score.

Follow-up clinical data are detailed in Table 3. When comparing the results of the MoCA scores at baseline and at the follow-up evaluations, no significant difference was found (mean scores: 22.6±4.2 vs. 22.2±5.5; p=non-significant [NS]), and the total number of patients with an abnormal score was also similar (11 (25.6%) vs. 10 (23.3%); p=NS). However, a significant reduction in the naming ability domain was observed during follow-up (97.4±9.0% vs. 90.5±17.8%; p=0.031) but this result was not considered clinically relevant (Table 2).

The mean HADS for depression and anxiety did not change significantly during follow-up (6.7±4.0 vs. 5.63±4.5; p=0.137; 6.98±4.4 vs. 6.57±4.4; p=0.624, respectively). However, some individual variation was found. During follow-up, 29 (67.4%) patients had a normal HADS for depression, 9 (20.9%) a borderline value and 5 (11.6%) an abnormal score; and 28 (65.1%) patients had a normal HADS for anxiety, 9 (20.9%) a borderline value and 6 (14.0%) an abnormal score (Figure 2).

Figure 2.

Hospital Anxiety and Depression Scale variation over the time for depression (A) and anxiety (B). Legend: HADS: Hospital Anxiety and Depression Scale.

(0.13MB).

During follow-up, one patient died (2.33%), 14 (32.6%) patients where readmitted (20.9% due to HF) and 15 (34.9%) patients had the composite endpoint of all-cause rehospitalization and/or mortality. The presence of mild CI at discharge (reflected by a MoCA <22) was highly predictive of adverse events during the follow-up period: in a multivariate analysis a MoCA score <22 (vs. ≥22) at discharge increases sixfold the risk of HF readmissions (HR=6.42 [1.23-32.61]; p=0.025) (Figure 3A). Also, the presence of CI was an independent predictor of all-cause readmissions (Hazard ratio [HR]=4.00 (1.15-13.95); p=0.03) (Figure 3B) and conferred an almost five fold greater risk of the composite endpoint of all-cause readmission or death (HR=4.63 [1.37-15.67]; p=0.014) (Figure 3C).

Figure 3.

Kaplan-Meier curves for (A) HF hospital readmission, (B) all-cause readmission, and (C) the composite endpoint of all-cause readmission or death, according to the Montreal Cognitive Assessment score.

(0.2MB).

There was no significant difference in KCCQ main domains (symptoms and global QoL) when comparing patients with vs. without CI (65.1±26.3% vs. 64.9±25.2%; p=NS; 60.5±18.4% vs. 64.4±19.4%; p=NS, respectively), but patients with a higher MoCA score performance at discharge, and therefore with better cognitive status, had a better ability to deal with their HF according to the assessment of this component in the KCCQ (p=0.038).

The authors recognize some limitations in their work, including the fact that this study was from a single center and included only 43 patients, which may also compromise the statistical analysis in a model with several covariates. However, this represents real-world practice including an HF population admitted to hospital due to acute HF and followed-up prospectively in an outpatient HF clinic setting. The observed data on rates of cognitive impairment, anxiety and depression were similar to those obtained in studies with larger HF populations. It is plausible that the 8±2 month period of follow-up may not have been long enough to capture changes in cognitive performance in HF even with therapeutic optimization, however, some of the previously published studies have similar or shorter follow-up.20 Similar studies with a longer follow-up are needed. Another concern is the possibility of false positives and negatives in MoCA and HADS results. Also, although patients with previously diagnosed neurological or psychiatric disease were excluded, brain imaging studies were not routinely performed, so the presence of chronic brain lesions that might contribute to our results cannot be fully ruled out. Nevertheless, no patients had evidence of neurologic deficits, either at inclusion or during the follow-up period.

It is also important to note that most of the previous published studies on the prognostic impact of CI in HF used batteries of neuropsychological tests instead of screening tools.7 Although neuropsychological tests provide information about function across multiple domains and have diagnostic acuity, their application usually takes several hours and requires specialized training in the instruments and interpretating them.55 Cannon et al.7 previously recommended an increasing use of cognitive assessment with standardized screening tools in all future HF studies. These cognitive screening tools provide initial insight into an individual's cognitive status and assist with a decision for referral for more comprehensive neuropsychological testing where the presence of CI will be determined.