How Different are Quality of Life Ratings for People with Dementia Reported by Their Family Caregivers from Those Reported by the Patients Themselves?

Abstract

Background: Measurements of patient quality of life (QoL) play a major role in the management of dementia.

Objective: We investigated the self-proxy discrepancy of QoL ratings in the elderly and the impact of dementia severity on the discrepancy.

Methods: QoL of 718 patients with dementia and 651 non-demented elderly were rated by themselves and their caregivers (CG) using the Quality of Life-Alzheimer’s Disease (QoL-AD). The impact of the rater on the total and item scores of QoL-AD was analyzed using repeated measures ANOVA and differential response patterns between self and proxy were analyzed using differential item functioning (DIF) analysis.

Results: Self-rated scores were higher than CG-rated scores in all diagnostic groups. The interaction between rater and diagnostic group was significant in total QoL-AD score and 5 item scores (‘memory’, ‘marriage’, ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’). The strength of the DIF increased with advancing dementia in these items.

Conclusion: Self-proxy rating discrepancy of QoL was influenced by the presence and severity of dementia only in five items.

Keywords

Caregiver dementia discrepancy proxy quality of life self

INTRODUCTION

Maximizing the quality of life (QoL) of persons with dementia (PWD) is a major goal in the care of PWD and in designing interventions for dementia [1], as QoL is heavily deteriorated in PWD and tends to worsen with the progress of dementia [2]. Typically QoL ratings are based on self-report, as this approach emphasizes the subjective nature of the person’s experience [2, 3]. However, it is difficult to reliably measure QoL in PWD due to the deterioration of the patient’s cognitive ability. With advancing severity of dementia, self-reported QoL becomes unreliable in PWD [4 –7]. To overcome this limitation, proxy reports on QoL of PWD from caregivers (CG) are commonly employed.

However, proxy-rated QoL was shown to be considerably discrepant from self-rated QoL in PWD [8 –11]. PWD usually gave higher QoL ratings than their CG [9], and the discrepancy between self- and proxy-rated QoL increased with the severity of dementia [12]. Although some previous studies attributed this severity-dependent discrepancy in the ratings of QoL between PWD and their CG to the progression of anosognosia and/or depression in PWD [12], such discrepancy may not solely be attributed to the changes in PWD, but also be due to factors affecting their CG. QoL is an umbrella concept that includes various components (social, psychological, and physical) and their interactions [13]. It may just be natural for ratings of QoL to vary to some degree with the rater’s point of view. In addition, psychiatric or medical conditions of CG may also influence the ratings of QoL of PWD reported by CG. However, there have only been a few studies investigating sources of self-proxy discrepancy in ratings of QoL in PWD. Furthermore, the detailed nature of self-proxy discrepancy in the rating of QoL of PWD has not been thoroughly studied.

In this study, we investigated the self-proxy discrepancy in QoL ratings to differentiate the QoL items that are differently rated due to progressive deterioration of dementia patients from those due to the characteristics of their CG.

METHODS

Design and study population

This study was conducted as a part of the Korean Longitudinal Study on Cognitive Aging and Dementia (KLoSCAD) [14]. The KLoSCAD is a nationwide community-based prospective cohort study on randomly sampled community-dwelling elderly Koreans aged 60 or over. The baseline evaluation of the cohort was conducted from November 2010 through September 2012 and follow-up evaluations will be conducted every two years until 2018. Among the 8,539 participants in the baseline evaluation, 1,369 pairs of elderly care recipient individuals (CR) and their caregivers (CG) completed both self- and caregiver-rated Quality of Life-Alzheimer’s Disease (QoL-AD) assessments. The CG was defined as the person who had an ongoing responsibility for helping the CR with activities of daily living [15]. All CG were relatives of the CR, were not paid for their care, and saw them in person 5 days or more per week.

Evaluation and diagnosis

Geriatric psychiatrists with expertise in dementia research administered a face-to-face standardized diagnostic interview, and physical and neurological examinations of each CR were conducted using the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet (CERAD-K) Clinical Assessment Battery (CERAD-K-C) [16]. A research neuropsychologist or trained research nurse administered the CERAD-K Neuropsychological Assessment Battery (CERAD-K-N) [16, 17], the Digit Span Test (DST) [18], and the Frontal Assessment Battery [19] to each CR. In addition, we also evaluated global cognition using the Mini-Mental Status Examination (MMSE) [20], the severity of subjective memory complaints using the Subjective Memory Complaints Questionnaire (SMCQ) [21], depression using the Geriatric Depression Scale Short Form (SGDS), behavioral and psychological symptoms using the Neuropsychiatric Inventory (NPI) [22], and activities of daily living using the Disability Assessment for Dementia (DAD) scale [23].

Computed tomography or magnetic resonance imaging (T1-weighted, T2-weighted, FLAIR) of the brain; laboratory tests, including complete blood cell count, chemistry profile, and a serological test for syphilis; echocardiography; and chest X-ray were conducted for subjects with a diagnosis of dementia to determine the subtype.

We diagnosed dementia according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) diagnostic criteria for dementia [24]. We determined the severity of dementia using the Clinical Dementia Rating (CDR) [25], with ratings of 0.5 as very mild, 1 as mild, and 2 or higher as moderate to severe. We classified the CR into four diagnostic groups: cognitively normal elderly (CNE), patients with very mild dementia (DEM-VM), patients with mild dementia (DEM-M), and patients with moderate to severe dementia(DEM-MS).

Assessment of QoL

We assessed QoL using the QoL-AD [7]. It is a brief 13-item measure designed to obtain both self- and proxy-rated QoL ratings for people with AD. Each item can be rated from 1 (poor) to 4 (excellent). The global score thus ranges from 13 (very poor QoL) to 52 (very good QoL). In this study, proxy-rated QoL was scored by a CG of each CR.

Statistical analyses

A descriptive analysis was carried out of the clinical and socio-demographic characteristics of the sample, using absolute and relative frequencies for qualitative variables and measures of central trend and dispersion for quantitative variables. We analyzed the correlation between dementia severity and both self- and CG-rated QoL-AD ratings using a Spearman test. A repeated measures analysis of variance (ANOVA) was then performed on QoL-AD, with the rater (either CR or CG) as the within-subject factor and diagnostic group, age, sex, education, living area, patient/caregiver relationship, and MMSE, NPI, DAD, SMCQ, and Geriatric Depression Scale (GDS) scores as between-subject factors. We also performed a Tukey post-hoc comparison for each item in each dementia group. Differences in the magnitude of differential responses to QoL-AD items between CR and CG were estimated using standard mean differences obtained from a hybrid logistic regression/differential item functioning (DIF) analysis [26, 27]. The proportional-odds logistic regression model was performed with the assumption that the outcome variable is ordinal. Based on the proportional odds assumption or the parallel regression assumption, a single set of regression coefficients was estimated for all cumulative logits with varying intercepts (αk). For each item, an intercept-only (null) model and three nested models were formed in hierarchy with additional explanatory variables. p-values were considered significant when they were lower than 0.05. All analyses were performed on R version 3.1.0 (The R Foundation) [28].

RESULTS

Characteristics of the subjects

As shown in Table 1, 1,369 pairs of CR and their CG were included in this study. 651, 256, 336, and 126 of the CR classified as CNE, DEM-VM, DEM-M, and DEM-MS, respectively. 41% of the CR were men, 75% lived in urban areas, and 4.7% were economically disadvantaged. The subjects’ age increased and their educational levels decreased significantly with increases in the severity of dementia (p < 0.001). Furthermore, MMSE and DAD scores decreased and their NPI, SMCQ, and GDS scores increased with increases in the severity of dementia (p < 0.001). Among the subjects’ CG, 54% were the spouse, 17% were the son, 21% were the daughter, and 8% were the daughter-in-law. The mean age of the CG was 58.8 years (SD = 13.4 years) and was not different between diagnostic groups (p = 0.363). The mean educational level of the CG was 12.3 years (SD = 4.5 years) and was slightly higher in the CG of the CR with DEM-VM (p < 0.001).

Self- and CG-rated QoL

As shown in Table 2, both self- and CG-rated total QoL-AD scores were negatively associated with the severity of dementia (r = –0.44, p < 0.0001 for self-rating; r = –0.55, p < 0.0001 for CG-rating). However, the CR reported higher QoL scores than their CG (mean discrepancy = 1.9, SD = 5.8). In the repeated-measures ANOVA (η² = 0.496), both self- and CG-rated total QoL-AD scores decreased with dementia severity (F = 371.537, p < 0.001), but self-rated scores were higher than CG-rated scores (F = 146.098, p < 0.001) in all diagnostic groups. The interaction between the rater (CR versus CG) and the diagnostic group (CNE, DEM-VM, DEM-M, and DEM-MS) was significant (F = 4.884, p < 0.005), indicating that the discrepancy between self- and CG-rated total QoL-AD scores may differ according to the presence and severity of dementia. The differences between self-rated and CG-rated total QoL-AD scores were 1.2, 1.6, 3.0, and 2.7 in the CNE, DEM-VM, DEM-M, and DEM-MS groups, respectively. The score differences in the DEM-VM group were comparable to those in the CNE group (p = 0.31), but score differences those in the DEM-M and DEM-MS groups were significantly higher than those in the CNE group (p < 0.001 for DEM-M and p < 0.05 for DEM-MS).

As a whole, compared to their CG, the CR gave higher ratings in the 13 items, except for “memory” in the CNE group. For 5 out of 13 items (‘memory’, ‘marriage’, ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’), the interaction between the rater and the diagnostic group was significant (Table 2 and Fig. 1). The CR-CG difference in the rating of ‘memory’ was significant beginning at the very mild stage of dementia, but those of ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’ were significant beginning at the moderate stage of dementia. In addition, ratings of ‘do things for fun’ and ‘life as a whole’ increase until the mild stage and then decrease thereafter. The CR-CG difference in the rating of ‘marriage’ was not significantly different based on the severity of dementia in the posthoc comparison. Figure 2 shows the strengths of DIF between CR and CG in these five item scores. The strength of DIF steadily increased with advancing dementia in the ‘memory’ and ‘chores around the house’ items, but peaked at the mild stage and then decreased thereafter in the ‘do things for fun’ and ‘life as a whole’ items. The strength of DIF in the ‘marriage’ item steadily increased with advancing dementia, but was already high in the CNE group (0.352). The self-proxy discrepancies in the other items were not influenced by the presence or severity of dementia. The ratings for seven items (‘physical health’, ‘energy’, ‘mood’, ‘situation’, ‘family’, ‘friends’, and ‘self as a whole’) were also significantly different between CR and CG in the absence of dementia.

DISCUSSION

In this study, self-rated scores were higher than CG-rated scores in all diagnostic groups (F = 146.098, p < 0.001). The total QoL-AD score (F = 4.884, p < 0.005) and scores for 5 out of 13 QoL-AD items (‘memory’, ‘marriage’, ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’) became more discrepant between CR and CG with the presence and severity of dementia in CR. CR without dementia consistently gave higher ratings on 7 items (‘physical health’, ‘energy’, ‘mood’, ‘situation’, ‘family’, ‘friends’, and ‘self as a whole’) than their CG. According to Whitehouse and Rabins [29], QoL should not be isolated from the many measurements of the benefits of care, but rather the central goal of our professional activity in care programs for PWD. QoL of patients suffering from chronic conditions like dementia is of particular importance for caregivers and healthcare providers, as well as for patients themselves [30]. However, measurement of QoL remains difficult in patients with dementia, as QoL is a complex concept that corresponds to the perceived quality of an individual’s daily life and includes numerous components (i.e., psychological, physical, and environmental) and their interactions. Recently, Conde-Sala and colleagues demonstrated that anosognosia was common in PWD and was associated with the increase in self-rated QoL scores in PWD. This suggests that this rating is biased and in need of careful consideration before it is used to draw clinical conclusions [15]. Proxy rating may thus provide a second estimation of QoL of PWD for health professionals in their management of PWD. The QoL-AD has become one of the most popular instruments for evaluating QoL of PWD, as it allows for both self and proxy ratings [2].

However, proxy ratings of QoL of PWD may also have limitations. For instance, the characteristics of CG may influence their point of view regarding the QoL of PWD. As QoL is a highly subjective concept, CG may focus on different aspects of QoL than the PWD. There have been numerous studies on the effect of CG characteristics on the rating of the QoL of their relative with dementia [31 –35]. In 2011, a Canadian analysis estimated that CG depression and burden were negatively associated with the rating of patient QoL [32]. Two years later, another study showed that CG age, money, and valuation of life were significant predictors of potential biases in the proxy assessment of patient QoL [34]. More recently, in 2014, Gräske and colleagues corroborated these findings as they discovered in a dementia care that nurse burnout and satisfaction with life as well as the circumstances of the ratings had a significant impact on the rating of the QoL of dementia people [35]. That being said, CR-CG discrepancy in the ratings of QoL of PWD may not be only due to cognitive deterioration of CR with dementia or burden in CG, but to differences in the viewpoints of CG and CR regarding QoL. Indeed, we found that the CG of CR without dementia also rated 12 out of 13 items lower than their CR. This discrepancy was particularly high for the ratings on ‘family’, ‘energy’, and ‘marriage’. In fact, such discrepancy between self- and proxy-ratings of QoL has already been reported in the elderly without dementia. A review by Von Essen indicates that CG usually report more problems and lower levels of function for elderly CR than the CR themselves. Interestingly, this discrepancy was more pronounced for ratings of emotional functions, suggesting that the interpretation of proxy ratings for these items may be a delicatematter [36].

It is also difficult to decide which rating should be relied on when the ratings by CG and CR are discrepant. For items such as ‘family’ or ‘energy’, in which the CR-CG rating discrepancy significant, but not influenced by the presence or severity of dementia, the CR’s ratings may be preferred, considering the subjective nature of QoL. On the contrary, for items in which the CR-CG rating discrepancy was higher in the CR with dementia and increased with advancing dementia of the CR, the CG’s ratings may be preferred as an alternative. There have been a number of studies on the discrepancy between QoL ratings by CG and those by CR with dementia. Buckley et al. reported that self- and proxy-reported QoL ratings were only modestly correlated [3]. Such self-proxy discrepancy may be attributed to various factors. Arons and colleagues discovered that CG project part of their own QoL onto patients when assessing patient QoL [34]. More recently, Dourado et al. reported that the patients’ mood and awareness of the disease were associated with changes in the ratings provided by the CG [37]. Conde-Sala et al. also found that the discrepancy between CR and CG regarding ratings of the CR’s QoL increased with the severity of dementia [12]. In their study, patients with a better cognitive status, more depression, and less anosognosia gave the lowest ratings, whereas those with a poorer cognitive status, less depression, and greater anosognosia gave the highest ratings of their QoL. Sneeuw et al. suggested a U-shaped relationship between the patient’s health condition and the level of self-proxy agreement. That is, large discrepancies tend to occur most frequently for patients with a moderately impaired health condition and less frequently for patients with either a very good or very poor health status. For patients with either a very good or very poor health status, the answers to many QoL questions will be evident, i.e. at either the top or the bottom end of the scale [38].

In the current study, we found that the CR-CG discrepancy on four items of QoL-AD (‘memory’, ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’) was negligible in the CR without dementia (i.e., CNE), but was significant in the CR with dementia and increased with increasing severity of dementia. The discrepancy between ratings by CR and CG on these items was much higher in the CR with moderate or severe dementia than in the CR with normal cognition. In contrast to the ratings of ‘chores around the house’, ‘do things for fun’, and ‘life as a whole’, those of the ‘memory’ item were also considerably discrepant between CR with very mild dementia and their CG. Even in the very mild stage, PWD may underrate their memory impairment possibly due to either anosognosia or denial of disease. Two other items (‘chores around the house’ and ‘do things for fun’) were associated with activities of daily living. Therefore, PWD may underrate their impairment if they are still engaging in such activities, even if their performance is impaired. However, their CG may be more likely to rate these items not based on the CR’s participation, but based on their performance. The rating of the item ‘life as a whole’ may reflect the evaluation of the overall impact of impairments in the other 12 items on the life of CR. The CR with dementia and their CG may rate this item based on different aspects of life. PWD may feel good by just participating in activities and/or by interacting warmly with others, even if their levels of performance or interaction are not as good as they were before. Again, the CG may judge the life of PWD not by how they feel but by how well they perform. Sometimes, CG may confuse the evaluation of QoL with that of function.

Our study has several limitations. First, this study was cross-sectional. Our results thus warrant further longitudinal observational studies. Second, anosognosia and denial of disease were not evaluated. Third, medical and/or psychiatric disorders of CG that may also influence the self-proxy rating discrepancy were not evaluated. Fourth, all subjects were Korean, and thus our observation may not be applicable to populations with different cultural backgrounds. Nonetheless, we characterized the nature of CR-CG discrepancy in the rating of QoL of CR with dementia, which may be a good reference for health professionals in understanding the results of QoL ratings in PWD.

In conclusion, self-proxy QoL rating discrepancy was evident in both dementia patients and normal elderly individuals. Furthermore, the self-proxy rating discrepancy increased with the presence and severity of dementia in only 5 out of 13 items. Therefore, we should not rashly attribute such discrepancy to PWD. Even in the evaluation of QoL of people with moderate to severe dementia, we should be careful in considering proxy-ratings instead of self-ratings.

Footnotes

ACKNOWLEDGMENTS

This work was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (grant number HI09C1379) (A092077).

Authors’ disclosures available online ().

References

Weyerer

, Schäufele

(2003) The assessment of quality of life in dementia. Int Psychogeriatr 15, 213–218.

Logsdon

, Gibbons

, McCurry

, Teri

(1999) Quality of life in Alzheimer’s disease: Patient and caregiver reports. J Ment Health Aging 5, 21–32.

Buckley

, Fauth

, Morrison

, Tschanz

, Rabins

, Piercy

, Norton

, Lyketsos

(2012) Predictors of quality of life ratings for persons with dementia simultaneously reported by patients and their caregivers: The Cache County (Utah) Study. Int Psychogeriatr 24, 1094–1102.

Sevush

, Leve

(1993) Denial of memory deficit in Alzheimer’s disease. Am J Psychiatry 150, 748–751.

Lopez

, Becker

, Somsak

, Dew

, DeKosky

(1994) Awareness of cognitive deficits and anosognosia in probable Alzheimer’s disease. Eur Neurol 34, 277–282.

Albert

, Del Castillo-Castaneda

, Sano

, Jacobs

, Marder

, Bell

, Bylsma

, Lafleche

, Brandt

, Albert

, Stern

(1996) Quality of life in patients with Alzheimer’s disease as reported by patient proxies. J Am Geriatr Soc 44, 1342–1347.

Logsdon

, Gibbons

, McCurry

, Teri

(2002) Assessing quality of life in older adults with cognitive impairment. Psychosom Med 64, 510–519.

Conde-Sala

, Garre-Olmo

, Turró-Garriga

, López-Pousa

, Vilalta-Franch

(2009) Factors related to perceived quality of life in patients with Alzheimer’s disease: The patient’s perception compared with that of caregivers. Int J Geriatr Psychiatry 24, 585–594.

Conde-Sala

, Turró-Garriga

, Garre-Olmo

, Vilalta-Franch

, Lopez-Pousa

(2014) Discrepancies regarding the quality of life of patients with Alzheimer’s disease: A three-year longitudinal study. J Alzheimers Dis 39, 511–525.

10.

Tay

, Chua

, Chan

, Lim

, Ang

, Koh

, Chong

(2014) Differential perceptions of quality of life (QoL) in community-dwelling persons with mild-to-moderate dementia. Int Psychogeriatr 26, 1273–1282.

11.

Huang

H-L

, Weng

L-C

, Tsai

Y-H

, Chiu

Y-CY

, Chen

K-H

, Huang

C-C

, Tang

JS-H

, Wang

W-S

(2015) Predictors of self- and caregiver-rated quality of life for people with dementia living in the community and in nursing homes in northern Taiwan. Int Psychogeriatr 27, 825–836.

12.

Conde-Sala

, Reñé-Ramírez

, Turró-Garriga

, Gascón-Bayarri

, Campdelacreu-Fumadó

, Juncadella-Puig

, Rico-Pons

, Garre-Olmo

(2014) Severity of dementia, anosognosia, and depression in relation to the quality of life of patients with Alzheimer disease: Discrepancies between patients and caregivers. Am J Geriatr Psychiatry 22, 138–147.

13.

Felce

, Perry

(1995) Quality of life: Its definition and measurement. Res Dev Disabil 16, 51–74.

14.

Kim

, Park

, Lee

, Jhoo

, Kim

B-J

, Kim

J-L

, Kim

, Youn

, Ryu

S-H

, Lee

, Kwak

, Lee

, Moon

, Cha

, Han

, So

, Jeong

H-G

, Kim

(2013) Overview of the Korean longitudinal study on cognitive aging and dementia. Alzheimers Dement 9, P626–P627.

15.

Conde-Sala

, Turró-Garriga

, Piñán-Hernández

, Portellano-Ortiz

, Viñas-Diez

, Gascón-Bayarri

, Reñé-Ramírez

(2016) Effects of anosognosia and neuropsychiatric symptoms on the quality of life of patients with Alzheimer’s disease: A 24-month follow-up study. Int J Geriatr Psychiatry 31, 109–119.

16.

Lee

, Lee

, Kim

, Jhoo

, Kim

, Yoon

, Woo

, Ha

, Woo

(2004) A normative study of the CERAD neuropsychological assessment battery in the Korean elderly. J Int Neuropsychol Soc 10, 72–81.

17.

Lee

, Lee

, Kim

, Jhoo

, Kim

, Lee

, Kim

, Han

, Woo

(2002) Development of the Korean version of the Consortium to Establish a Registry for Alzheimer’s Disease Assessment Packet (CERAD-K): Clinical and neuropsychological assessment batteries. J Gerontol B Psychol Sci Soc Sci 57, P47–P53.

18.

Blackburn

, Benton

(1957) Revised administration and scoring of the Digit Span Test. J Consult Psychol 21, 139–143.

19.

Kim

, Huh

, Choe

, Jeong

, Park

, Lee

, Jhoo

, Lee

, Woo

, Kim

(2010) Korean version of frontal assessment battery: Psychometric properties and normative data. Dement Geriatr Cogn Disord 29, 363–370.

20.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

21.

Youn

, Kim

, Lee

, Jhoo

, Lee

, Park

, Choi

, Choe

, Jeong

, Choo

, Woo

(2009) Development of the subjective memory complaints questionnaire. Dement Geriatr Cogn Disord 27, 310–317.

22.

Yesavage

, Brink

, Rose

, Lum

, Huang

, Adey

, Leirer

(1982) Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res 17, 37–49.

23.

Gélinas

, Gauthier

, McIntyre

, Gauthier

(1999) Development of a functional measure for persons with Alzheimer’s disease: The disability assessment for dementia. Am J Occup Ther 53, 471–481.

24.

American

Psychiatric Association

(2000) Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, American Psychiatric Association, Washington, DC.

25.

Morris

(1993) The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414.

26.

Choi

, Gibbons

, Crane

(2011) Lordif: An R package for detecting differential item functioning using iterative hybrid ordinal logistic regression/item response theory and monte carlo simulations. J Stat Softw 39, 1–30.

27.

Fletcher

(2009) Detecting Differential Item Functioning (DIF) in the Diabetes Risk Perception Survey (RPS-DM), ProQuest.

28.

R Core Team (2014) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/.

29.

Whitehouse

, Rabins

(1992) Quality of life and dementia. Alzheimer Dis Assoc Disord 6, 135–137.

30.

Ready

(2002) Quality of life in dementia. Med Health R I 85, 216–218.

31.

Karlawish

, Casarett

, Klocinski

, Clark

(2001) The relationship between caregivers’ global ratings of Alzheimer’s disease patients’ quality of life, disease severity, and the caregiving experience. J Am Geriatr Soc 49, 1066–1070.

32.

Naglie

, Hogan

, Krahn

, Black

, Beattie

, Patterson

, Macknight

, Freedman

, Borrie

, Byszewski

, Bergman

, Streiner

, Irvine

, Ritvo

, Comrie

, Kowgier

, Tomlinson

(2011) Predictors of family caregiver ratings of patient quality of life in Alzheimer disease: Cross-sectional results from the Canadian Alzheimer’s disease Quality of Life Study. Am J Geriatr Psychiatry 19, 891–901.

33.

Heggie

, Morgan

, Crossley

, Kirk

, Wong

, Karunanayake

, Beever

(2012) Quality of life in early dementia: Comparison of rural patient and caregiver ratings at baseline and one year. Dementia (London) 11, 521–541.

34.

Arons

, Krabbe

, Schölzel-Dorenbos

, van der Wilt

, Rikkert

MGO

(2013) Quality of life in dementia: A study on proxy bias. BMC Med Res Methodol 13, 110.

35.

Gräske

, Meyer

, Wolf-Ostermann

(2014) Quality of life ratings in dementia care – a cross-sectional study to identify factors associated with proxy-ratings. Health Qual Life Outcomes 12, 177.

36.

von

Essen L

(2004) Proxy ratings of patient quality of life–factors related to patient-proxy agreement. Acta Oncol 43, 229–234.

37.

Dourado

, Sousa

MF de

, Santos

, Simões Neto

, Nogueira

, Belfort

, Torres

, Dias

, Laks

(2016) Quality of life in mild dementia: Patterns of change in self and caregiver ratings over time. Rev Bras Psiquiatr. doi: 10.1590/1516-4446-2014-1642.

38.

Sneeuw

, Aaronson

, Sprangers

, Detmar

, Wever

, Schornagel

(1998) Comparison of patient and proxy EORTC QLQ-C30 ratings in assessing the quality of life of cancer patients. J Clin Epidemiol 51, 617–631.