Cope Study

8/14/2019 Cope Study

1/9

ORIGINAL CONTRIBUTION

A Biobehavioral Home-Based Interventionand the Well-being of PatientsWith Dementia and Their CaregiversThe COPE Randomized TrialLaura N. Gitlin, PhD

Laraine Winter, PhD

Marie P. Dennis, PhD, EdM

Nancy Hodgson, PhD, RNWalter W. Hauck, PhD

AMONG THE MORE THAN 5 MIL-lion dementia patients in theUnited States, most live athome, cared for by family

members.1 Functional decline, a coredisease feature, represents a risk fac-tor for poor quality of life, high healthcare costs,institutionalization, and mor-tality.2-4Withdisease progression, fami-lies increasingly provide hands-on

physical assistance with activities ofdaily living (ADLs) and instrumentalADLs (IADLs). This often results inheightened caregiver distress,a riskfac-tor for patient nursing home place-ment.5

Few large randomized trials evalu-ate treatments for supporting physicalfunction of patients with dementia.Trials of antidementia medicationsshow few if any benefits for physicalfunction or caregiver burden and havesubstantial adverse effects.6-8 In 1 study,twice-yearly comprehensive care plan-ning in memory clinics showed no ad-ditional positive effects on functionaldecline.9 Previous nonpharmacologicintervention trials (exercise, use ofpleasant activities, home environmen-tal modifications) had promising find-ings, yet studies reported small effectsizes and outcomes other than func-

Author Affiliations:Jefferson Center for Applied Re-search on Aging and Health, Thomas Jefferson Uni-versity, Philadelphia, Pennsylvania (Drs Gitlin, Win-ter, Dennis,and Hodgson); Sycamore Consulting, NewHope, Pennsylvania (Dr Hauck).

Corresponding Author:Laura N. Gitlin, PhD, Jeffer-son Center for Applied Research on Aging andHealth, Thomas Jefferson University, 130 S NinthSt, Ste 513, Philadelphia, PA 19130 ([email protected]).

Context Optimal treatment to postpone functional decline in patients with demen-tia is not established.

Objective To test a nonpharmacologic intervention realigning environmental de-

mands with patient capabilities.Design, Setting, and Participants Prospective 2-group randomized trial (Care ofPersons with Dementia in their Environments [COPE]) involving patients with demen-tia and family caregivers (community-living dyads) recruited from March 2006 throughJune 2008 in Pennsylvania.

Interventions Up to 12 home or telephone contacts over 4 months by health pro-fessionals who assessedpatient capabilities and deficits; obtainedblood and urine samples;and trained families in home safety, simplifying tasks, and stress reduction. Controlgroup caregivers received 3 telephone calls and educational materials.

Main Outcome Measures Functional dependence, quality of life, frequency of agi-tated behaviors, and engagement for patients and well-being, confidence using ac-tivities, and perceived benefits for caregivers at 4 months.

Results Of 284 dyads screened, 270 (95%) were eligible and 237 (88%) random-ized. Data were collected from 209 dyads (88%) at 4 months and 173 (73%) at 9months. At 4 months, compared with controls, COPE patients had less functional de-pendence (adjusted mean difference, 0.24; 95% CI, 0.03-0.44; P=.02; Cohen d=0.21)and less dependence in instrumental activities of daily living (adjusted mean differ-ence, 0.32; 95% CI, 0.09-0.55; P=.007; Cohen d=0.43), measured by a 15-item scalemodeled after the Functional Independence Measure; COPE patients also hadimproved engagement (adjusted mean difference, 0.12; 95% CI, 0.07-0.22; P =.03;Cohen d=0.26), measured by a 5-item scale. COPE caregivers improved in their well-being (adjusted mean difference in Perceived Change Index, 0.22; 95% CI, 0.08-0.36;P =.002; Cohend=0.30) and confidence using activities (adjusted mean differ-ence, 0.81; 95% CI, 0.30-1.32; P=.002; Cohen d=0.54), measured by a 5-item scale.By 4 months, 64 COPE dyads (62.7%) vs 48 control group dyads (44.9%) eliminated1 or more caregiver-identified problems (21=6.72,P = .01).

Conclusion Among community-living dyads, a nonpharmacologic biobehavioral en-vironmental intervention compared with control resulted in better outcomes for COPE

dyads at 4 months. Although no group differences were observed at 9 months forpatients, COPE caregivers perceived greater benefits.

Trial Registration clinicaltrials.gov Identifier: NCT00259454

JAMA. 2010;304(9):983-991 www.jama.com

2010 American Medical Association. All rights reserved. (Reprinted) JAMA,September 1, 2010Vol 304, No. 9 983

wnloaded From: http://jama.jamanetwork.com/ on 10/08/2013


2/9

tional dependence or required replica-tion.10,11 Recent psychosocial care-giver interventions showed caregiverimprovements, but benefits either didnot extend to patients with dementiaor did not address functional depen-

dence.

12,13

Building on previous nonpharma-cologic approaches and best clinicalpractices,14 we designed the Careof Persons with Dementia in theirEnvironments (COPE) trial to test anonpharmacologic, biobehavioralapproach to support physical func-tion and quality of life for patientswith dementia and the well-being oftheir caregivers. The COPE programtargeted modifiable environmentalstressors to decrease sensorial, physi-cal, and cognitive demands and align

with patient capabilities and alsoruled out underlying medical condi-tions that could lead to reducedpatient functioning. The interventionsought to re-engage patients in dailyactivities and increase functionality,thereby alleviating caregiver burden.

We hypothesized that COPE pa-tients, compared with those in a con-trol group, would show reduced func-tional dependence, improved quality oflife, and enhanced engagement in ac-tivities at 4 months (main study end

point). We also hypothesized thatCOPE caregivers, compared with con-trol caregivers, would report im-proved well-beingand confidence usingactivities at 4 months. Also consid-ered was whether COPE reduced oc-currences of agitated behavior andeliminated problem areas identified bycaregivers.Because the COPE study in-cluded a brief medical screento rule outundiagnosed medical conditions, preva-lence of these conditions are reportedfor intervention patients. Secondarily,

we evaluated long-term effects (at 9months).

METHODS

Study Population

Patients with dementia and familycare-givers were recruited from March 2006to June 2008 through media announce-ments and mailings by social agencies

targeting caregivers. Study proce-dures wereexplained to interested care-givers contacting the research team(telephone, return postcard),and a brieftelephone eligibility screen was admin-istered. Eligible patients had a physi-

cian diagnosis of probable dementia(using criteria from NINCDS/ADRDA[National Institute of Neurological andCommunicative Disorders and Stroke/Alzheimers Disease and Related Dis-orders Association]) or a Mini-MentalState Examination (MMSE)15 score lessthan 24; theyalso were 21years or olderand English speaking, needed helpwithdaily activities or had behavioral symp-toms, and lived with or within 5 milesof family caregivers. Eligible care-givers provided oversight or care for 8or more hours weekly, planned to live

in the area for 9 months, were not seek-ing nursing home placement, and re-ported difficultymanaging patientfunc-tional decline or behaviors.

Exclusioncriteria for dyads were ter-minal illnesses with life expectancy ofless than 9 months, active treatmentsfor cancer, more than 3 acute hospital-izations in the past year, or involve-ment in another caregiver trial. Pa-tients were excluded if they hadschizophrenia or bipolar disorder, haddementia secondary to probable head

trauma, orhad an MMSEscore of0 andwere bed-bound.

Written informed consent was ob-tained from caregivers prior to base-line interviews using forms approvedby the institutional review board. Care-givers provided proxy patient consentand patient assentwas obtained for eachpatient-related assessment usingscriptsapproved by the institutional reviewboard. Families were compensated $20at each interviewfor their participation.

Following baseline interviews, dy-

ads were randomized to the COPE orcontrol group and reassessed by tele-phone at 4 and 9 months by interview-ers masked to participant group. Con-sistent with other trials,13,16 caregiversof patients placed in nursing homesprior to 4 months (n=7) were reas-sessed at 4 months (but not 9 months)in areas amenable to reporting. Care-

givers of patientswho died (n= 21) werenot reassessed at 4 months (n=9) or 9months (n= 12) nor included in analy-ses, as outcome measures were not rel-evant.

Randomization

Dyads were stratified by living arrange-ment (alone/together) and random-ized within each stratum using ran-dom permuted blocks to control forpossible changesin participant mixovertime. The blocking number was devel-oped by the project statistician and un-known to others. Randomization listsand 2 sets of randomization forms wereprepared using opaque envelopes. Theproject director randomized each dyadwithin 48 hours of the baseline inter-view.

Treatment Conditions

The COPE program sought to supportpatient capabilities by reducing envi-ronmental stressors and enhancingcaregiver skills. In this multicompo-nent intervention, all COPE dyads re-ceived exposure to each treatment ele-ment: assessments (patient deficits andcapabilities, medical testing, home en-vironment, caregiver communication,and caregiver-identified concerns);caregiver education (patient capabili-

ties, potential effects of medications,pain, constipation, dehydration); andcaregiver training to address caregiver-identified concerns and help themreduce stress. Training in problem-solving, communication, engaging pa-tients in activities, and simplifying taskswas tailored to address caregiver-identified concerns and patientcapabilities.

COPE dyads received up to 10 ses-sions over 4 months with occupationaltherapists and 1 face-to-facesession and

1 telephone session with an advancepractice nurse. Occupational thera-pists initially interviewed caregivers toidentify patient routines, previous andcurrent roles, habits and interests, andcaregiver concerns. They also con-ducted cognitive and functional testingto identify patient strengths and defi-cits in attention, initiation and perse-

HOME-BASED INTERVENTION FOR PATIENTS WITH DEMENTIA AND THEIR CAREGIVERS

984 JAMA,September 1, 2010Vol 304, No. 9 (Reprinted) 2010 American Medical Association. All rights reserved.



3/9

veration, construction, conceptualiza-tion, and memory.17,18 Occupationaltherapists then trained caregivers tomodify home environments, daily ac-tivities, and communications to sup-port patient capabilities; use problem-

solving to identify solutions forcaregiver-identified concerns; and re-duce stress. For each targeted concern,a written action plan was provided11,13,16

describing treatment goals, patientstrengths, and specific strategies. In ahome visit, the nurse provided care-givers health-related information (paindetection, hydration), obtained patientblood and urine samples, and exam-ined patients for signs of dehydration.Laboratory evaluations included com-plete blood cell count, blood chemis-try, thyroid testing of serum samples,

and culture and sensitivity testing ofurine samples. Patientmedications werereviewed for appropriateness, polyphar-macy, and dosing using publishedguidelines.19 Caregivers were informedof results by telephone and mailedcopies to share with the patientsphysicians.

Dyads assigned to the control groupreceived up to three 20-minute tele-phone calls from trained research staffmembers (not occupational therapistsor nurses). Using scripts, staff asked

caregivers about carechallenges,mailedrelevant informational brochures, andreviewed the materials in subsequentcalls. Materials included tips from theAlzheimers Association and govern-ment agencies on home safetyand man-aging patient behaviors, functional de-cline, and caregiver stress. Thiscontrolledfor professional attention andtailoring of information.

Treatment Implementation

Interventionists for both treatment

groups were independently trained inprotocols through readings, didacticsessions, and practices. For the COPEgroup, treatment fidelity was moni-tored through twice-monthly supervi-sion and audiotapes submitted by in-terventionists, which were reviewed byinvestigators. For the control group,randomly selected telephone calls were

monitored for protocol adherence inreal time. In both groups, intervention-ists completed documentation of du-ration and delivery content for eachcontact, which was reviewed for ad-herence.The COPE interventionistsdid

not have contact with the control groupinterventionists.

Measures

Characteristics of dyads that were as-sessed included living arrangement(alone/together), sex, education, race,age, financial difficulty (1, not very dif-ficult, to 3, very difficult paying for ba-sics like food), and use of 10 formal ser-vices (eg, home health aide). Todescribe the racial background of par-ticipants, caregivers identified them-selves and the patient with dementia as

Caucasian/white, black/African Ameri-can, or other.

Patient Outcomes

For functional dependence, we used a15-item measure modeled after theFunctional Independence Measure,20

previously shown as psychometricallysound and corresponding to objectivedeterminations of dependence and as-sistance required.21,22 Items included 8IADLs (telephone, shopping, mealpreparation, housework, laundry,

travel, medicine, managing finances)and 7 self-care ADLs (bathing, dress-ing upper/lower body, toileting, groom-ing, eating, getting in/out of bed). Foreach item, caregivers chose a score toindicate the following: patients werecompletely independent (a score of 7);there was a safety concern, excessivetime required, or assistive devices used(6); patients needed supervision, set-up, or cueing but no physical help (5);or patients needed physical help (4 fora little help, 25% assistance; 3 for mod-

erate, 50% assistance; 2 for a lot of help,75% assistance; or 1 for complete help,75% assistance). A total mean func-tional dependence score was derived bysumming across items and dividing bynumber of items(actual range of means,1.0-6.3). Lower scores representedgreater dependence (=.92). Sub-scale scores for IADL dependence

(=.81) and ADL dependence (=.93)were similarly derived.

We us ed th e 12 -i te m va li da te dQuality of LifeAlzheimer Diseasescale to assess caregiver perceptions ofpatient quality of life (1, poor, to 4,

excellent).

23

Overall mean responsewas calculated by summing acrossitems and dividing by number ofitems. Higher scores indicated betterquality of life (=.78).

Activity engagement was measuredusing a validated 5-item scale24 (eg,showed signs of pleasure/enjoy-ment), with items rated 1 for never to3 for often.Scores were derived by sum-ming across items and dividing by num-ber of items, with 1 item reverse coded(actual range of means, 1.0-2.8). Higherscores indicated greater engagement

(=.62).We used the 16-item Agitated Be-

havior in Dementia scale to assess agi-tated behaviors in the past month.25 Atbaseline, caregivers indicated whetheragitated behaviors occurred (yes/no)and, if yes, the number of times. Totalnumber of agitated behaviors was de-rived by summingyes items; a mean fre-quency score was derived by sum-ming across items and dividing bynumber of items (actual range, 0.0-121.1). Higher scores indicated greater

number of agitated behaviors and fre-quency.

Caregiver Outcomes

Caregiver well-being (improvement/worsening) was evaluated using the13-item Perceived Change Index,26

fashioned after pharmacologic trialmeasures and shown to have strongpsychometric properties. Caregiversrated change in ability to manage de-mentia, emotional status (anger, dis-tress), and somatic symptoms (en-

ergy, sleep quality) in the past monthusing 5-point scales(1,gotmuch worse,to 5, improved a lot). Total mean scorewas derived by summing across itemsand dividing by number of items.Higher scores indicated greater im-provement (=.86).

Caregiver confidence using activi-ties over the past month was mea-





4/9

sured by 5 investigator-developeditems (identify daily activities patientcan do, involve patient in activities,use activities to distract patient, man-age boredom, set up activities) withratings from 0 for not confident to 10

for very confident.

27

Mean scores werederived across items (actual range ofmeans, 0.60-10.00), with higherscores indicating greater confidence(=.87).

We used a targeted measurement ap-proach employed in medical, pharma-cologic, psychotherapeutic, and behav-ior management trials to capture themost challenging problems (eg, behav-

iors, dependence, respite) for care-givers.28,29 For each identified prob-lem at baseline, caregivers indicated at4 months whether that problem hadbeen eliminated.

At 9 months, we evaluated care-

giver appraisal of study benefits usingan 11-item survey.13,16,29 Items con-cerned satisfaction (yes/no) with par-ticipation (study clearly explained,treated respectfully, effort required, rec-ommend to others); and used ratingsof not at all, some, and a great deal forperceived benefits (overall benefit, de-mentiaunderstanding,confidence man-aging care, enhanced skills, life easier)

and perceived patient benefits (im-proved daily life, helped keep patienthome).

Statistical Analysis

Based on previous research, we based

sample size calculation on assump-tionsof 25.0% attrition by 4 months andstudy hypothesis tested at 90% powerto detect moderateeffect sizes(d=0.45).We used =.05 level test. Given ex-pected attrition, we planned to ran-domize 230 dyads.2 andWilcoxon rank-sum tests were

used to compare intervention and con-trol participants on baseline character-

Table 1.Characteristics of Patients With Dementia and Their Caregivers Who Completed 4-Month Assessment

Characteristic

Control Group

(n = 107)

Intervention Group

(n = 102)

Total

(N = 209) 2

Z

P

ValuePatients with dementia

Age, mean (SD), y 81.8 (9.9) 83.1 (7.8) 82.4 (8.9) 1.00 .33

Sex, No. (%) 3.42 .06

Male 40 (37.4) 26 (25.5) 66 (31.6)

Female 67 (62.6) 76 (74.5) 143 (68.4)

Race, No. (%) 2.18 .34

White 72 (67.3) 75 (73.5) 147 (70.3)

African American 31 (29.0) 26 (25.5) 57 (27.3)

Other 4 (3.7) 1 (1.0) 5 (2.4)

Living arrangement, No. (%) 0.07 .79

Alone 5 (4.7) 4 ( 3.9) 9 ( 4.3)

With caregiver 102 (95.3) 98 (96.1) 200 (95.7)

No. of agitated behaviors, mean (SD) 6.0 (3.0) 6.8 (3.0) 6.4 (3.0) 1.98 .048

MMSE score, mean (SD) 13.6 (7.9) 13.1 (8.2) 13.4 (8.1) 0.51 .61Caregivers

Age, mean (SD), y 62.4 (11.7) 62.0 (12.4) 62.2 (12.0) 0.31 .83

Sex, No. (%) 0.62 .43

Male 10 (9.3) 13 (12.7) 23 (11.0)

Female 97 (90.7) 89 (87.3) 186 (89.0)

Race, No. (%) 5.27 .07

White 71 (66.4) 75 (73.5) 146 (69.9)

African American 31 (29.0) 27 (26.5) 58 (27.8)

Other 5 (4.7) 0 5 (2.4)

Relationship to patient, No. (%) 1.69 .19

Spouse 45 (42.1) 34 (33.3) 79 (37.8)

Nonspouse 62 (57.9) 68 (66.7) 130 (62.2)

Education, No. (%) 7.06 .03

High school 26 (24.3) 38 (37.3) 64 (30.6)Some college 42 (39.3) 24 (23.5) 66 (31.6)

College 39 (36.4) 40 (39.2) 79 (37.8)

Time caregiving, mean (SD), y 3.9 (2.8) 4.0 ( 4.4) 4.0 ( 3.7) 0.57 .84

Financial difficulty, mean (SD) a 2.2 (1.0) 2.2 (1.0) 2.2 (1.0) 0.65 .52

No. of formal services used,mean (SD)

2.39 (1.3) 2.45 (1.27) 2.42 (1.28) 0.38 .70

Abbreviation: MMSE, Mini-Mental State Examination.aValue represents median response category with 2 indicating not very difficult paying for basics like food, housing, medical care (range, 1-3).





5/9

istics and to compare those who stayedin vs those who dropped out by 4months (main end point). These pro-cedures were also used to examine po-tential differences at screening be-tw een el i g i b l e dy ads w i l l i ng to

participate and those not willing.Means, standard deviations, and rangesfor outcome measures were com-puted. The normality assumption foreach dependent measure was tested byexamining the distribution of residuals.

For main treatment effects, the out-come measure was 4-month score withdesign variable, living arrangement(alone/together), and baseline value ofthe outcome measure entered as covar-iates. For the 4-month sample, wefound statistically significant differ-ences between treatment groupsat base-

line for caregiver education and num-ber of agitated behaviors (TABLE 1).Weran additional analyses of covariancewith these variables as covariates. As re-sults did not differ from the primaryanalyses, they are not reported. Co-hendwas calculated to measure effectsize.

Consistent with other trials, to evalu-ate clinical significance for outcomesreaching statistical significance at 4

months, we used the criterion of a0.50-SD improvement from baseline tofollow-up.13 This also represents theup-per endof thedistribution of effect sizesreported in the literature. We calcu-lated number of dyads improving by

0.50 SD or more from baseline to 4months and compared proportions be-tween treatment groups using Mantel-Haenszel2 analyses, controlling for liv-ing arrangement. We also comparedproportion of COPE and control groupcaregivers eliminating 1 or more care-giver-identified problem by 4 monthsusing 2 analysis, controlling for liv-ing arrangement.

To evaluate 9-month effects, inter-vention and control groups were com-pared on adjusted mean differences(baseline to 9 months) for each out-

come using the same procedures as for4-month effects. We also compared in-tervention and control group care-giver perceived benefit at 9 monthsusing Mantel-Haenszel2 analyses, con-trolling for living arrangement.

Statistical analysis was performedwith SPSS version 17.0 (SPSS Inc, Chi-cago, Illinois) with the significance levelset atP.05. All analyses were 2-sided.Analyses included all caregivers ac-

tively caregiving (not bereaved) andproviding 4-month data. Following in-tention-to-treat principles, we in-cluded participants regardless of expo-sure level to treatment.

We adjusted for 6 outcome mea-

sures (functional dependence, activityengagement, quality of life, frequencyof agitated behaviors, and caregiverwell-being and confidence) using amethod controlling for false discoveryrate (ie, proportion of rejected hypoth-eses expected to be erroneous).30 Be-cause .05 significance wasused, we con-trolled the false discoveryrate to be notmore than 5%. Reported numerical Pvalues were not corrected for multipleend points, but impact of adjustmentis noted in TABLE 2.

RESULTSOf 284 screened, 270 dyads (95.1%)were eligible, of whom 237 (87.8%)were willing to participate. No statis-tically significant demographic differ-ences were found between the en-rolled dyads and the 33 dyads eligiblebut unwilling to participate. Study at-trition was low, with 28 dyads (11.8%)lost by 4 months. A higher percentageof male caregivers (n=12, 34.3%)

Table 2. Comparison of Intervention (n = 102) and Control (n = 107) Group Patients and Caregivers at 4 Months a

Mean (SD) Score

AdjustedMean DifferenceBetween Groups

(95% CI)P

ValueCohen

d

Baseline 4-Month Follow-up

Control GroupIntervention

Group Control GroupIntervention

Group

Patient outcomesOverall functional

dependenceb2.8 (1.3) 3.0 (1.2) 3.3 (1.3) 3.7 (1.3) 0.24 (0.03 to 0.44) .02 0.21

IADL dependence 1.8 (1.0) 1.8 (1.0) 2.5 (1.1) 2.8 (1.2) 0.32 (0.09 to 0.55) .007 0.43

ADL dependence 4.1 (1.8) 4.3 (1.7) 4.3 (1.7) 4.6 (1.6) 0.16 (0.09 to 0.42) .21

Activity engagement 2.0 (0.4) 1.9 (0.4) 1.9 (0.5) 2.0 (0.4) 0.12 (0.07 to 0.22) .03 0.26

QOL-AD score 2.1 (0.5) 2.1 (0.4) 2.1 (0.5) 2.2 (0.5) 0.10 (0.00 to 0.20) .06 0.14

ABID score 9.8 (10.7) 11.0 (14.6) 5.5 (8.0) 6.7 (10.6) 0.65 (3.05 to 1.74) .59

Caregiver outcomesPerceived change in

well-being

2.8 (0.5) 2.7 (0.5) 2.9 (0.5) 3.1 (0.6) 0.22 (0.08 to 0.36) .002 0.30

Confidence usingactivitiesc

7.0 (2.2) 6.6 (2.1) 6.9 (2.5) 7.5 (1.9) 0.81 (0.30 to 1.32) .002 0.54

Abbreviations: A BID, Agitated Behaviors in Dementia scale; ADL, activities of daily living; CI, confidence interval; IADL, instrumental activities of daily living; QOL-AD, Quality ofLifeAlzheimer Disease scale.

a Refer to the Methods section for descriptions of the scales used in all outcome measures. All analyses controlled for living arrangement (alone vs with caregiver) and baselinevalue of dependent variable. After adjustment for multiple comparisons by the method of Benjamini and Hochberg, 30 thePvalues for the 6 primary measures (not counting ADLand IADL subscales) were .006 for perceived change in well-being and confidence using activities, .04 for overall functional dependence and activity engagement, .07 for qualityof life, and 0.59 for ABID score.

bThis measure was assessed for 202 patients because 7 patients were placed in nursing homes and the caregivers were not asked functional dependence items at 4 months.cThis measure was assessed for 106 caregivers in the control group because 1 caregiver was unable to respond to items.





6/9

dropped out compared with femalecaregivers (n =16, 7.9%; 21 =19.9;P .001) (FIGURE 1).

By 9 months, an additional 36 dy-ads (17.2% from 4 months) were lostto follow-up. Total study attrition by 9

months was 64 dyads (27.0%). This in-cluded 20 deaths (13 control group pa-tients [65%], 7 intervention patients[35%]) and 10 nursing home place-ments (5 control patients [50%], 5intervention patients [50%]); groupdifferences were not statistically sig-nificant.

For the 4-month sample, patientshada mean (SD) age of 82.4(8.9)years anda mean (SD) MMSE score of 13.4 (8.1).Most were female (n =143, 68.4%) andlived with caregivers (n=200, 95.7%).Caregiversreported managingmanyagi-

tated behaviors (mean [SD], 6.4 [3.0])and high functional dependence(mean[SD],2.9 [1.3]). Most patientswere tak-ing medications: 95 were taking antide-pressants (40.1%); 77, medications tomanage behavioral symptoms (32.5%);108, pain medications (45.6%); and173,memory enhancers (73.3%).

Caregivers were a mean (SD) age of62.2 (12.0) years. Most were female(n= 1 8 6 , 8 9 .0 %), w hi te (n= 1 4 6 ,69.9%), and nonspouses (n =130,62.2%; primarily adult sons anddaugh-ters [n= 115, 88.5%]) (Table 1).

Treatment Implementation

Of 102 COPE dyads, 80 (78.4%) com-pleted 8 to 12 sessions; 3 dyads (2.9%)had fewer than 3 sessions. Overall, dy-ads received a mean (SD) of 9.31 (1.54)face-to-face sessions (mean [SD] length,68.24 [38.34] minutes) and3.25 (0.79)telephone sessions (mean [SD] lengths,20.15 [13.12] minutes for occupa-tional therapists; 6.27 [16.50] min-utes for nurses). Intervention cost wasestimated as $537.05 per dyad based onnational hourly salary or fringerates for

occupational therapists ($42.83) andnurses ($74.41), patient laboratorycosts($120),and themean number andlength of contacts.31 Control group dy-ads received a mean (SD) of 2.83 (0.42)telephone contacts lasting 15 (8.39)minutes as per protocol.

Undiagnosed Medical Conditions

Among 117 COPE patients, nurse as-sessments were obtained for 107 pa-tients (91.4%) and blood or urinesamples for 92 patients (85.9%; 3 re-

fused and samples were unattainablefrom 12). Undiagnosed illnesses oc-curred in 40 patients (37.3%); 3 pa-tients (2.8%) had 2 or more coexist-ing undiagnosed medical illnesses.Conditions included bacteriuria (n= 6;15%), anemia (n=4; 9%), and hyper-glycemia (n=2; 5%). For the 40 pa-tients with undiagnosed medical ill-nesses, 39 caregivers (97.5%) followedup with physicians; 1 refused. Amongthe 39 caregivers following up withphysicians, 1 patient was admitted to

a hospital and 29 patients were outpa-tients.

4-Month Outcomes

Statistically significant improvementswere observed in functional depen-dence for COPE patients (baseline to4 months) compared with controlgroup patients (adjusted mean differ-

Figure 1.Flowchart of Study Design

284 Patients with dementia andfamily caregiver dyads screenedfor eligibility

270 Eligible

237 Randomized

117 Randomized to receive intervention 120 Randomized to control group

102 Included in primary analysis(4-month follow-up)

15 Excluded (did not complete4-month follow-up)

107 Included in primary analysis(4-month follow-up)

13 Excluded (did not complete4-month follow-up)

9-Month follow-up

85 Assessed

22 Discontinued study participation7 Patient died

10 Caregiver refused4 Nursing home placement before

4 months so 9-month assessmentnot offered

1 Nursing home placement andcaregiver refused

9-Month follow-up

88 Assessed


3 Nursing home placement before4 months so 9-month assessmentnot offered

6 Caregiver refused

4-Month follow-up

102 Assessed


6 Caregiver refused1 Caregiver changed

2 Nursing home placement andcaregiver refused

3 Dissatisfied with study

4-Month follow-up

107 Assessed


3 Caregiver refused1 Caregiver changed

3 Dissatisfied with study

14 Excluded7 Nursing home placement planned1 Nursing home placement before

baseline2 MMSE score too high2 No dementia (had stroke)1 Patient died before baseline1 No caregiver distress

33 Excluded (unwilling to participate)

Of the 102 dyads in the intervention group, 3 patients were placed in nursing homes and caregivers receiveda modified 4-month assessment. Of the 107 dyads in the control group, 4 patients were placed in nursinghomes andcaregiversreceived a modified 4-monthassessment. MMSE indicates Mini-MentalStateExamination.





7/9

ence, 0.24; 95% confidence interval[CI], 0.03-0.44;P=.02; Cohend=0.21),representing a small effect. Improve-

ment occurred mostly for IADLs (ad-justed mean difference, 0.32; 95% CI,0.09-0.55; P = .007; Cohen d=0.43), amoderate effect. COPE patients im-proved slightly more in ADL function-ing than controls, but this was not sta-tistically significant(Table 2). Similarly,we observed small but statistically sig-nificant improvements in engagementfor COPE compared with control pa-tients (adjusted mean difference, 0.12;95% CI, 0.07-0.22; P=.03; Cohend=0.26). We did not find statistically

significant benefitsfor frequency of agi-tated behaviors or quality of life.

Compared with control group care-givers, COPE caregivers reported im-provement in well-being (adjustedmean difference, 0.22; 95% CI, 0.08-0.36;P=.002; Cohend=0.30) and en-hanced confidence using activities (ad-justed mean difference, 0.81; 95% CI,0.30-1.32; P=.002; Cohen d=0.54),small to moderate effects (Table 2).TABLE 3 shows proportions of par-

ticipants with clinically meaningful

changes (0.50 SD)for statistically sig-nificant 4-month outcomes. Net im-provement across measures favoredCOPE participants over controls, withdifferences reaching statistical signifi-cance for all except activity engage-ment. Differences in net improve-ments ranged from 14.6% to 26.5%. Of112 caregivers (53.8%) reporting 1 or

more caregiver-identified problemseliminated by 4 months, 64 (62.7%)

were COPE and 48 (44.9%) were con-trol group caregivers (21=6.72, P=.01).

9-Month Outcomes

We did not find statistically signifi-cant differences between COPE andcontrol group participants from base-line to 9 months for any outcome mea-sure. Both intervention and control

group caregivers considered study par-ticipation worthwhile and not timecon-

suming, felt they were treated respect-fully, and would recommend the studyto others (allP .14). However, COPEcompared with control caregivers re-ported a great deal of improvementin their lives overall (70.9% vs 38.5%,22=20.5,P .001), disease understand-ing(66.3% vs 43.6%,22=15.0, P=.001),confidence managing behaviors (72.1%

Figure 2.Perceived Benefits of Intervention and Control Group Caregivers at 9 Months

COPE (n=

88)

Control (n=85)Treated with respect P=.34Study satisfaction

Study clearly explained P=.62

Too much work P=.50

Recommend to others P=.14

Benefited you P


8/9

vs 37.2%,22=25.4, P .001), made lifeeasier (45.3% vs 17.9%, 22 =34.2,P .001), ability to care for patients(54.7% vs 26.9%, 22=25.7, P .001),patients quality of life (32.6% vs 10.3%,22=17.0,P .001), and ability to keep

patients home (39.5 vs 20.8%,2

2=7.5,P=.02) (FIGURE 2).

COMMENT

These findings addto an increasing evi-dentiary base for nonpharmacologicmanagement of patients with demen-tia. We tested a multicomponent inter-vention that helped caregivers attendto patients medical well-being and sim-plify everyday tasks to align with pa-tient capabilities. COPE addresses coreelements of dementia care: optimizingphysical health and function, engag-

ing in daily activities, maintaining qual-ity of life, and supporting caregivers.32

At 4 months, COPE improved patientfunctioning, especially IADLs; patientengagement; and caregiver well-beingand confidence using activities. COPEdidnot improve caregiver ratings of pa-tient quality of life or frequency of agi-tated behaviors, although change wasin the right direction.

Improvement in patient function, al-beit small, compares favorably withpharmacologic trials, yet with no

adverse events or known risks. Al-though different functional measureswere used, trials of dimebon33 andtarenflurbil6 showed no functionalimprovement, and benefits reported ford o n e z e p i l w e r e s m a l l ( C o h e nd0.10)34 compared with COPE (Co-hen d=0.21 for overall function, Co-hend=0.43 for IADL). Other studies ofcholinesterase inhibitors show statis-tically significant but small benefits forIADLs and a trend in ADL improve-ment, as in COPE.7 A multisite study

found no differences in functioningfrom clinic-based treatments.9 In con-trast, COPE decreased severity of over-all dependence by 0.7 points and IADLdependence by 1 point. Control groupcaregivers also reported small func-tionalgainsof 0.5points overall and0.7points for IADLs, although differ-ences were statistically significant fa-

voring intervention. As points on thescale reflect increments of 25%in physi-cal assistance required by caregivers, a1-point reduction may be clinicallymeaningful. Poor patient functioningis a predictor of disease progression,

heightening risk of caregiver burdenand nursing home placement.12 Also,dependencies are associated with in-creased health care costs.3 Thus, evensmall reductions in physical depen-dence may ease caregiver burden.

As to caregiver effects, pharmaco-logic interventions have shown onlysmall benefits in caregiver burden(Cohen d=0.18),8 whereas in thisstudy COPE participants showedhigher effects compared with controls,from Cohen d=0.29 for well-beingan d d =0.54 for confidence using

activities to engage patients. Theseimprovements appear to be clinicallymeaningful. More intervention dyadsimproved 0.50 SD or more than con-trols on outcome measures. Also,more COPE caregivers than controlsreported eliminating at least 1 prob-lem initially identified as challenging.

Consistent with recent studies,16,35 ahigh prevalence (close to 40%) wasfound of undiagnosed, treatable medi-cal conditions for intervention pa-tients with all but 1 dyad (97.5%)

following up with physicians for treat-ment. However, effects of their treat-ment are unclear. A comparison ofCOPE patients with identified andtreated medical problems (n= 39) withCOPE patientswithout identified medi-cal problems or treatment (n=63)showed similar 4-month gains. Never-theless, managing physical health is animportant aspect of dementia care. Highrates of untreated conditions suggest theneed for more frequent routine medi-cal examinations because symptoms

maypresent atypically and patientsmaynot be able to report adequately.At 9 months, there were no statisti-

cally significant differences in out-come measures. Nevertheless, per-ceived benefits favored intervention.Compared with controls, COPE care-givers reported a great deal of im-provement in many areas, including

managing care better and keeping pa-tients home. Lack of findings for stan-dardized measures contrasts with per-ceived benefits, highlighting thecomplexity of measuring improve-ments in quality of life.36

Of importance is that neither groupreported finding the study burden-some, and both groups participantswere equally willing to recommend itto others. Training and telephone edu-cation were equally well received.

Study limitations include an inabil-ity to determine active treatment com-ponents. The trial was not designed toanswer this question and COPE re-flects the integration of multiple com-ponents. COPE may primarily affectcaregiver appraisals. As outcome mea-sures relied on proxy report, it is dif-

ficult to rule out this pathway.Another limitation is study general-

izability. Because caregivers volun-teered for participation, they may havebeen more aware of their role and moremotivated to learn skills than nonvol-unteers.37 Only 15% of study care-givers were male and a higher propor-tion of male caregivers than femalecaregivers dropped out, so it is un-clear how best to address their needs.1

A concern may be the placebo con-dition. Controls received information

tailoredto their needs,12

buttheamountof time staff spent providing informa-tion was not equivalent to that in COPE.Nevertheless, our approach is an ad-vance over previous studies employ-ing no-treatment comparison groups.

Because most patients live at homewith functional decline, a nonpharma-cologic, biopsychosocial-environmen-tal intervention may positively contrib-ute to disease management. Futureresearch needs to examine effects ofunderlying medical conditions, ways

to boost treatment effects, cost-effectiveness, COPE in combinationwith pharmacologic treatments, andtranslational potential.

Author Contributions:Dr Gitlin had full access to allof the data in the study and takes responsibility forthe integrity of the data and the accuracy of the dataanalysis.Study concept and design:Gitlin, Hauck.Acquisition of data:Gitlin, Winter.





9/9

Analysis and interpretation of data:Gitlin, Winter,Dennis, Hodgson, Hauck.Drafting of the manuscript:Gitlin, Winter.Critical revision of the manuscript for important in-tellectual content:Gitlin, Dennis, Hodgson, Hauck.Statistical analysis:Gitlin, Winter, Dennis, Hodgson,Hauck.Obtained funding:Gitlin.Administrative, technical, or material support:Gitlin,

Winter.Study supervision:Gitlin.Financial Disclosures:None reported.Funding/Support:Research reported was supportedin part by funds from the National Institute onAging and the National Institute on NursingResearch (RO1 AG22254) and the PennsylvaniaDe p a rtme n t o f He a lth , T o b a cco S e tt le me n t(SAP100027298).Role of the Sponsor: Funding agencies had no rolein the design and conduct of the study; in the col-lection, analysis, and interpretation of the data; orin the preparation, review, or approval of themanuscript.Additional Contributions:Barry Rovner, MD, Jeffer-son Hospital for Neuroscience, provided patient con-sultation, for which he did not receive additional com-pensation besides his salary. The interventionists whomadeimportant contributionswere Michele Rifkin, MA,OTR/L, Health Through Action; Nicole Davis, MS,OTR/L; Lauren Lapin, OTR/L; Catherine Piersol, MA,OTR/L; Geri Shaw, OTR/L; and Tracey Vause-Earland, MA, OTR/L, and the nurse interventionist,Kathy Czekanski,RN, PhD. These individualswereem-ployees or contractors for Thomas Jefferson Univer-sityand weresupported in part byfundsfrom thelistedgranting agencies. We also acknowledge the contri-butions of our interviewing staff and thank the fami-lies for their study participation.

REFERENCES

1. 2010 Alzheimers disease facts and figures. Alzhei-mers Association. http://www.alz.org/documents_custom/report_alzfactsfigures2010.pdf. Accessed Au-gust 4, 2010.2. AtchisonTB, Massman PJ, Doody RS. Baselinecog-nitivefunctionpredictsrate of decline in basic-careabili-

ties of individuals with dementia of the Alzheimerstype. Arch Clin Neuropsychol. 2007;22(1):99-107.3. HillJ, Fillit H, ThomasSK, Chang S. Functional im-pairment, healthcare costs and the prevalence of in-stitutionalisation in patients with Alzheimers diseaseand other dementias. Pharmacoeconomics. 2006;24(3):265-280.4. Soto ME, Andrieu S, Gillette-Guyonnet S, CantetC, Nourhashemi F, Vellas B. Risk factors for func-tional decline and institutionalisation among commu-nity-dwelling older adults with mild to severe Alzhei-mersdisease:one yearof follow-up.AgeAgeing. 2006;35(3):308-310.5. Spillman BC, Long SK. Does high caregiver stresspredict nursing home entry? Inquiry. 2009;46(2):140-161.6. Green RC, Schneider LS, Amato DA, et al; Taren-flurbil Phase 3 Study Group. Effect of tarenflurbil on

cognitive decline and activities of daily living in pa-

tients with mildAlzheimerdisease: a randomized con-trolled trial.JAMA. 2009;302(23):2557-2564.7. Trinh NH, Hoblyn J, Mohanty S, Yaffe K. Efficacyof cholinesterase inhibitors in the treatment of neu-ropsychiatric symptoms and functional impairmentinAlzheimer disease: a meta-analysis. JAMA. 2003;289(2):210-216.8. Lingler JH, Martire LM, Schulz R. Caregiver-specific outcomes in antidementia clinical drug trials:

a systematic review and meta-analysis.J Am GeriatrSoc. 2005;53(6):983-990.9. Nourhashemi F, Andrieu S, Gillette-Guyonnet S,et al; PLASA Group. Effectiveness of a specific careplan in patients with Alzheimers disease: clusterrandomised trial (PLASA study). BM J. 2010;340:c2466.10. Teri L, Gibbons LE, McCurry SM, et al. Exerciseplus behavioral management in patients with Alzhei-merdisease:a randomized controlledtrial.JAMA. 2003;290(15):2015-2022.11. Gitlin LN, Winter L, Corcoran M, Dennis MP,Schinfeld S, Hauck WW. Effects of the home envi-ronmental skill-building program on the caregiver-care recipient dyad: 6-month outcomes fromthe Phila-delphia REACH Initiative. Gerontologist. 2003;43(4):532-546.12. SrensenS, Duberstein P, Gill D, PinquartM. De-mentia care:mental health effects, intervention strat-egies, and clinical implications. Lancet Neurol. 2006;5(11):961-973.13. Belle SH, Burgio L, Burns R, et al; Resources forEnhancing Alzheimers Caregiver Health (REACH) IIInvestigators. Enhancingthe qualityof lifeof demen-tia caregivers from different ethnic or racial groups: arandomized, controlled trial. Ann Intern Med. 2006;145(10):727-738.14. Ballard CG,GauthierS, Cummings JL,et al.Man-agement of agitation and aggression associated withAlzheimer disease. Nat Rev Neurol. 2009;5(5):245-255.15. Folstein MF, Folstein SE, McHugh PR. Mini-mental state: a practical method for grading thecog-nitive state of patientsfor theclinician.J Psychiatr Res.1975;12(3):189-198.16. GitlinLN, WinterL, DennisMP, Hodgson N, HauckWW. Targeting and managing behavioral symptomsin individuals with dementia: a randomized trial of anonpharmacological intervention[published online July

19, 2010].J Am Geriatr Soc. doi:10.1111/j.1532-5415.2010.02971.x.17. Earhart CA.Allen Diagnostic Modules: Manual.2nd ed. Colchester, CT: S&S Worldwide; 2006.18. MillerJM, PliskinNH. Theclinicalutility of theMat-tis Dementia Rating Scale in assessing cognitive de-cline in Alzheimers disease. Int J Neurosci. 2006;116(5):613-627.19. Fick DM, Cooper JW, Wade WE, Waller JL,MacleanJR, Beers MH.Updatingthe Beers criteria forpotentiallyinappropriate medicationuse in olderadults:results of a US consensus panel of experts. Arch In-tern Med. 2003;163(22):2716-2724.20. Guide for the Uniform Data Set for Medical Re-habilitation (including the FIM instrument), version5.1. Buffalo, NY:StateUniversity of NewYork at Buf-falo; 1997.21. GitlinLN, Roth DL,Burgio LD,et al.Caregiverap-praisals of functional dependence in individuals with

dementiaand associated caregiverupset: psychomet-

ricproperties of a newscaleand response patterns bycaregiver and care recipient characteristics. J AgingHealth. 2005;17(2):148-171.22. Cotter EM, Burgio LD, Roth DL, Gerstle J,Richardson P. Comparison of caregiver and occupa-tional therapist ratings of dementia patients perfor-mance of activities of daily living. J Appl Gerontol.2008;27(2):215-225.23. Logsdon RG, Gibbons LE, McCurry SM, Teri L.

Assessing quality of life in older adults with cognitiveimpairment. Psychosom Med. 2002;64(3):510-519.24. Albert SM,Del Castillo-CastanedaC, Sano M, etal.Quality of life in patients with Alzheimers disease asreported by patient proxies.J Am Geriatr Soc. 1996;44(11):1342-1347.25. Logsdon RG, Teri L, Weiner MF, et al. Assess-ment of agitation in Alzheimers disease: the Agi-tated Behavior in Dementia scale: Alzheimers Dis-ease Cooperative Study. J Am Geriatr Soc. 1999;47(11):1354-1358.26. GitlinLN,WinterL, Dennis MP, Hauck WW.As-sessing perceived change in the well-being of familycaregivers: psychometric properties of the PerceivedChange Index and response patterns. Am J Alzhei-mers Dis Other Demen. 2006;21(5):304-311.27. Gitlin LN, Winter L, Burke J, Chernett N, DennisMP, Hauck WW. Tailored activities to manage neu-ropsychiatric behaviors in persons with dementia andreducecaregiver burden: a randomized pilot study.AmJ Geriatr Psychiatry. 2008;16(3):229-239.28. BattleCC, ImberSD, Hoehn-SaricR, Nash ER,FrankJD. Target complaints as criteria of improvement. AmJ Psychother. 1966;20(1):184-192.29. SchulzR, OBrien A, CzajaS, etal. Dementia care-giver intervention research: in search of clinicalsignificance.Gerontologist. 2002;42(5):589-602.30. Benjamini Y, Hochberg Y. Controllingthe false dis-covery rate: a new and powerful approach to mul-tiple testing. J R Stat Soc Ser B. 1995;57:1289-1300.31. Chartbook: occupational employment and wages,May 2006 (2008, bulletin 2703). US Dept of Labor.http://www.bls.gov/oes/. Accessed June 1, 2010.32. Prince MJ, Acosta D, Castro-Costa E, Jackson J,Shaji KS. Packages of care for dementia in low- andmiddle-income countries. PLoS Med. 2009;6(11):e1000176.

33. Pfizer and Medivation announce results fromtwophase 3 studies in dimebon (latrepirdine*) Alzhei-mers diseaseclinical developmentprogram. Pfizerme-diaroom. March 3, 2010.34. R oman GC, Wilkinson DG, Doody RS, Black SE,Salloway SP, Schindler RJ. Donepezil in vascular de-mentia: combined analysis of two large-scale clinicaltrials.Dement Geriatr Cogn Disord. 2005;20(6):338-344.35. Lppnen MK, Isoaho RE, Rih IJ, et al. Undi-agnosed diseases in patients withdementia:a poten-tial targetgroup for intervention.Dement Geriatr CognDisord. 2004;18(3-4):321-329.36. Rabins PV, Black BS. Measuring quality of life indementia: purposes, goals, challenges and progress.Int Psychogeriatr. 2007;19(3):401-407.37. Pruchno RA, Brill JE, Shands Y, et al. Conve-nience samples and caregiving research: how gener-alizable are the findings? Gerontologist. 2008;

48(6):820-827.



Documents

Cope Study