Analyzing the Costs of Informal Care for Persons with Dementia in Spain

Abstract

The objective of this article is to estimate the time and associated costs of informal caregiving for the elderly with different levels of dementia. In a cross-sectional observational study with 242 subjects, we use the Resident Assessment Instrument Home Care (RAI-HC) to compile information on socio-demographic variables, informal care, comorbidities, hearing and vision function, use of formal support services, use of locomotion aids, and dementia. We construct a multivariable regression model to determine the cost of informal care due dementia. Findings show a positive association of dementia severity and costs of informal care.

KEYWORDS:

• care services
• dementia
• informal care cost
• LTC management

INTRODUCTION

Health sector is one of the most important in The Organization for Economic Co-operation and Development (OECD) countries. In 2010 European Union members' states devoted an average of 9% of their gross domestic product (GDP) on health expending (OCDE, 2012). On the other hand, long term care (LTC) expenditure accounts for 1.5% of GDP in the OECD (Colombo, F., Llena-Nozal, A., Mercier, J., Tjadens, F., 2011), although, relatively, is a small sector in the economy is expected to grow in the next years. First, because population aging is a generalized process worldwide and elderly represent the largest percentage of dependent people; therefore, demand for LTC services will increase. Second, up to now, care has been provided by informal care, especially women, but female participation in the formal labor market is increasing and in the feature, the availability of informal caregivers will decrease and the need for paid care, will increase.

In this context, dementia deserves special attention because is a chronic degenerative condition that affects mental functions such as memory, orientation, language, visual recognition, and conducts that lead to dependence for activities of daily living (Carpenter, Hastie, Morris, Fries, & Ankri, 2006). Furthermore, several studies show that the prevalence of dementia increases exponentially with age, almost doubling every 5 years and more rapidly in women (Comas-Herrera, Wittenberg, Pickard, & Knapp, 2007). The total societal worldwide costs of dementia are estimated to be $422 billion in 2009 and per case is estimated at $12,200 (Wimo, Winblad, & Jönsson, 2010). In addition, one of the most important direct non-health care costs of dementia is informal care. Nordberg, von Strauss, Kareholt, Johansson, and Wimo (2005) indicate that the cost of informal care for patients with dementia outweighs the cost of health care, although Wimo, Winblad, and Jönsson (2007) do not confirm this finding. In any case, the cost of informal care is a large percentage of the total cost of care of persons with dementia (Ernst & Hay, 1997).

Nowadays, there is an important structural change as states reduce their expenditure, and managers and policy makers of health and LTC sector have to find ways of coping with the growth of the elderly people with dementia population. However, public organizations have difficulty adapting and much of the literature has highlighted the importance of establishing a process for change and creating a consensus around this need (Barton Cunningham & Kempling, 2009). Additionally, it is necessary to consider that to help maintain and improve their performance, organizations need to introduce innovations in the organization's production or operating systems and managerial process (Damanpour, Walker, & Avellaneda, 2006). In that sense, it is suitable to know how to improve the management of related services of long term care. To do this, the first area of research is determining the financial cost of providing for the care of people with dementia.

To our knowledge, this cost has not been estimated in Spain. Internationally, there are studies related to management, health, and care, which do not analyze costs (Drake & Davies, 2006; Mital, 2010; Cepeda-Carrión, Cegarra-Navarro, & Leal-Millán, 2012; Sheng, Chang, Teo, & Lin, 2013). Although management literature is wide and varied (Hatchuel, 2009), we can consider some tasks required by managers (Nienaber, 2010): planning, organizing, command, coordination, and control. To perform these tasks successfully, cost information is relevant.

This article estimates the time and associated costs of informal caregiving for the elderly with different levels of dementia by using a Spanish population. This information is economically and socially useful to health care managers and government officials as the population aging and dependency ratio are increasing. To overcome this challenged, innovation in health and LTC is necessary (Thakur, Hsu, & Fontenot, 2011). For our purpose, the factors associated with more or fewer hours of informal care were identified and used as control variables in a multiple regression analysis to isolate the time devoted to dementia care and to determine the associated costs. These issues are carried out in the Method section. The third section presents the principal results and the final sections provide discussion and conclusions.

METHOD

Conceptual Framework of Informal Caregiving for Dementia

The growth in the number of people with dementia implies an increase of persons who depend on others to carry out daily activities of personal care and household tasks.

The costs of care for patients with dementia are important to society, but it is not easy to identify and estimate the costs of informal care for a given disease. For example, factors such as age, gender, the availability of informal caregivers, and/or the possibility of using other care options, such as professional services, can also increase the time devoted to informal care. To establish the financial cost amount of informal care for disabled people, it is necessary to determine the amount of time provided to this population of dependent people.

In general, studies as Boaz and Hu (1997), Langa et al. (2001), Jönsson (2004), or Wimo et al. (2007, 2010) consider that needs for help are related to the number of Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) disabilities and cognitive impairment. Cognitive impairment results in limitation in the independence performance of ADL and IADL. ADLs are listed in the sequence in which a person is likely to lose the ability to perform them without help: bathing, dressing, getting around the house, getting to and using the toilet, continence, getting in and out of bed, and eating. IADLs are activities that enable individuals to live in the community: housework, laundry, meal preparation, shopping for groceries, getting around outside the house, managing money, using the phone, and taking oral medicine (Boaz & Hu, 1997). Additionally, elderly individuals with dementia may also have comorbid conditions.

To determine people with dementia, authors have used national surveys such as Boaz and Hu (1997), Ernst and Hay (1994, 1997); Nordberg et al. (2005) or international surveys such as Wimo et al. (2007, 2010).

Differences among studies regarding associated cost for dementia care can be due to the method used to value informal caregiver time. To value costs of informal care, different methods have been proposed. There is no consensus as to which methodology should be recommended (Jönsson, 2004). One of the methods is “cost opportunity” and another one is to use the market price of an equivalent service, identified as “replacement cost.”

In the review of economic evidence in dementia developed by Jönsson (2004) when informal care is considered, the second option is utilized more frequently; that is, the value of the average salary of a housemaid. The Swedish study was the only study in which the informal care costs were valued at the opportunity of the caregivers' time.

In general, results show that costs of informal care for the elderly with dementia increase according to the level of dementia.

Sample

A cross-sectional, observational study was conducted in five randomly selected towns in the province of Cuenca, Spain. The towns' populations ranged from 1,500 to 3,000 inhabitants. The percentage of dependents in this province and region is similar to Spain' overall population (26.7% and 25.8%, respectively). One nurse from each center evaluated the dependency after receiving training from the research team. Subjects included in the study were residents older than 64 years with a score of three or more on the Pfeiffer index and 95 or less on the Barthel test. These are tests to assess mental status and the ability to function in elderly patients. Meeting the inclusion criteria involved a minimum degree of dependence. This fact implicated a high probability to receive informal care, which is necessary to achieve study objectives. The information was collected in 2007 by personal interviews conducted by the team of nurses with the participants.

A total of 242 subjects met the screening criteria and agreed to participate in the study. All participants were informed of the study objectives and signed and informed consent in which they agreed to participate.

Dependent Variable: Informal Care Hours

Dependent variable of our model is informal care hours. We have used the comprehensive Resident Assessment Instrument for Home Care (RAI-HC) to evaluate older adults living at home. Several studies have demonstrated the reliability and validity of this instrument (Hirdes et al., 2004; Morris et al., 1997). It is being used in Canada, the United States, and in eleven European countries (Sorbye et al., 2009).

The RAI-HC defines informal care as the amount of time family, friends, and neighbors dedicate to assistance in Instrumental Activities of Daily Living (IADL) and Activities of Daily Living (ADL) during the last seven days. The literature has established that the amount of hours is related to the number ADL and IADL disabilities. We calculated the intensity (number of weekly hours) of informal care (ICh/wk) using the sum of two RAI-HC items dedicated to the care provided (Section G3, items 1 and 2 of RAI-HC). Any values in excess of 16 hours were reduced to 16 hours to avoid the well-known overestimation of the number of hours of informal care received. This limit is also used in the studies with the aim of allowing informal caregivers at least 8 hours of sleep daily (Ernst & Hay, 1994; Langa et al., 2001). This adjustment affected 5% of the sample.

Independent Variables: Evaluation of Dementia

The presence and severity of dementia was measured using the Cognitive Performance Scale (CPS). It is a predictive algorithm on the basis of a decision tree that grades according to seven levels of cognitive performance from 0 (intact) to 6 (seriously deteriorated). This scale derives from RAI-HC and correlates well with other frequently used and accepted measurement instruments, including the Mini-Mental State Examination (MMSE) (Paquay et al., 2007).

On the basis of prior studies, we defined a cut-off score of three as the level of cognitive impairment consistent with dementia. In order to assess dementia severity, we created dummy variables representing three stages of dementia: Stage 1 for moderate impairment (CPS = 3; MMSE = 15.4), stage 2 for moderate-severe impairment and severe impairment (CPS = 4–5, MMSE = 6.9–5.1), and stage 3 for very severe impairment (CPS = 6, MMSE = 0.4). Scores 4 and 5 were overlapped due to the low number of subjects in one scoring group.

Empirical Model and Costs of Informal Care

Since the goal of the analysis was to identify the additional informal caregiving attributable to dementia, we construct a multivariable regression model. Others studies such as Langa et al. (2001) used this type of model. The model was calculated using the Napierian logarithm of the hours of informal care as the dependent variable and the following independent variables: a group of categorical variables indicating the presence of moderate, moderate-severe, or very severe dementia; categorical variables indicating socio-demographic characteristics; and categorical variables indicating the presence of comorbidities and hearing or vision problems. In the first step, all the variables were introduced in the model, and later we used the backward method with the aim of developing the most cost-effective model. Variables were selected through a review of the literature.

In the resulting statistical model, the regression coefficients for the dementia severity variables indicate the additional caregiving associated with moderate, moderate-severe, or very severe dementia compared to those with normal cognitive function, after adjusting for differences across groups.

As the hours of care were corrected by the Napierian logarithmic transformation to yield a normal distribution, it was necessary to perform a transformation to calculate the time of informal care in each stage of dementia. To do this, we used a technique that allows improved estimation efficiency (Manning & Mullahy, 2001; Moya-Martínez et al., 2009). After applying transformations, we obtain the estimates of the informal care hours attributable to the different stages of dementia adjusted by the factors and covariates included in the final regression model.

The factors and covariates included are others' health measures. Possible hearing impairment and vision deficits were evaluated in each subject using items C1 and D1 of RAI-HC. Furthermore, the RAI-HC has a section dedicated to the diagnosis of diseases. The presence or absence of the diseases that were frequent enough for comparative analyses to be made between persons with and without dementia was assessed in the study. The variables of the physical function domain to assess the use of locomotion aids were also included.

Additionally, other measures were included. Participants or the principal caregiver, in the case of subjects who could not ask due to dementia or other comorbidities, were asked about their socio-demographic characteristics and the specific characteristics of their caregivers.

The amount of formal care was not available through the RAI-HC, but it includes information about the use of the next formal services: house-aid, at-home nurses, and domestic service.

Costs of Informal Care

To value informal care time, depending on the objectives of economic evaluations, different methods can be used such as opportunity cost method, proxy good method, contingent evaluation, conjoint analysis, and the replacement costs (Koopmanschap, van Exel, van den Berg, & Brouwer, 2008). All of these methods have advantages and disadvantages.

Opportunity cost is the preferred measurement of cost for economic analyses in health care (Glick, Doshi, Sonnad, & Polsky, 2007). The opportunity cost of an informal caregiver's time is sometimes assigned using the average hourly wage of working individuals with similar characteristics, but for some groups of caregivers (retired elderly, for example) there is no appropriate wage data. An alternate approach is to use the market price of an equivalent service as an estimate of the opportunity cost of a caregiver's time. Due to the nature of data, we used the replacement cost approach. Using this method, the yearly cost of family caregiving for each level of dementia was estimated (Glick et al., 2007).

To apply this method, the cost of informal care for persons in each stage of dementia was calculated by multiplying the adjusted hours of informal care in each stage by the mean cost per hour of care. We used two plausible scenarios through the cost per formal care hour offered by a specialized enterprise in Spain (mean €12; range €10 to €14) and the cost per hour of contracting a support worker to do the task (mean €8.5; range €7 to €10). The hourly cost was obtained from a special report by the company DBK, S.A.. The confidence intervals for these costs were obtained with the confidence intervals of each dementia stage in the regression model.

RESULTS

This study analyzed a total of 242 assessments of people aged over 64 years. Mean subject age was 81.09+/−7.08 years. The mean CPS score of subjects with dementia was 4.70+/−1.06 (CPS scale grades from 0 to 6). The prevalence of dementia was 39.3%. There were no significant differences in the prevalence of dementia between men (44%) and women (36.4%).

The distribution of the subjects without dementia and in different stages of dementia is shown in Figure 1. We can see that 20% of the subjects with dementia were in stage 1, 57.89% were in stage 2, and 22.11% were in stage 3.

Figure 1 Distribution of subjects without dementia and in different stages of dementia (n = 242). *p values were calculated by the way of different in adjusted hours of informal care of each group.

Socio-demographic and clinical characteristics of the participants with and without dementia were similar, except for hearing and vision, the diagnosis of osteoporosis, and the diagnosis of emphysema/asthma/COPD (Table 1).

Table 1 Descriptive Analysis of the Sample (n = 242)

		No dementia n (%)	Dementia n (%)	p
Gender	Women	96 (63.6)	55 (36.4)	0.245
	Men	51 (56.0)	40 (44.0)
Age (years)	64–75	39 (65.0)	21 (35.0)	0.051
	75–85	79 (65.3)	42 (34.7)
	+85	29 (47.5)	32 (52.5)
Marital status	Widow(er)	51 (53.7)	44 (46.3)	0.046
	Married	90 (67.7)	43 (32.3)
	Unmarried	5 (38.5)	8 (61.5)
Education	Elementary school	27 (75.0)	9 (25.0)	0.058
	No schooling	120 (58.3)	86 (41.7)
Primary caregiver	Lives with subject	119 (58.3)	85 (41.7)	0.059
	Spouse	69 (67.6)	33 (32.4)	0.061
Hearing	Adequate or minimal difficulty	128 (73.6)	46 (26.4)	<0.001
	In special situations / severely altered	19 (27.9)	49 (72.1)
Vision	Adequate or minimally altered	121 (79.1)	32 (20.9)	<0.001
	Moderately to seriously altered	26 (29.2)	63 (70.8)
Comorbidities^a	Coronary artery disease	5 (38.5)	8 (61.5)	0.091
	Osteoporosis	42 (76.4)	13 (23.6)	0.007
	Emphysema /asthma /COPD	32 (84.2)	6 (15.8)	0.001
Locomotion aid	Cane	61 (69.3)	27 (30.7)	0.039
	Walker	40 (58.8)	28 (41.2)	0.702
	Wheelchair	9 (42.9)	12 (57.1)	0.079
Formal care	House-aid	44 (71.0)	18 (29.0)	0.052
	At-home nurses	36 (72.0)	14 (28.0)	0.063
	Domestic service	10 (50.0)	10 (50.0)	0.312

^aOnly comorbidities which significant differences between dementia groups were included. All evaluated were: congestive heart failure, coronary artery disease, hypertension, cardiac arrhythmias, Parkinsonism, arthritis, osteoporosis, cancer, diabetes, emphysema /asthma /COPD, renal failure, and thyroid disease.

The best adjustment variables used to estimate ICh/wk derived from dementia were found after thorough literature review and by calculating means differences of ICh/wk among subjects with and without dementia and means differences among other characteristics.

The final model (Table 2) was the most parsimonious model, but the estimated effects of ICh/wk were not significantly different in the models with the full set of controls. The model explained 30% of the variation in the Napierian logarithmic of ICh/wk. The estimations were made using the following control variables: hearing problems, walker use, wheelchair use, at-home nursing care, and domestic service. We observed that hearing problems, walker use, wheelchair use, and at-home nursing care increased ICh/wk (44.6%, 89.7%, and 41.7%, respectively).

Table 2 Final Regression Model with the Dependent Variables (ln) of Weekly Hours of Informal Care

		β	95% confidence interval		p
Dementia
	Stage 1 (moderate impairment)	0.441	0.069	0.813	0.020
	Stage 2 (moderate severe and severe impairment)	0.538	0.269	0.807	<0.010
	Stage 3 (very severe impairment)	0.664	0.278	1.049	<0.010
Hearing deficit	0.356	0.108	0.604	<0.010
Walker-assisted locomotion	0.446	0.220	0.672	<0.010
Wheelchair-assisted locomotion	0.897	0.531	1.264	<0.010
At-home nurses	0.417	0.170	0.664	<0.010
Domestic service	–0.449	–0.813	–0.085	0.016
Constant	2.316	2.153	2.480	<0.010
R square	0.325
Corrected R square	0.301

However, this effect was reversed by the availability of domestic service that reduces the time devoted to informal care in 44.9%. The primary explanatory variables of interest in the model, stages of dementia, showed an important positive association with an increment in informal care time of 44.1% for stage 1, 53.8% for stage 2, and 66.4% for stage 3. However, the wheelchair variable had the most positive association, assuming a major increment in ICh/wk. We estimated that those subjects without dementia received 700.78 adjusted annual hours of informal care and those with dementia increased the number of hours by 388.41, 499.36, and 660.52 depending on the dementia stage.

Figure 2 represents the replacement costs in the three stages of dementia through the two-cost scenarios. The cost of informal care due to dementia was between €3301.52 and €4660.97 per person/year for moderate dementia (stage 1), €4244.5 and €5992.32 per person/year for moderate-severe dementia (stage 2), and €5614.39 and €7926.20 per person/year for very severe dementia (stage 3).

Figure 2 Annual cost of informal care at different stages of dementia calculates using replacement cost.

DISCUSSION AND CONCLUSIONS

The present study quantifies the time committed and activity performed by informal caregivers, highlighting the importance of informal care, not only from a social point of view, but also from an economic perspective, because of the limited existing resources in Spain regarding caregiving for these types of patients (geriatric homes, health institutions, and home care).

Our study shows that dementia increases the need for informal care time between 44.1% and 66.4% with a cost between €3000 and €8000 per person/year depending on the service provider, enterprises, or individuals and supply the information and tools needed for politicians and administrators to evaluate the most cost-effective policies. These results are according to other studies; for example, in the Unites States, the estimated annual cost of informal care is $2,610 to 23,310 per case depending on the degree of deterioration (Langa et al., 2001) and in the European countries the cost of informal care varies in a range of €624.72 to €17184 depending on the country, methodology, and dementia grade (Jönsson, 2004).

In the case of Spain, the annual cost of dependents is €16449. 61, 2.6 times more than budget devoted to this task by the Central Government (Escribano-Sotos, Moya-Martínez, & Pardo-García, 2012).

We used a regression model that through dummy variables discriminates between subjects with dementia (or in different stages) and subjects without dementia and adjusts for the heterogeneity of individuals through a set of covariates and factors.

Bivariate and multivariate analysis of the covariates and factors associated with informal care time revealed that hearing problems, walker-assisted locomotion, wheelchair-assisted locomotion, at-home nursing care, and domestic service were the best control variables for predicting the quantity of informal care time required for subjects with dementia. In our final model, the adjustment variable of formal at-home nursing care increased ICh/wk, reflecting the seriousness of the patient's comorbidities, rather than a usual care situation. In a longitudinal study, it has been shown that medical formal services decrease informal care time in the early years of aging, but the amount of informal care to control functional status eventually stabilizes, thus supporting our findings (Li, 2005). In addition, ICh/wk was negatively associated with the use of domestic household services, which seemed to be a substitute for informal care. Similar results have been found in other studies (Arksey & Glendinning, 2007; Hammar, Rissanen, & Perälä, 2008).

This is the first study in Spain that evaluated the costs of informal care due to dementia. The main strength of the study lies in the bottom-up approach. Also, because the costs estimated over moderate, moderate-severe, and severe dementia rather than being limited to overall patients with dementia, the results are a more accurate representation of the cost of informal care of dementia.

There are limitations to the validity of the data we collected and to the results we have presented herein.

The data came from a single region, which may limit generalizations. However, because there is no reason to assume that the time consumed in the study would be substantially different from elsewhere in Spain, it is reasonable to assume that the estimated costs are representative. Another limitation is that we could not differentiate between hours of informal care for activities of daily living (ADL) such as eating or showering and instrumental activities of daily living (IADL) such as shopping or housework. It should be noted that the present study use replacement cost to value informal care time. Although it is the common solution in the literature, results may be slightly different if we use another method as the opportunity cost. Additionally, new research is necessary to compare data from 2007 onward.

In conclusion, the main objective is to determine informal care costs of dementia in order to shape care policies. We found that dementia increased with age and was associated with more comorbidity and hearing and vision problems. Other studies in the literature support this finding (Spijker et al., 2008).

The comparison of informal care cost of people with dementia across studies is difficult due to methodological and socio-cultural between countries differences. For example, in Belgium is estimated at €624.72, in France at €2114 for those with MMS>20 and at €11276 for those with MMSE between 11 and 15 or Ireland at IR£4097 (Jönsson, 2004). Despite the difficulties in comparison, our study is in the same framework of costs, but our calculations are more similar to studies in the literature with the lowest estimates. This is due to the use of regression models to isolate the cost of informal care of dementia, which allows for a better fit.

These results provide information to managers in order to avoid costly institutional care and to identify best-practice strategies for long term care related to dementia. In this sense, this study provides important information that allows government administrators and the community to understand what it would cost to reduce the number of hours that informal caregivers dedicate to persons with dementia and what would make it possible to lighten the burden on health and the quality of life that this excess care load entails. Future interventions in the field of dementia may include this cost in their cost-effectiveness analysis.

If relatives were not available as a source for providing care, the public institutions would have the obligation of financing and providing continuous care services. The lack of home care would force the public services to use paid human resources for the care of those patients, as well as to increase, remarkably, the institutionalization resources that are already available. In that sense, it is necessary to consider that the World Health Organization stresses that strategies should be drawn up for providing support to patients and caregivers at the community level in order to avoid costly institutional care (Drake & Davies, 2006).