The Association Between Mild Cognitive Impairment and Self-care in Adults With Chronic Heart Failure

Journal of Cardiovascular NursingVol. 00, No. 0, pp 00Y00 x Copyright B 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

The Association Between Mild CognitiveImpairment and Self-care in Adults WithChronic Heart FailureA Systematic Review and Narrative Synthesis

Kay Currie, PhD, RN; Andrew Rideout, MPH, RN; Grace Lindsay, PhD, RN;Karen Harkness, PhD, RN

Background: Emerging evidence suggests that heart failure (HF) patients who have mild cognitive impairment

(MCI) may experience greater difficulty with self-care. Objective: This article reports a systematic review that

addressed the objective ‘‘What is the evidence for an association between MCI and self-care, measured in

1 or more of the self-care domains related to HF, in adults who have a diagnosis of chronic HF?’’ Method:

We adopted Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for the review and

synthesis of quantitative research studies that formally measured both cognitive function and self-care in HF

patients and sought to describe the relationship between these factors. Results: Ninety-one potentially relevant

studies were located; 10 studies (2006Y2014) were included. Because of heterogeneity in the retrieved studies,

meta-analysis was not possible. Narrative synthesis found growing evidence regarding the association between

MCI and adverse effects on self-care in HF. Nine studies reported significant positive associations between MCI

and self-care in HF, either specifically in relation to medication adherence or more generic measures of self-care

activity. One study reported a significant, negative correlation between cognitive function and self-care,

suggesting that worse cognitive function was associated with better self-care; however, this is partially explained

by a small sample size and mixed methodology. Conclusions: These findings have implications for clinical

practice. It is known that HF patients have difficulty with self-care, and the influence of cognitive function needs

to be considered when providing professional support. Further research to determine the feasibility and

acceptability of cognitive assessment in routine clinical care is recommended.

KEY WORDS: heart failure, mild cognitive impairment, review, self-care, systematic

Background

Heart failure (HF) is a syndrome of reduced cardiacoutput, such that the heart cannot maintain adequatetissue perfusion.1 Quality of life for individuals with

HF may be reduced by symptoms of breathlessness,limb edema, and fatigue, often resulting in disturbedsleep, reduced capacity for coping with normal daily ac-tivities or employment, and deteriorating mental well-being.2 The prevalence of HF is 1% to 2% of the generalpopulation, but that increases significantly with age, risingsharply to 10% for individuals around 70 years of age.3

In a comprehensive literature review of epidemiologicalstudies of HF, the overall prevalence ranges from 1% to12%, with variation caused mainly by methodologicaldifferences between studies.4 As the population ages,the burden of HF can be expected to increase in manycountries.3,5

There is growing recognition of the importance ofeffective self-care in the management of HF.3,6 The in-ability to undertake self-care activities is well recognizedas a contributing factor for increased hospital admissionin patients with HF.7,8 Self-care in this context has beendescribed as a naturalistic decision-making process involv-ing knowledge, experience, skill, and compatibility withpersonal values.9 Theoretically, the concept of self-care

1

Kay Currie, PhD, RNReader in Applied Health Research, Nursing & Community Health,Glasgow Caledonian University, Scotland, UK.

Andrew Rideout, MPH, RNResearcher, Nursing & Community Health, Glasgow CaledonianUniversity, Scotland, UK.

Grace Lindsay, PhD, RNReader in Nursing, Nursing & Community Health, GlasgowCaledonian University, Scotland, UK.

Karen Harkness, PhD, RNAssistant Clinical Professor, School of Nursing, McMaster University,Ontario, Canada.

The authors have no conflicts of interest to disclose.

CorrespondenceKay Currie, PhD, RN, School of Health & Life Sciences, GlasgowCaledonian University, Cowcaddens Road, Glasgow G4 0BA,Scotland, UK ([email protected]).

DOI: 10.1097/JCN.0000000000000173

Copyright © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

in HF has been explained in terms of 3 discrete yet over-lapping components. These are self-care maintenance(involving activities associated with symptom monitoringand treatment adherence), self-care management (wherepatients recognize and respond to their symptoms by im-plementing self-care activities), and self-care confidence(acts as a mediator and moderator of the outcomes ofthe self-care process).9 Recommendations for patient be-haviors or activities to promote self-care include adviceregarding medication, fluid and sodium management,nutrition and weight management, smoking cessation,alcohol limitation, and physical activity,10 commonly de-scribed as the ‘‘domains’’ of self-care in the context of HF.

Specific tools have been developed and validated tomeasure self-care in patients with HF, the most com-monly cited being the Self-care in Heart Failure Index(SCHFI)11 and the European Heart Failure Self-care Be-haviour Scale.12 Both tools were developed based on aninitial concept analysis to construct scale items and sub-sequent psychometric testing using pooled data drawnfrom multicenter studies; however, it should be noted thatthese tools rely on patient self-report, which may under-report or overreport actual self-care ability, and a singleassessment may not reflect variation in self-care over time.

Self-care in HF patients involves a complex cognitivedecision-making process; however, the pathophysiologyand symptoms associated with HF may have an adverseimpact on cognitive functioning.13 Therefore, it is rea-sonable to consider the potential contribution of cognitivefunction on engagement in self-care in HF. Cognitiondescribes the intellectual functions required to manageindependently within one’s environment. These includememory, language, attention, visuospatial ability, behav-ior, and executive functions.13 Cognition in aging per-sons follows along a spectrum ranging from normalfunction to severe impairment, known as dementia. Mildcognitive impairment (MCI) is described as an interme-diate step between normal cognitive function and de-mentia that make performance of some activities of dailyliving more difficult than usual (eg, unable to organizemedications or appointments without a memory aid ortake care of finances) but are not severe enough to im-pair basic activities of daily living (eg, dressing, eating,toileting).14 Formal diagnosis of MCI requires evidenceof abnormalities on neuropsychological testing (1.5 SDbelow age-standardized mean) in at least 1 cognitivedomain, with or without memory impairment.15

The presence of cognitive impairment is well docu-mented in patients with HF, with a prevalence rangingbetween 25% and 75%,16Y19 depending on the defi-nition used and the cognitive domains tested. The Mini-Mental State Examination (MMSE)20 is a widely usedinstrument for cognitive testing in older persons, with orwithout HF.16,21,22 However, it is acknowledged thatthe MMSE is not sensitive for detecting MCI and haslimited sensitivity to executive functions.22,23 In other

words, people with HF and MCI will often score withinthe normal range on the MMSE, and therefore, the pre-sence of MCI is missed.21,24,25 The Montreal CognitiveAssessment (MoCA) is recommended for use in patientswith vascular disease26 and is sensitive to cognitive de-ficits detected in older HF patients. Recent studies havereported a prevalence of MCI, as defined by a totalMoCA score lower than 26, in 54% to 75% of olderpatients with HF.27Y29 Difficulties with memory, atten-tion, psychomotor speed, verbal learning, and executivefunctions are the most common types of cognitive de-ficits reported.30,31 It is often difficult to detect MCI ina clinical setting without formal cognitive testing or spe-cifically asking about signs such as a decline in managingfinances, organizing medications, or performing otherusual instrumental activities.14

Self-care imposes a high cognitive demand on patientswith HF, and emerging evidence suggests that patientswith MCI may experience difficulty with self-care,25,32,33

although the evidence for this association is arguablylimited. Evaluation and synthesis of current evidenceare important to inform clinical practice; identifyingwhether MCI can affect specific domains or generalaspects of self-care in patients with HF will enable clin-icians who support patients in developing knowledge,skill, and confidence in self-care to take account of theinfluence of cognitive function.

This article reports a systematic review that aimedto synthesize the best available evidence for an asso-ciation between MCI and self-care in adults who havebeen diagnosed with chronic HF. The review questionwas as follows: What is the evidence for an associationbetween MCI and self-care, measured in 1 or more of theself-care domains related to HF, in adults who have adiagnosis of chronic HF? A review protocol was reg-istered with the Joanna Briggs Institute (protocol 642).

Methods

We adopted the Preferred Reporting Items for System-atic Reviews and Meta-Analyses approach34 to system-atic review and narrative synthesis. The inclusion criteriawere primary quantitative studies that recruited adultpatients (Q18 years) who have a diagnosis of chronicHF, at any stage of the New York Heart Association(NYHA) classification of HF, regardless of sex, age,ethnic origin, or any comorbidity. Our inclusion criteriawere articles that excluded patients who had preexistingdementia or known severe cognitive impairment and thatused validated instruments to measure cognitive functionor impairment that were sensitive to MCI and also mea-sured self-care capability or activities then sought todescribe the relationship between these factors. For thepurposes of this review, MCI was defined as evidence ofimpairment on a validated tool or diagnostic battery inindividuals whodo not have cognitive impairment that

2 Journal of Cardiovascular Nursing x Month 2014


is severe enough to interfere with their basic activitiesof daily living; studies that conducted cognitive testingusing only the MMSE were excluded, as this test is notsensitive to MCI. Definitions of self-care could includemanagement of diet (including healthy diet, salt, and fluidintake), medicines management, avoidance of risk factorsfor disease progression/deterioration, and symptom aware-ness and management. Measurement of self-care may beperformed using validated tools, or more simple mea-sures (such as percentage of missed medications).

Our search strategy was reviewed by an informationmanagement specialist and aimed to find both publishedand unpublished (indexed) studies. We conducted a sys-tematic search of electronic databases, including Medline,EMBASE, CINAHL, PsycINFO, CSA Sociological Ab-stracts, AARPAgeline, SocINDEX, ISI Web of Science:Social Sciences Citation Index and Science Citation IndexExpanded, Ethos, and MedNar. As the concept of self-care in HF developed in the mid-1990s, a date limit of1995 was set for the database search; because of re-source constraints, only articles with English abstractwere included in the search. A 3-phase search strategywas used. The initial search used 3 keywords or phrases:heart failure, cognitive, and self-care. These terms werecombined using the ‘‘AND’’ Boolean operator and en-tered into Medline, CINAHL, and MedNar databases.Retrieved articles were reviewed to identify keywordscontained in the title and abstract, as well as index termsused to describe the article. A second search using allidentified keywords and index terms plus theoreticallyderived terms from the self-care literature was then per-formed across all included databases (Table 1). Third,the reference list of all identified reports was searchedfor additional studies. There was no other attempt toidentify gray literature in the area of interest (search com-pleted March 14, 2014).

All citations located by the search were screened by2 reviewers for relevance based on the title and abstract.Thereafter, the full text of selected articles or theses wasassessed independently by 2 reviewers using the speci-fied inclusion criteria, before appraisal of methodologicalquality using a standardized critical appraisal instru-ment from the Joanna Briggs Institute (JBI-MAStARI).The Figure demonstrates the selection process adopted.

Data from included studies were extracted onto astandardized template by 1 reviewer and independentlyconfirmed by a second, with any discrepancies resolvedthrough discussion with a third reviewer. Data itemsincluded study setting; method; demographic data suchas age, sex, and cohabitation status; disease-specificdata (commonly NYHA classification); and a wide va-riety of outcome measures pertaining to the 2 areas ofinterest for this review (cognitive impairment and self-care), including descriptive statistics and results fromregression analyses indicating association between out-come measures.

A heterogeneous range of data collection tools wereused for assessment of variables and data aggregationor meta-analysis was not possible; instead, findings ex-tracted from all included studies were inserted into amatrix for ease of comparison for narrative synthesis.Measures used to assess cognitive function and self-carewere highlighted to identify possible similarities; outcomeswere classified as self-care related to either medicationadherence (a specific focus of several studies) or a broaderrange of self-care activities. Studies were further cate-gorized into those that demonstrated a statistically sig-nificant association between MCI and self-care and thosethat did not. Table 2 presents summary details of the10 studies25,35,39Y43 included in this comparative matrix.All were observational correlation designs, published asfull-text articles between 2006 and 2014; 2 studies42,43

included mixed methods; and 1 study38 was a compa-rative cohort design. Eight studies were conducted in theUnited States,35Y39,41,43 1 was conducted in Australia,25

and 1 in Canada.40

Results

Ten studies were included in the review, involving atotal of 1842 patients. Sample size ranged from 29 to557 participants. One study42 purposively sampled ex-treme cases (poor or expert in self-care; n = 29), andanother43 used a purposive homogenous sampling tech-nique. All other studies recruited convenience samplesthat met inclusion criteria.

TABLE 1 Summary of Search Terms Used Across

All Included Databases

S1 MM ‘‘Heart Failure +’’S2 ‘‘Cardiac failure’’S3 ‘‘Ventricular failure’’S4 ‘‘Pulmonary oedema’’ OR ‘‘Pulmonary Edema’’S5 ‘‘Dyspnoea+’’S6 S1 OR S2 OR S3 OR S4 OR S5S7 MM ‘‘cognitive impairment’’S8 ‘‘Mild cognitive impairment’’S9 ‘‘Cognitive disorders+’’

S10 ‘‘Cognitive dysfunction’’S11 ‘‘Cognitive function’’S12 ‘‘Executive function’’S13 S7 OR S8 OR S9 OR S10 OR S11 OR S12S14 MM ‘‘Self-care+’’S12 ‘‘Self-management’’S13 ‘‘Medicines management’’S14 ‘‘Medication management’’S15 ‘‘Medication adherence’’S16 ‘‘Exercise’’S17 ‘‘Diet’’S18 ‘‘Diet, sodium-restricted’’S19 ‘‘Fluid intake’’S20 ‘‘Fluid restriction’’S21 ‘‘Smoking’’S22 S14ORS15ORS16S17ORS17ORS18ORS19ORS20ORS21S23 S6 AND S13 AND S22

Mild Cognitive Impairment and Self-care in Chronic HF 3


Two studies25,39 recruited HF patients during anhospital admission; the other 8 studies recruited from HFoutpatient clinics or management programs.35Y38,40Y43

Most of both inpatient and clinic participants were re-cruited from single facilities, with only 3 studies25,38,43

recruiting from more than 1 site. Participant ages rangedfrom 25 to 93 years, with mean ages of 49 to 73 yearsreported. Two studies targeted older patients (945 years25

and 955 years40) to increase representativeness of theHF population. Most participants were male. One largersample study36 recruited from a veterans’ facility, and menconstituted 99% of the sample. Eight studies25,35,36,38,40Y43

provided NYHA classification details for participants,indicating functional severity of HF; most participantshad at least NYHA stage III HF, suggesting marked lim-itation in activity due to HF symptoms. Where reported,participants were urban dwelling and predominantlywhite, although minority groups were represented inmost studies. Five studies25,37,39,41,42 reported whetherparticipants lived alone or with another person; in thesesamples, most lived with another person, with 13% to49.7% living alone.

A variety of cognitive assessment tools were used inthe reported studies; each included study used at least1 technique known to be sensitive in detecting MCI.One study used MMSE and MoCA25; 2 used MoCAalone40,41; 1 used MMSE in combination with a range

of other cognitive assessment tools35; and the remaining6 studies used various other assessment tools excludingthe more commonly applied MMSE or MoCA.36Y39,42,43

Most articles did not differentiate between degrees ofcognitive impairment in their findings (except for Hawkinset al37). However, as we included only studies that ex-cluded people with dementia or significant cognitive im-pairment and that used tools sensitive to MCI, we caninfer that the findings represent the association betweenMCI and self-care.

All included studies measured some dimension ofself-care capability or activities as outcomes of interest,addressing either medication adherence or a more com-prehensive assessment of self-care activities. Four studiesmeasured medication adherence using various report-ing scales35Y38; 3 studies used the validated SCHFIalone25,40,42; 1 study used the European Heart FailureSelf-care Behaviour scale39; 1 study used SCHFI and theEuropean Heart Failure Self-care Behaviour scale41; and1 study used SCHFI and the Dutch Heart Failure Knowl-edge Scale.43

Review Findings

From the 10 studies reviewed, 8 studies25,35Y41 reportedstatistically significant results indicating a positive corre-lation between MCI and self-care ability, with 1 study43

reporting a negative correlation between MCI and self-care (suggesting poorer cognitive function leads to betterself-care). One study42 did not report P values. Only4 studies37,39,40,43 reported confidence intervals for thefindings related to cognitive function alone; these werewide, suggesting that the effect size may be rather smallor of poor precision because of small sample size. How-ever, there is persuasive evidence that MCI affects HF-self-care ability adversely.

Measurement of selected variables was largely hetero-geneous in the included studies, with the variety of as-sessment tools and outcome measures, meaning thatmeta-analysis was not possible. Therefore, the follow-ing results are presented as a narrative synthesis of (a)those studies that found a statistically significant asso-ciation between MCI and some dimension of self-careand (b) the single study that reported a negative cor-relation between MCI and self-care (see Table 2).

Impaired Cognitive Function Leads toPoorer Self-care

Four studies35Y38 demonstrated a statistically significantassociation between MCI and poorer medication ad-herence. These studies had reasonable sample sizes (n =122Y280), and each used a combination of validated teststo measure cognitive function, thus increasing the like-lihood of detection of MCI. In 3 of 4 studies, medication

FIGURE. Selection process.



TABLE2

IncludedStu

dies

Author

Design

Setting

Age;Gender

NYHA

Class

Lived

Alone

Cognition

Outcome

Measu

res

Self-care

Outcome

Measu

res

Resu

ltsand

Conclusions

Studiesthatdemonstratedpositive

correlationbetw

eenMCIandpoorermedicationadherence

1.

Aloscoetal

(2012)35

Retrospective

observational

correlational

studyvia

hierarchical

regression

analysis

HFmanagement

program

inOhio,

United

States

59Y7

7y

(mean,68y)

F:M

29:93

2Y3

Notgiven

MMSE

TMT-A,

TMT-B

LawtonBrody

ADLScale

Functionalability

measuredand

arangeof

activities

including

medication

management

Cognitivefunctionshowed

increm

entalp

redictive

validityfor

medicationmanagement(R

2

change=0.14,PG.001;no

confidence

intervalsreported).

Significan

tassociationsare

TMT-A

withmed

ications,MMSEwith

driving,andgenderwithfinances.

Convenience

sample

of

consecutive

casesfrom

largeHF

database

InpersonswithHF,

cognitive

perform

ance

isanindependent

predictorofindependence

inmedicationmanagement.Persons

withpoorercognitivetest

perform

ance

(TMT-A)reported

greaterdifficultywithmed

ication

management.

Exclusioncriteria

relatedto

brain

function,

eg,stroke,

headinjury

withLO

CN=122

2.

Aloscoetal

(2013)36

Cross-sectional,

descriptive

studyusing

hierarchical

regression

analysis

OPDcardiology

clinicsin

Ohio

Mean,68y

As%

:1,0.5%

;2,84%;3,14%;

4,1.5%

Notgiven

TMT-A,

TMT-B

LawtonBrody

ADLScale

Significantbivariate

correlationan

dbetween(a)executive

function

(TMT-A)andmedicationmanagem

ent

(correlationcoefficient=0.18,

PG.05)and(b)attention(LNS)

and

medicationmanagem

ent(correlation

coefficient=0.15;PG.05).

Hierarchicalregressionsrevealed

that

reducedexecutive

functionwas

indepen

den

tlyassociated

witha

decreasedabilitytomanagemedications.

Noconfidence

intervalgiven.

N=197

F:M

70:127

DigitSymbol

Coding

Functionalability

measuredand

arangeof

activities

including

medication

management

Excluded

peo

ple

withdementia,

otherneurological

history

LNS

3.

Haw

kinsetal

(2012)37

Prospective

observational

correlational

studyvialinear

regression

HFveterans

OPDfacilityin

theUnited

States

33Y9

3y

(mean,66y)

Notgiven

27.3%

(68)

SaintLouis

University

Mental

Status

MA

Estim

ator

Compared

withnoCI,MAsignifican

tly

worsen

edby8percentagepoints

(78%

to70%,P

=.017)forpatients

withMCIbutdid

notcontinueto

worsen

forpatients

withsevere

CI

inthedem

entiarange.

(continues)



TABLE2

Included

Stu

dies,

Continued

Author

Design

Setting

Age;Gender

NYHA

Class

LivedAlone

CognitionOutcome

Measu

res

Self-care

Outcome

Measu

res

Resu

ltsand

Conclusions

Convenience

sample

F:M

4:247

Battery

ofAssessmentof

neuropsychologicalstatus;

TMTs;Wechsler

TestofAdult

Reading;GroovedPegboard

Test;phonem

icandsemantic

verbalfluency

tasks;

Similarities,Matrix

Reasoning,

LNS,andDigitSpansubtests

from

theWechsler

Adult

Intelligence

Scale,4

thed

ition

Reg

ressionan

alyses

revealed

arobustassociationbetween

thepresence

ofCIandMA

(P=.017).Amongthe

studyvariab

les,onlyCIw

asfoundto

haveastatistically

significantassociationwith

MA.

Nosubjectshad

knownCI

beforeenrolment

Studyfoundthatunrecognized

CIw

ashighlyprevalent

andassociated

withpoorer

adheren

ceto

med

ication.

(Theseresultsarenot

necessarilysupported

bythe

confidence

intervals,which

suggestnodifference

betweengroups)

N=251

4.

Riegeletal

(2011)38

Prospective

comparative

observational

correlationstudy

vialogistic

regressionmodels

HFoutpatient

clinicsin

3UShospitals

50Y7

4(m

ean62)

Majority

3Notgiven

TelephoneInterview

of

CognitiveStatus

MAVBasel

Assessm

entof

Adherence

Scale

Theonlycognitionmeasure

significan

tlyassociated

withMAwas

attentionPV

T(P

=.047;noconfiden

ceintervalsreported).

Astratified

sampling

techniquewas

used,based

on

EDSan

dCI

F:M

100:180

PVT

Nonadherence

was

significantly

more

commonin

those

withEDS,

regardless

of

cognitivestatus(P

=.035).

N=280(242

completed)

TMT-B

Secondary

modelsusingthe

EDSscoreandtheindividual

cognitiontestscores

showed

thatcognitive

status

wasnotsignificantlyassociated

withMA(P

=.752).

Reportedattrition

andstates

cohorts

becam

esaturated

butnospecific

justificationfor

samplesize

DigitSymbol-SubstitutionTest

EDSmay

causelack

ofvigilance/

attentionan

dmedication

nonadherence,ie,picture

morecomplexthan

CIalone.

Theau

thorsacknowledge

thatthose

withCIm

ay

beunderreportingtheir

medicationnonad

heren

ce.

Excluded

ifdem

entia

orneurological

even

t

Probed-M

emory

RecallTask

LNStest

(continues)



TABLE2

Included

Stu

dies,

Continued

Author

Design

Setting

Age;Gender

NYHA

Class

LivedAlone

Cognition

Outcome

Measu

res

Self-care

Outcome

Measu

res

Resu

ltsand

Conclusions

Studiesthatdemonstratedpositive

correlationbetw

eenMCIandpoorercomprehensive

self-care

activity

5.

Cameronetal

(2010)25

Prospective

observational

correlational

studyvia

multiplelinear

regression

analysis

1inpatient

facilityin

Australia

62Y8

4y

(mean,73y)

45%

NYHA

Class

342%

(39)

MMSEandMoCA

SCHFI

Sixty-eightpatients

(75%

)were

codedashavingMCIandhad

significantlylowerself-care

management(P

=.01)and

self-confidence

scoresP=.05)

(noconfidence

intervalsgiven).

Convenience

sample

of

consecutive

patients

F:M

27:66

5%

NYHA

Class

4

MCImadethelargest

contribution,

explaining9%

ofthevariance

inself-care

management.

N=93

CI,ahiddencomorbidity,may

imped

epatients’abilityto

make

appropriate

self-caredecisions.

Dementia,

neurological

excluded

6.

Hadjuketal

(2013)39

Prospective

observational

cohortstudy

Patients

hospitalized

with

prim

ary

orsecondary

diagnosis

ofHF

Meanage,71y

Notgiven

287(49.7%

)5word

immediate

anddelayed

memory

test

(MoCA),

ControlledOral

Word

Association

Test,DSST

EHFScBS-9

Memory,processingspeed,and

executive

functionwere

impairedin

33.3%

,40.0%

,and

56.0%

ofpatients,respectively;

more

thanthree-quarters

(79%

)ofpatients

were

impairedin

at

least

1ofthe3domains.Overall

cognitive

statuswas

notassociated

withperform

ance

ofself-care

activities

inmultivariableadjusted

regressionmodels;however,

impairmentin

specificdomains

affectedself-care.

n=577

Multiple

linear

regression

analysis

Screenedfor

delirium

5inpatient

facilities

inthe

UnitedStates

F:M

254:323

Memory

impairm

entwas

associated

withsignifican

tlypoorerself-care

(P=.006)("

=j1.87;95%

confiden

ceinterval,j3.20to

j0.54).Im

pairmentsin

processing

speedandexecutive

function

were

notsignificantlyassociated

withself-care

intheadjusted

models.

(continues)



TABLE2

IncludedStu

dies,

Continued

Author

Design

Setting

Age;Gender

NYHA

Class

LivedAlone

Cognition

Outcome

Measu

res

Self-care

Outcome

Measu

res

Resu

ltsand

Conclusions

7.

Harkness

etal

(2013)40

Observational

cross-sectional

studyusing

backward

regression

model

Patients

with

HFattending

aheartfunction

clinicinCan

ada

Mean(SD),

72.4

(9.8)y

1Y3

Notgiven

Geriatric

Depression

Scale,MoCA

SCHFI

MCI,asdefinedbyaMoCAscore

G26,

waspresentin

73%

ofpatients;21%

had

anad

equateself-careman

agem

ent

SCHFIscore;an

d12%

reported

symptomsofdep

ression.Participan

tswithaMoCAscore

G26vs

Q26scored

significan

tlylower

ontheself-care

man

agem

entsubscaleoftheSC

HFI

(mean[SD],48.1

[24]vs

59.3

[22],

respectively;P=.035).Usingbackw

ard

regression,thefinal

model

was

fitted

toself-caremanagem

entwhile

controlling

forageandsexandwas

significant(with

F=7.04,df=3,96,andPG.001),

accountingfor18%

ofthetotalvariance

inself-care

managem

ent(R2=18.03%;

confidence

interval,0.787Y2

.781).

Convenience

sample

ofconsecutive

patients

F:M

32:68

N=100

Excludedif

documented

history

ofCI

8.

Leeetal

(2013)41

Prospective

observational

correlational

studyvia

generalized

linear

modelingand

hierarchicallinear

modeling

Single

HFOPD

facilityin

the

UnitedStates

44Y6

9y

(mean,57y)

F:M

57:91

58.8%

NYHA

Class

3Y4

13%

(9)

MoCA(range,

0Y3

0),using

cutoffscores

forthegeneral

population

(26)andfor

adultswith

cardiovascular

disease

(24)

SCHFI

UsingMoCAscoresof26and24,respectively,

33.1%

and14.2%

ofthesample

hadMCI.Controllingforcommon

confounders,participantswithMoCA

scoreslowerthan

26reported

self-care

comparablewiththat

ofparticipan

tswithMoCAscoresof26orhigher.

ParticipantswithMoCAscoresG24,

however,reported21.5%

worse

self-care

man

agem

ent(P

=.014),

and51%,worseconsultingbehaviors

(P=.001),comparedwithparticipants

withMoCAscoresofQ24

(noconfidence

intervalsgiven

).

Convenience

sample

N=148

EHFScBS-9

(for

consulting

behavior)

Adisease-specificcutoff

forMCIreveals

markeddifferencesin

patients’ability

torecognizeandrespondto

HF

symptomswhentheyoccur.

Dem

entiaV

excluded

(continues)



TABLE2

IncludedStu

dies,

Continued

Author

Design

Setting

Age;Gender

NYHA

Class

LivedAlone

Cognition

Outcome

Measu

res

Self-care

Outcome

Measu

res

Resu

ltsand

Conclusions

9.

Riegeletal

(2007)42

Cross-sectional

mixed-m

ethods

designusing

extremecase

sampling

Single

HF

OPDfacility

intheUnited

States

Givenper

category

F:M

11:18

Majority

3Y4

21%

(6)

PMRTask

DSST

SCHFI

Patients

poorin

HFself-care

had

worsecognition,more

sleepiness,

higherdepression,andpoorerfamily

functioning.Theprim

aryfactors

distinguishingthose

goodvs

expert

inself-careweresleepinessandfamily

engagem

ent(althoughnotestsof

statisticalsignificance

wereconducted).

N=29

Patientswith

severely

impairedcognition

wereexcluded.

Studythatfoundnostatistically

significantcorrelationbetw

eenCIandpoorercomprehensive

self-care

activity

10.

Dicksonetal

(2011)43

Concurrent

triangulation

mixed-m

ethods

design;contentand

correlationalanalysis

with

linearregression

2OPD

HFclinics

inalarge

urbanmedical

centerin

theUnited

States

25Y6

5y

(mean,49y)

F:M

15:26

41.5%

NYHA

Class

258.5%

NYHA

Class

3

Notgiven

DSST

PMR

DutchHeart

Failure

Knowledge

Scale

There

wasasignificantnegative

correlationbetw

eencognitive

function(DSST)andself-care

maintenance

(r=j0.33P=.03).

Purposive

homogenous

sampling

technique

Excluded

those

withdementiaor

neurologicalevent

SCHFI

95%

Confidence

intervalsareallnegative

(ie,d

onotcross0as

theothersdo),

sothisresultmaybeconsidered

significant,althoughtheconfid

ence

intervalsareverywide,suggestin

gpoorprecisiondueto

smallsam

ple

size

(n=41).Therewas

also

asignificant,negativecorrelationbetween

cognitive

function(PMR)andself-care

managem

ent(r=j0.39,P=.049),

indicatingthatworsecognitive

functionwas

associated

withbetter

self-care,

explained

inpartby

mixed

methodologyan

dthe

qualitativenarratives.Itcouldbe

that

someindividualswithMCI

overestim

ated

theirresponseson

theSC

HFI.Reexaminationofqualitative

resultsrevealed

that

socialsupport

(eg,spouses,children,andfriends)

playedim

portan

trolesin

self-care.

N=41

Excludetestsnot

welldescribedin

thesis

Abbreviations:

ADL,

activitiesofdaily

living;CI,cognitiveim

pairment;DSST,DigitSym

bolSubstitutionTest;EDS,excessivedaytim

esleepiness;EHFScBS-9,EuropeanHeartFailure

Self-care

BehaviourScale;

HF,

heartfailure;LN

S,LetterNumberSequencing;LO

C,loss

ofconsciousness;MA,medicationadherence;MCI,mild

cognitiveim

pairment;MMSE,Mini-MentalState

Examination;MoCA,MontrealCognitive

Assessment;NYHA,New

York

Heart

Association;OPD,outpatient;PMR,ProbedMemory

Recall;

PVT,PsychomotorVigilance

Task;SCHFI,Self-care

inHeartFailure

Index;

TMT,TrailMakingTest.



adherence was found to be adversely affected by MCI35Y37;however, although the fourth38 found a relationshipbetween the cognitive function of attention, as measuredby the Psychomotor Vigilance Test, results indicated thatexcessive daytime sleepiness, rather than measured cog-nitive status, is more influential on medication adherence.These findings suggest that the concepts and measure-ment of cognitive function and excessive daytime sleep-iness may require further study to untangle their relativeinfluence.

A further 5 studies25,39Y42 used the Self-care Heart Fail-ure Index or European Heart Failure Self-care BehaviourScale to measure self-care and demonstrated a statisticallysignificant association between MCI and an adverse ef-fecton self-careactivities. Four studies25,39Y41 had relativelylarge samples (n = 93Y577) and reported a statisticallysignificant negative association between MCI (based onMoCA score) and self-care management and help-seekingbehaviors. The other study in this grouping, a mixed-methods descriptive study,42 had a smaller sample size(n = 29), and although reporting that ‘‘Patients poor inHF self-care had worse cognition, more sleepiness, higherdepression, and poorer family functioning’’ (p. 235), onlymean scores for each type of patient (poor, good, ex-pert in self-care) were provided and statistical testingfor differences in scores between patient types was notconducted,

Impaired Cognitive Function Is Not Relatedto Poorer Self-care

Despite the significant association between MCI andself-care described above, 1 study43 did not detect thisrelationship.

Dickson et al43 acknowledge the ‘‘surprising’’ findingof a negative correlation between their measures of cog-nitive function and self-care, indicating that poorer cog-nitive function improves self-care ability, which is intuitivelyperverse. However, this mixed-methods study had asmall sample (n = 41) of relatively young (25Y65 years)participants, which may affect the generalizability of find-ings. In addition, the authors note, ‘‘It could be that someindividuals with MCI overestimated their responses onthe SCHFII Re-examination of qualitative results rev-ealed that social support (eg, spouses, children, and friends)played important roles in self-care’’ (p. 183).

Discussion

Despite the established relationship between HF andcognitive impairment, there is a limited body of evidenceinvestigating the association between MCI and self-carein HF; sensitive measurement of both MCI and self-carecan be challenging, and the lack of consistency in thetools used in existing studies hampers the synthesis or

comparison of findings. In addition, the quality of thestudies reviewed here may have been affected in somecases by the use of small convenience samples drawnfrom single sites and the use of multiple regression whereexplanatory variables were not conditioned to removeimplicit bias, as discussed below.

In relation to the influence of sample size, the resultsof studies reported here have been obtained largelythrough the use of various regression methodologies.Regression is an appropriate methodology and buildson the concept that changes in explanatory variablestypically generate proportionate changes in the outcome(or response) process. Broadly speaking, the accuracywith which a regression coefficient is identified is pro-portional to the square root of the sample size; only themost influential explanatory variables are likely to beidentified from small samples. On the basis of the rec-ommendations of Good and Hardin,44 judging by thesample sizes and number of explanatory variables usedin the included articles, at least 2 studies,37,43 may havesome multiple regressions that were potentially tooambitious.

One potential weakness in a number of the arti-cles25,36,41Y43 under review is the absence of multipleregressions in which the explanatory variables are con-ditioned to remove implicit bias; when investigating thedependence of self-care on explanatory variables, a hier-archical multiple regression is preferable, where the ex-planatory variables are modified, or conditioned, to removethe effect of explanatory variables that potentially intro-duce bias, for example, age, gender, depression, and ex-cessive daytime sleepiness.

There are challenges in measuring both cognitive func-tion and self-care in HF; the reported studies used a widevariety of tools to measure cognitive function, limitingcomparability and preventing meta-analysis. The MMSEwas used in combination with other cognitive tests in2 of the studies25,35 included in this review. The MMSEis not sensitive for detecting MCI that impacts on self-care in HF,21Y25 and studies exploring the relationshipbetween cognitive function and self-care need to sup-plement the MMSE with cognitive tests sensitive to MCI.The MoCA has been shown to be a more sensitiveindicator of MCI in HF.26 The 4 studies25,39Y41 that usedMoCA demonstrated an association between MCI andpoorer self-care.

Self-care measures in all studies were self-reports,and, as acknowledged by some authors38,43 reviewedhere, individuals may not reliably report self-care, witheither overreporting or underreporting of ability orlack of acknowledgement of the role of carers in sup-porting self-care activities, particularly when MCI ispresent.

Finally, evidence presented in some studies suggeststhat predictors of poor self-care are more complexthan MCI alone with depression and excessive daytime



sleepiness also being implicated and their relative influ-ence difficult to untangle from cognitive function.

Conclusions

This systematic review and narrative synthesis has founda growing body of evidence regarding the associationbetween MCI and adverse effects on self-care in HF,either in its broadest sense or more specifically in re-lation to medication adherence. Although some of theresearch conducted thus far may have limitations dueto small sample size and design, the stronger evidence(produced by the use of a wider range of more sensitivemeasures of MCI, within larger sample studies) sug-gests that MCI does adversely affect self-care; however,this is based on findings from only 9 studies, 4 of whichrelate specifically to medication adherence. Further re-search using sensitive measures to detect MCI in ap-propriately selected and sized samples for multiple/hierarchical regressions is warranted. While acknowl-edging the limitations of the studies included in this re-view, nevertheless, our report represents a useful state ofbest evidence in this important aspect of HF.

These findings have implications for clinical practice.It is known that HF patients have difficulty with self-care and the influence of cognitive function needs to beconsidered when providing professional support for self-care in HF, with further research to determine the fea-sibility and acceptability of cognitive assessment inroutine clinical care. Once MCI is recognized, tailoredstrategies should be developed and evaluated to helppatients adapt to the increased challenges in self-carecaused by MCI, to improve clinical outcomes for thispatient group.

Acknowledgment

We acknowledge the support and advice offered by Pro-fessor Alex Clark, University of Alberta, in the prepara-tion of this article.

REFERENCES

1. ICD-10 [Database on the Internet]. World Health Orga-nization. 2010. http://apps.who.int/classifications/icd10/browse/2010/en#/I30-I52. Accessed February 3, 2012.

2. Zambroski C, Moser DK, Bhat G, Zeigler C. Impact ofsymptom prevalence and symptom burden on quality of lifein patients with heart failure. Eur J Card Nurs. 2005;4:198Y206.

3. McMurray JJ, Adamopoulos S, Anker SD, et al. ESC guide-lines for the diagnosis and treatment of acute and chronic heartfailure 2012: the Task Force for the Diagnosis and Treatmentof Acute and Chronic Heart Failure 2012 of the EuropeanSociety of Cardiology. Developed in collaboration with theHeart Failure Association (HFA) of the ESC. Eur Heart J.2012;14:803Y869.

4. Roger VL. Epidemiology of heart failure. Circ Res. 2013;113(6):646Y659. doi:10.1161/CIRCRESAHA.113.300268.

5. McCullough P, Philbin E, Spertus J, Kaatz S, Sandberg K,Weaver W. Confirmation of a heart failure epidemic: findingsfrom the Resource Utilization Among Congestive Heart Failure(REACH) study. J Am Coll Cardiol. 2002;39(1):60Y69.

6. Riegel B, Moser DK, Anker SD, et al. State of the science:promoting self-care in persons with heart failure: a scientificstatement from the American Heart Association. Circulation.2009;120(12):1141Y1163.

7. Riegel B, Dickson VV, Goldberg LR, Deatrick JA. Factorsassociated with the development of expertise in health failureself-care. Nurs Res. 2007;56(4):235Y243.

8. Silcock J. The Self Management of Heart Failure: A Placein Practice? [dissertation]. Leeds: University of Leeds; 2009.

9. Riegel B, Jaarsma T, Stromberg A. A middle-range theory ofself-care of chronic illness. ANS Adv Nurs Sci. 2012;35(3):194Y204.

10. Lainscak M, Blue L, Clark A, et al. Self-care managementof heart failure: practical recommendations from the PatientCare Committee of the Heart Failure Association of the EuropeanSociety of Cardiology. Eur J Heart Fail. 2011;13(2):115Y126.

11. Riegel B, Lee CS, Dickson VV, Carlson B. An update on theself-care of heart failure index. J Cardiovasc Nurs. 2009;24(6):485Y497.

12. Jaarsma T, Arestedt K, Martensson J, Dracup K, Stromberg A.The European Heart Failure Self-care Behaviour scale revisedinto a nine-item scale (EHFScB-9): a reliable and valid interna-tional instrument. Eur J Heart Fail. 2009;11:99Y105.

13. Harkness K, Heckman G, Akhtar-Danesh N, Gunn E, Demers C,McKelvie R. Cognitive function and self-management inolder patients with heart failure [published online aheadof print June 3, 2013]. Eur J Cardiovasc Nurs. doi:10.1177/1474515113492603.

14. Chertkow H, Nasreddine Z, Joanette Y. Mild cognitive im-pairment and cognitive impairment, no dementia, part A:concept and diagnosis. Alzheimer Dement. 2007;3(4):266Y282.

15. Petersen RC, Knopman DS, Boeve BF, et al. Mild cognitiveimpairment: tenyears later.ArchNeurol. 2009;66(12):1447Y1455.

16. Bennett SJ, Sauve MJ. Cognitive deficits in patients with heartfailure: a review of the literature. J Cardiovasc Nurs. 2003;18(3):219Y242.

17. Pressler SJ. Cognitive functioning and chronic heart failure:a review of the literature (2002YJuly 2007). J Cardiovasc Nurs.2008;23(3):239Y249.

18. Vogels RL, Scheltens P, Schroeder-Tanka JM, Weinstein HC.Cognitive impairment in heart failure: a systematic review ofthe literature. Eur J Heart Fail. 2007;9(5):440Y449.

19. Athilingam P, King KB. Heart and brain matters in heart failure:a literature review. J N Y State Nurses Assoc. 2007;38(2):13Y19.

What’s New and Important

h This article represents a useful and important ‘‘stateof best evidence’’ on the association between MCI andself-care in HF.

h The evidence for such an association is limited at present;however, 8 studies that used larger samples and moresensitive indicators of MCI found a statistically significantassociation suggesting that MCI has an adverse impacton self-care in HF.

h We recommend further research to investigate thefeasibility and acceptability of routine assessment ofcognitive function in HF patients to enable practitionersto tailor interventions to support self-care in the presenceof MCI.



http://apps.who.int/classifications/icd10/browse/2010/en#/I30-I52

http://apps.who.int/classifications/icd10/browse/2010/en#/I30-I52

20. Folstein M, Folstein S, McHugh P. ‘‘Mini-mental state’’: apractical method for grading the cognitive state of pa-tients for the clinician. J Psychiatric Res. 1975;12(3):189Y198.

21. Riegel B, Bennett JA, Davis A, et al. Cognitive impairmentin heart failure: issues of measurement and etiology. Am JCrit Care. 2002;11(6):520Y528.

22. Tombaugh T, McIntyre N. The Mini-Mental State Examina-tion: a comprehensive review. J Am Geriatr Soc. 1992;40(9):922Y935.

23. Bauer LC, Johnson JK, Pozehl BJ. Cognition in heart failure:an overview of the concepts and their measures. J Am AcadNurs Pract. 2011;23(11):577Y585.

24. Athilingam P, King KB, Burgin S, Ackerman M, Cushman L,Chen L. Montreal Cognitive Assessment and Mini-mentalStatus Examination compared as cognitive screening toolsin heart failure. Heart & Lung. 2011;40(6):521Y529.

25. Cameron J, Worrall-Carter L, Page K, Riegel B, Lo SK,Stewart S. Does cognitive impairment predict poor self-carein patients with heart failure? Eur J Heart Fail. 2010;12(5):508Y515.

26. Nasreddine Z, Phillips N, Bedirian V, et al. The MontrealCognitive Assessment, MoCA: a brief screening tool for mildcognitive impairment. J AmGeriatr Soc. 2005;53(4):695Y699.

27. Athilingam P. Validation of an Instrument to Measure Cog-nitive Function in PatientsWith Heart Failure [dissertation].Rochester, NY: University of Rochester School of Nursing;2008.

28. Cameron J, Worrall-Carter L, Page K, Stewart S, Ski C.Screening for mild cognitive impairment in patients with heartfailure: Montreal Cognitive Assessment versus Mini MentalState Exam.Eur J Cardiovasc Nurs. 2012;11(2). doi: 10.1177/1474515111435606.

29. Harkness K, Demers C, Heckman GA, McKelvie RS. Screen-ing for cognitive deficits using the Montreal Cognitive Assess-ment tool in outpatients 9=65 years of age with heart failure.Am J Cardiol. 2011;107(8):1203Y1207.

30. Bauer LC, Pozehl B, Hertzog M, Johnson JA, Zimmerman L,Filipi M. A brief neuropsychological battery for use in chronicheart failure population. Eur J Cardiovasc Nurs. 2012;11:23.doi:10.1016/j.ejcnurse.2011.03.007.

31. Vogels RL, Oosterman JM, van Harten B, et al. Profile ofcognitive impairment in chronic heart failure. J Am GeriatrSoc. 2007;55(11):1764Y1770.

32. Dickson VV, Tkacs N, Riegel B. Cognitive influences on self-care decision making in persons with heart failure.AmHeart J.2007;154(3):424Y431.

33. Sloan RS, Pressler SJ. Cognitive deficits in heart failure: re-cognition of vulnerability as a strange new world. J CardiovascNurs. 2009;24(3):241Y248.

34. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA state-ment for reporting systematic reviews and meta-analyses ofstudies that evaluate health care interventions: explanationand elaboration. PLOS Med. 2009;7:e1000100. doi:http://dx.doi.org/10.1136/bmj.b2700.

35. Alosco ML, Spitznagel MB, Cohen R, et al. Cognitive im-pairment is independently associated with reduced instrumen-tal activities of daily living in persons with heart failure.J Cardiovasc Nurs. 2012;27(1):44Y50.

36. Alosco ML, Spitznagel MB, Naftali R, et al. Executive dys-function is independently associated with reduced functionalindependence in heart failure. J Clin Nurs. 2013;23:829Y836.

37. Hawkins LA, Kilian S, Firek A, Kashner TM, Firek CJ, Silvet H.Cognitive impairment and medication adherence in outpatientswith heart failure. Heart Lung. 2012;41(6):572Y582.

38. Riegel B, Moelter ST, Ratcliffe SJ, et al. Excessive daytimesleepiness is associated with poor medication adherence inadults with heart failure. J Card Fail. 2011;7(4):340Y348.

39. Hadjuk AM, Lemon SC, McManus DD, et al. Cognitiveimpairment and self-care in heart failure. Clin Epidemiol.2013;5:407Y416.

40. Harkness K, Heckman GA, Akhtar-Danesh N, Demers C,Gunn E, McKelvie RS. Cognitive function and self-caremanagement in older patients with heart failure [publishedonline ahead of print June 3, 2013]. Eur J Cardiovasc Nurs.doi:10.1177/1474515113492603.

41. Lee CS, Gelow JM, Bidwell JT, et al. Blunted responses toheart failure symptoms in adults with mild cognitive dysfunc-tion. J Cardiovasc Nurs. 2013;28(6):534Y540. doi:10.1097/JCN.0b013e31826620fa.

42. Riegel B, Vaughan Dickson V, Goldberg LR, Deatrick JA.Factors associated with the development of expertise in heartfailure self-care. Nurs Res. 2007;56(4):235Y243.

43. Dickson VV, Lee CS, Riegel B. How do cognitive functionand knowledge affect heart failure self-care? J Mixed MethodsRes. 2011;5(2):167Y189.

44. Good PI, Hardin JW. Common Errors in Statistics (andHow to Avoid Them). 3rd ed. Hoboken, NJ: Wiley; 2009.



http://dx.doi.org/10.1136/bmj.b2700

http://dx.doi.org/10.1136/bmj.b2700

The Association Between Mild Cognitive Impairment and Self-care in Adults With Chronic Heart Failure

Documents

Transcript of The Association Between Mild Cognitive Impairment and Self-care in Adults With Chronic Heart Failure