The Utilization of Home Care by the Elderly in Brazil's Primary Health Care System

Abstract

Objectives. We assessed the utilization of home care by the elderly in Brazil after implementation of the Family Health Strategy (FHS).

Methods. Data were derived from a cross-sectional study in a southern city in Brazil. Using the χ² test and a logistic regression with different levels of determination, we tested the hypothesis that the FHS increased the utilization of home care compared with utilization under the Traditional Primary Health Care (TPHC) system.

Results. We interviewed 1593 residents aged 60 years and older. Home care utilization under the FHS was 2.7 times the rate of utilization under the TPHC (95% confidence interval = 1.5, 4.7; P = .001), and utilization increased among the older group, the less educated, those with history of hospitalization, and those with functional limitations.

Conclusions. Improvement in access to care resulted in greater utilization of home care. Our findings have policy implications that include expanding the coverage of the FHS throughout big cities where coverage is limited. These findings are important because the population is aging and the family strategy operates in poorer areas; thus, it can promote equity in access to home health care among the elderly.

The aging of the population represents a challenge to governments around the world, which are faced with the task of designing and implementing national strategies for elderly care, including improving primary health care through home health care systems and social networks.¹ The growth of disabled elderly populations is likely to increase future demand for home health care globally.^2,3

In Brazil, access to health care is a universal social right guaranteed by the government since 1988 through the Unified Health System (Sistema Único de Saúde).⁴ The Unified Health System is a decentralized system, and the organization and delivery of health care is the responsibility of municipal authorities with financial and technical support provided by state and federal governments. The local government organizes the physical structure, equipment, and professional teams to provide the population with access to health services. The goal is to invest in strong primary health care, and to this end an optional alternative model, the Family Health Strategy (FHS), was developed during the 1990s.^4,5 The FHS has subsequently been adopted by many municipalities in Brazil and has replaced the Traditional Primary Health Care (TPHC) system in many areas. TPHC was introduced in the Brazilian public system during the 1980s, after the Alma-Ata Declaration in 1978 identified primary health care as the key to the attainment of the goal of Health for All.⁶

Both TPHC and FHS are part of the same public health system at the primary level. However, the FHS was introduced to reorganize or restructure primary care in the Brazilian health system. The planning and design of the FHS had the benefit of the earlier development of models of family care (in Quebec, Canada; Cuba; Sweden; and England), which served as guides for the formulation of the Brazilian model.⁷ There are important differences between the FHS and TPHC. Under the FHS, a team of 1 general practitioner, 1 nurse, 1 or 2 nurse assistants, and 4 or 5 community health workers are responsible for meeting the heath care needs of approximately 1000 families in a defined geographical area. TPHC teams, by contrast, do not have a fixed structure. They contain more medical professionals than do FHS teams, sometimes including specialists in internal medicine, pediatrics, and obstetrics and gynecology as well as general practitioners, nurse assistants, and nurses. TPHC teams also do not serve a defined number of families or geographical area, and not all TPHC teams include community health workers.⁵

Traditional care focuses on the specific disease, and the health team usually remains in the medical facilities from which it dispenses curative care, acting on emerging demands. With an emphasis on curative medicine, the TPHC treats the individual as a subject of concern, with TPHC teams having little ability to solve the health problems of a family or to take a social perspective. The TPHC is usually limited to sectoral actions and there is little involvement by TPHC service staff with the community. In the FHS, the household is part of the environment of care and FHS workers are expected to be proactive professionals who identify the most vulnerable members of the population.

The implementation of the FHS is a political process undertaken by municipal governments that can choose either to maintain the TPHC model or to implement the new strategy.⁵ In 2008, family health teams were operating in 5218 municipalities, representing 94% of Brazilian municipalities and offering coverage to 49% of the national population,⁸ and were organized consistently across Brazil. However, coverage differed depending on municipality size. In municipalities with fewer than 20 000 inhabitants, the coverage averaged approximately 82%. In municipalities with more than 100 000 inhabitants, especially in the south and southeast regions of Brazil, the FHS covered only 35% of the population. The remaining population is covered by the TPHC. With coverage increasing, the Ministry of Health has demanded an assessment of FHS performance compared to TPHC performance in delivering home care for the elderly.

Recent decades have witnessed an increase in home health care utilization globally. The use of home care increases with increasing age, functional limitations, chronic disease, stroke, heart disease, and falls.^3,9–11 In Brazil, home health care has not been investigated in previously published epidemiological studies about health service utilization.^12–14 On the basis of activities of daily living and self-reported needs, some studies have estimated the need for home health care to range from 11%–25% of the population.^10,15

Statements by the Ministry of Health and by other government officials have indicated that the policy of the primary health services, as explained in the Brazilian National Policy for the Elderly, is to improve health care service delivery by keeping elderly citizens in their communities, close to their family environments.^15–17 Other investigators have found that utilization of home health care will be affected by perceptions of care needs, by the ease of access to home health care, by the population's acceptance of home health care, and by personal decisions to use home health care.^18–21 Assuming that FHS teams are available, that FHS teams support family involvement in care, and that FHS teams increase awareness about health as a right, utilization rates are expected to improve in populations under the care of a family workers’ team.

The primary aim of this paper was to test the hypothesis that the FHS increases the utilization of home health care by the elderly when compared with the TPHC. A secondary aim was to investigate the sources of the home health care provided (i.e., whether public or private services were used).

METHODS

We conducted a cross-sectional epidemiologic study during 2008 to examine utilization of home health care by a representative sample of elderly residents aged 60 years and older¹⁶ of Bagé, a southern city of Rio Grande do Sul state with approximately 120 000 inhabitants. Accounting for expected refusals to participate and an anticipated design effect of approximately 1.3, we estimated that a sample size of 1530 individuals would result in 80% power to detect a minimum relative risk of 1.5 for exposures affecting up to 4% of the population.

The decision to implement the FHS model in Bagé was made by municipal government leaders in 2003. At that time, 13 TPHC centers served the city. Three TPHC centers adopted the FHS in 2003, followed by 2 centers in 2004, 2 in 2006, and 1 in 2007. In the period from 2004 to 2008, 7 new primary health centers were built to increase FHS coverage in Bagé. By 2008, Bagé had 20 health service centers, 15 of them utilizing the FHS model and 5 following the TPHC model (Table 1). Half of the city's population (51%) was covered by the FHS model in 2008.

TABLE 1.

Characteristics of Primary Health Care Services in Bagé: Brazil, 2008

Characteristic	Traditional Primary Health Care	Family Health Strategy
No. of primary health care centers in the urban area	5	15
Estimated no. of elderly living in the primary health care center area	5585	6464
Estimated no. of elderly per primary health care center	1117	430
Community health workers living in the Primary Health Care center area or neighborhood	Sometimes present; no requirement	4–6 per team
Continuity of care per team and days per week	Same team works 4 h/d, 5 d/wk	Same team works 8 h/d, 5 d/wk

The sample was located within urban districts around the primary health centers. First, the areas of the primary health centers were delimitated. After that, the areas were divided into micro areas, with enumeration of the blocks. To guarantee equiprobability at the household level, different starting points were randomly selected. The next household to the left was selected and then every other fifth household, until a total of 5 were chosen. No replacements were admitted. All those persons aged 60 years and older living in the selected households were included in the study. Field workers made at least 3 attempts to interview household members. Given the proportion of elderly individuals in the general population (8%–10%), the probability of locating an elderly household member was estimated at 1 in every 3 households, thus ensuring good distribution of the sample within the area of coverage of each primary health center.

The data were collected through a structured questionnaire with precoded questions after a pilot study. Household interviews were carried out by 15 trained interviewers under the supervision of 3 supervisors. In cases of partial disability, another household member was asked to help the participant respond to the questionnaire. In cases of total disability, questions requiring patient self-reporting were not included.

Outcome Measure

The primary outcome was utilization of home health care. The data were obtained by asking the question, “During the past 3 months, did you receive care at home from a health care professional?” Respondents answered yes or no. If the answer was yes, we asked a follow-up question: “Did you receive the care from a professional from the public services (yes or no), a professional from the private services (yes or no), or a professional from private health insurance (yes or no).” The data were classified into 3 categories: public services (care provided by public employees), private services (care provided by private professionals paid out-of-pocket or by private health insurance), and mixed (care provided by public and by private services).

Covariates

The socioeconomic and demographic variables included in this study were gender, age (complete years and stratified in 4 groups), skin color (self-reported as white, black, brown, yellow, or indigenous and categorized in 3 groups: white; black; or brown, yellow, and indigenous), education (having attended school; yes or no), marital status (married or with a partner, widowed or never married, or divorced), health insurance status, and household income in the last month (collected in Brazilian money and converted to US dollars and analyzed by tertiles). Morbidity variables included self-reported medical diagnoses of heart disease and stroke and also included questions about hospitalization in the previous year. Symptoms of depression were measured according to the 15-item Geriatric Depression Scale,²² and the results were dichotomized as no symptoms of depression present (score: 0–5) versus moderate to severe symptoms of depression present (score ≥ 6). A Mini-Mental test was used to assess cognitive function and carry out the screening for dementia, and the results were dichotomized in 2 categories: with or without dementia signs according to the status of school attendance.²³ Functional limitation was assessed through 5 indicators (difficulty in walking, standing, moving from bed to chair, shopping, and using bus or taxi) and was stratified in 3 categories: no limitation, mild limitation, and severe limitation (unable to perform 3 or more of the 5 indicators alone). Self-rated health status was collected in 5 categories and was categorized in 2 groups: worse, bad, and regular, versus good and very good. The primary health centers were classified in 2 categories: TPHC and FHS.

Analyses

The data were entered with dual input by using Epi Info version 6.04 software (Centers for Disease Control and Prevention, Atlanta, GA) and were analyzed by using Stata version 10.0 (StataCorp, College Station, TX). Characteristics of home health care in the FHS and TPHC models were compared by using the χ² test and Fisher exact test. Crude and multivariate analyses assessed the strength of the association between each independent variable and home health care utilization by use of logistic regression. We used the Wald test for heterogeneity and for linear trend in the case of ordinal variables, producing odds ratios (ORs) and their respective 95% confidence intervals (CIs).

Adjusted analysis was based on a conceptual model with 4 levels of determination. The first level included gender, age, education, skin color, marital status, household income, and health insurance. The second level included heart disease, stroke, depression, dementia, hospitalization, and functional limitation. The third level included self-perceived health. The fourth level included the primary health centers (FHS vs TPHC). For confounder control, the effect of each variable was controlled for all other variables in the same or higher levels with P values ≤ .20.

RESULTS

We identified 1713 elderly household members and interviewed 1593, for a response rate of 93%. Losses represented 4% and refusals 3% of the identified household members. Of the total sample, 852 respondents (54%) lived in areas utilizing the FHS model. The respondents’ mean age was 70.6 years (SD = 8.1 years) in the FHS areas and 71.8 years (SD = 8.4 years) in the TPHC areas (P = .001).

Descriptive statistics for the sample, including demographic, socioeconomic, and morbidity characteristics as well as health status and home health care utilization, are shown in Table 2. The TPHC group and the FHS group were similar in terms of gender, marital status, history of stroke, perceived health status, and hospitalization in the previous year. Compared with the TPHC sample, the sample in the FHS areas included smaller percentages of those aged 75 years or older (28% and 35%, respectively), of those with white skin color (73% and 85%, respectively), of those who had been to school (68% and 85%, respectively), and of those who had private health insurance (27% and 45%, respectively), as well as a smaller percentage of elderly in the third tertile of household income (27% and 45%, respectively). In the FHS areas, the elderly displayed higher rates of heart disease, depression, and dementia than in the TPHC areas. The rate of home health care utilization in the total sample was 7%. In the FHS areas the utilization rate was 10%, compared with 4% in the TPHC areas (P < .001).

TABLE 2.

Sample Description, by Demographic, Socioeconomic, and Morbidity Variables, and Primary Health Care Center Areas: Brazil, 2008

	Total		TPHC		FHS
Variables	%	No.	%	No.	%	No.	P
Gender		1593		741		852	.1
Men	37.2		35.1		39.1
Women	62.8		64.9		60.9
Marital status		1592		740		852	.76
Divorced or never married	14.9		14.8		15.1
Married	51.3		50.5		51.9
Widowed	33.8		34.7		33
Age, y		1593		741		852	.01
60–64	25.1		22		27.8
65–69	23.5		22.5		24.2
70–74	20.2		20.7		19.8
≥75	31.2		34.8		28.1
Skin color		1557		720		837	< .001
White	78.6		85.1		72.9
Black	8.7		7.2		10
Brown, Yellow, or Indigenous	12.7		7.7		17.1
Attended school		1593		741		852	< .001
No	23.7		14.6		31.6
Yes	76.3		85.4		68.4
Household Income, R$		1565		726		839	< .001
1st tertile (≤ 410.00)	46.5		37.9		53.9
2nd tertile (410.01–650.99)	20.5		21.6		19.6
3rd tertile (≥ 651.00)	33		40.5		26.6
Private health insurance		1586		739		847	< .001
No	64.6		54.7		73.3
Yes	35.4		45.3		26.7
Morbidity
Heart disease	29.6	1593	26.3	741	32.4	852	.008
Stroke	9.8	1593	9.5	741	10.2	852	.61
Depression	14.9	1512	12.9	703	16.6	809	.049
Dementia	13.1	1514	10.4	703	15.4	811	.003
Hospitalization last year	17.7	1592	17.3	741	18.1	851	.668
Functional limitations		1593		741		852	< .001
No	77.8		82.3		73.8
Mild	16.1		11.5		20.1
Severe	6.1		6.2		6.1
Perceived health		1540		713		827	.11
Worse, bad, or regular	41.2		39		43.1
Good or very good	58.8		61		56.9
Home health care		1593		741		852	< .001
No	93.2		96.5		90.2
Yes	6.8		3.5		9.8

Note. FHS = Family Health Strategy; R$ = Brazilian reais; TPHC = Traditional Primary Health Care. At the time of this study, R$ 1.80 = US$ 1.00. P values are based on the χ² test.

The results of the crude and adjusted analyses examining the odds of home health care utilization during the past 3 months and the independent variables are shown in Table 3. The crude analyses showed that gender, household income, and private health insurance were not associated with utilization of home health care. The odds of utilization of home health care by respondents aged 75 years and older was 3.3 times (95% CI = 1.8, 5.9; P < .001) that for respondents aged 60 to 64 years. For respondents reporting brown, yellow, or indigenous skin color, the odds of using home health care was 1.8 times that for the white group (95% CI = 1.1, 3.1; P = .02). If a respondent had not attended school, the odds of home health care use was double that of the group that had attended school (95% CI = 1.4, 3.1; P = .03). All of the variables related to health problems showed a strong association with utilization of home care. Respondents with severe functional limitations were 20.4 times as likely to use home care as were those without functional limitations. The odds of using home health care when respondents characterized their health condition as bad, worse, or regular was 2.3 times the odds when reporting health condition as good or very good (95% CI = 1.5, 3.6; P < .001). Also, in the crude analyses, the utilization rate of home care in the FHS areas was roughly 3.0 times the rate in the TPHC areas (95% CI = 1.9, 4.7; P < .001).

TABLE 3.

Crude Odds Ratios (ORs) and Adjusted Odds Ratios (AORs) for Home Health Care Utilization, by Demographic, Socioeconomic, and Morbidity Variables: Brazil, 2008

Model Level and Variables	Crude OR (95% CI)	P	AOR (95% CI)	P
Level 1 modela
Gender		.47
Men (Ref)	1.00
Women	1.17 (0.77, 1.76)
Marital status		.001		.12
Divorced or never married (Ref)	1.00		1.00
Married or with a partner	1.55 (0.75, 3.22)		1.63 (0.76, 3.53)
Widowed	2.79 (1.35, 5.76)		2.17 (0.99, 4.74)
Age, y		.001b		< .001b
60–64 (Ref)	1.00		1.00
65–69	0.93 (0.44, 1.98)		0.98 (0.45, 2.13)
70–74	2.07 (1.07, 4.00)		2.13 (1.07, 4.21)
≥ 75	3.32 (1.85, 5.97)		3.36 (1.81, 6.23)
Skin color (self-reported)		.03		.04
White (Ref)	1.00		1.00
Black	1.15 (0.56, 2.35)		0.91 (0.42, 1.96)
Brown, Yellow, or Indigenous	1.82 (1.08, 3.06)		1.93 (1.15, 3.25)
Attended school		.001		.03
Yes (Ref)	1.00		1.00
No	2.01 (1.37, 3.08)		1.63 (1.04, 2.54)
Household income, R$		.34b		.06b
1st tertile (≤ 410.00; Ref)	1.00		1.00
2nd tertile (410.01–650.99)	0.65 (0.36, 1.16)		0.57 (0.31, 1.03)
3rd tertile (≥ 651.00)	0.84 (0.54, 1.29)		0.63 (0.40, 1.00)
Private health insurance		.91
No (Ref)	1.00
Yes	0.98 (0.65, 1.47)
Level 2 modelc
Morbidityd
Heart disease	2.06 (1.39, 3.05)	< .001
Stroke	5.00 (3.21, 7.82)	< .001
Dementia	4.15 (2.63, 6.54)	< .001
Depression	3.38 (2.13, 5.36)	< .001	1.56 (0.91, 2.66)	.1
Hospitalization last year	4.04 (2.67, 6.05)	< .001	3.13 (1.92, 5.10)	< .001
Functional limitations		< .001		< .001
No (Ref)	1.00		1.00
Mild	8.81 (5.44, 14.28)		6.69 (3.83, 11.70)
Severe	20.43 (11.74, 35.54)		11.16 (5.02, 24.79)
Level 3 modele
Health condition		< .001
Worse, bad, or regular	2.33 (1.52, 3.57)
Good or very good (Ref)	1.00
Level 4 modelf
Primary health center		< .001		.001
TPHC (Ref)	1.00		1.00
FHS	2.98 (1.89, 4.68)		2.65 (1.52, 4.63)

Note. CI = confidence interval; FHS = Family Health Strategy; R$ = Brazilian reais; TPHC = Traditional Primary Health Care. The sample size was n = 1593. P values are based on the Wald test for heterogeneity. Adjusted R² = 0.2285.

^aAdjusted for variables in the same level 1.

^bP values are based on the Wald test of tendency.

^cAdjusted for variables in the same level and level 1.

^dComparison group was no health problems (1.0).

^eAdjusted for variables in the same level and levels 1 and 2.

^fAdjusted for all levels.

The right column in Table 3 shows the multivariate logistic regression models. In the first level of regression, the variables gender and private health insurance did not stay in the model after adjustment (P > .20). Age, skin color, and school attendance remained significantly associated with home health care utilization. The significance of household income increased after adjustment and stayed in the model to control for the other variables. In the second level, after adjustment, only hospitalization and functional limitation remained significantly associated with home health care utilization. Depression lost its association with home health care utilization after adjustment (P = .11). In the third level, the association between self-reported health conditions disappeared. In the fourth level, even after adjustment for all the previous variables, the odds of home health care utilization in the FHS areas was approximately 2.7 times (95% CI = 1.5, 4.7; P = .001) that in the TPHC areas.

The status of the home health care providers in relation to the different types of primary health centers is shown in Table 4. In FHS centers, most of the home health care was provided by public services (86%), whereas in TPHC centers, more elderly used private services (77%) than public services. This difference was statistically significant (P < .001).

TABLE 4.

Home Health Care Providers in Bagé, by Type of Primary Health Care: Brazil, 2008

	Total, No.	TPHC, No. (%)	FHS, No. (%)	P
Home health care provider	109	26	83	< .001
Public	75	4 (15.4)	71 (85.5)
Private	28	20 (76.9)	8 (9.6)
Mixed	6	2 (7.7)	4 (4.8)

Note. FHS = Family Health Strategy; TPHC = Traditional Primary Health Care. P values are based on the Fisher exact test.

DISCUSSION

The results of our study showed that the FHS, as defined in the guidelines of the Brazilian Unified Health System and the National Policy of the Elderly, increases the utilization of home health care, thus supporting our hypothesis. The utilization of home health care was higher in areas where the FHS operated than in the TPHC areas (OR = 2.7; 95% CI = 1.5, 4.8), even after control for gender, age, skin color, education, marital status, household income, private health insurance, morbidity, hospitalization in the past year, functional limitations, and self-reported health condition.

We also found that the majority of home health care was delivered by public services in the FHS areas but by private providers in the TPHC areas. This difference may be explained by 3 factors. First, 1 FHS team is responsible, on average, for 430 elderly residents, whereas under the TPHC model, 1 team was found to care, on average, for 1117 elderly residents (Table 1), thus increasing the need for additional (nonpublic) resources. Second, in the FHS areas, community health workers conducted monthly home health visits and other activities to connect elderly residents with the health team. Third, the FHS model includes follow-up care.

These findings about the utilization of home health care are important because the FHS operates in poorer areas of the municipalities (In Bagé, 54% of elderly individuals are poor), thus strengthening its relevance for the promotion of equity in health care. These findings are similar to those of other Brazilian studies showing that the FHS provided more equitable care to the elderly with chronic conditions and also contributed to reducing infant mortality.^{10,13,14,24,25}

If we look at the FHS as a new technology for delivering health care, we can say that these results indicated its acceptance by the populations served. The patterns of illness, emphasizing chronic conditions, associated with home care utilization in this study were similar to those found in the United States²⁶ and with the need patterns reported by the World Health Organization.²⁷ Cardiovascular disease, followed by other chronic diseases, has been found to increase the need for home health care.^27,28

In Brazil, the organization of home health care is in its early stages, in both the public and private health systems.^29,30 The country does not have institutional providers of in-patient long-term care, which are referred to as nursing homes in the United States.³¹ Nor are Brazilian health care officials currently capable of analyzing home care expenditures. But health officials are confident that life expectancy will continue to increase, especially among those older than age 80 years. This will present strong social and economic challenges in the coming decades.^5,32,33

The strengths of our study were its large sample size (n = 1593), the use of validated instruments, and the high response rate of 93%. Our robust findings have important policy implications for improving Brazil's health system. There were, however, some limitations to our study. The small number of elderly in the TPHC who made use of home health care made it impossible to do further stratification. We lost information about depression and dementia because the Geriatric Depression Scale and the Mini-Mental test could not be applied for respondents unable to self-report these answers. In these cases, the other elements of the questionnaire were filled in by caregivers.

The results of our study showed that the elderly and their families use public health services when they are available and that the utilization of home health care has increased among the older group, the less educated, those with history of hospitalization in the previous year, and those with functional limitations. We also estimated the rate of home health care utilization by Brazil's elderly population. This is an important result and showed that improvement in access to home health care results in an improvement in utilization of these services and, perhaps, in quality of life.

Delivering care to the elderly is a problem not confined to developing countries but represents a global challenge. Brazil faces the task of finding, in a short period of time, a solution to improve access to health care for the poor and frail elderly who reside in the TPHC areas. Brazilian health officials can capitalize on the information presented here to increase access by encouraging local officials currently utilizing the TPHC system to switch to the FHS system.

Although our results confirm increases in utilization related to the delivery model, they do not inform Brazilian health officials about potential related impacts on the quality and length of life. These questions will require additional, more complex research.

Future research should also examine more thoroughly various aspects of care including those related to the health systems, such as the types of professionals involved or procedures performed, the economic impact, and the availability of social support networks. For example, being married and living with one's spouse has been associated with a lower likelihood of using home health care, according to a home care study in the United States.¹⁰ Findings in these areas could be used to estimate the expenditures of time and money needed to deliver care at home and to improve the quality of the care.