Services on Demand
On-line version ISSN 1518-8787
Rev. Saúde Pública vol.39 no.1 São Paulo Jan. 2005
Moisés GoldbaumI; Reinaldo José GianiniI; Hillegonda Maria Dutilh NovaesI; Chester Luiz Galvão CésarII
de Medicina Preventiva. Faculdade de Medicina. Universidade de São Paulo
(USP). São Paulo, SP, Brasil
IIDepartamento de Epidemiologia. Faculdade Saúde Pública. USP. São Paulo, SP, Brasil
Family Health Program (FHP) is a strategy for reorganizing the healthcare attendance
system within the Brazilian National Health System. The objective of the study
was to assess whether there had been changes in the utilization profile of the
healthcare services following implementation of the program, and to identify
factors associated with any changes observed.
METHODS: Data on service utilization and demand for attendance were analyzed by means of two cluster-based population samples, representing areas covered (n=1865) and not covered (n=2036) by the FHP, in two districts of the municipality of São Paulo. The data formed part of a population survey carried out in 2001. Statistical methods for cluster analysis were used.
RESULTS: In the area covered by the FHP, no statistically different prevalence ratios (PR) according to schooling and income levels were observed for service utilization. In the area not covered by the FHP, service utilization was positively associated with greater schooling and income. Among individuals with illnesses, the demand for attendance in the area covered by the FHP was higher (higher PR) among those with severe physical limitations. In the area not covered, the PR was higher among those with greater schooling and lower among those who were inactive (unemployed or retired).
CONCLUSIONS: In the areas studied, for the population covered by the FHP, income and schooling levels did not constitute factors that significantly differentiated the utilization profile of the healthcare services and the demand for attendance. This indicates that the program may be contributing towards greater equity under these conditions.
Keywords: Health services, utilization. Family health. Morbidity. Equity. Data collection. Delivery of health care. Health services coverage. Health services research. Population survey.
The Family Health Program (FHP) is a strategy for reorganizing the healthcare attendance system within the Brazilian National Health System (Sistema Único de Saúde - SUS). It has been described as a tool for promoting equity in the offer of services.5 This program started to be disseminated at a national level in 1996. It consists of a method of primary attendance that is implemented, as a priority, for populations at high risk, both from the biological and socioeconomic points of view. The proposition behind the FHP is characterized by family-centered actions, so as to seek integration with the community that it forms part of; perform active searches for cases, for early and opportune intervention; and to give emphasis to prevention and education in health. It stands for extension of coverage and facilitation of access, continuity of healthcare actions, multiprofessional teamwork and increased powers to solve problems.8 The FHP in the municipality of São Paulo began with the Qualis program in Itaquera, in 1996, and São Lucas Park, Sapopemba and the Nova Cachoeirinha district, in 1997. It was differentiated from the national program in how it was managed, in establishing working agreements with social organizations, and in incorporating other attendance resources for supporting family healthcare teams, such as oral health, mental health and specialized outpatient services.10
There is a pressing need to verify what possibilities the program has for producing an impact in a large metropolis like São Paulo. The impact can be measured through the changes in the utilization profile of the healthcare services among the populations that it covers, and by analyzing whether these changes converge with the objectives of the program. The need for verification comes firstly because it has to be taken into consideration that the populations on the periphery of São Paulo may have a variety of other resources available for healthcare.11 In this municipality, there is a high degree of development of healthcare attendance, with great complexity of the network of services present. The FHP is therefore not the only alternative in the basket of options that these populations have. Secondly, verification is needed because the impact of the FHP has to be distinguished from the trend towards expanding the utilization of services that has been described for the period 1989-1997.12 This effect was the result from the implementation of SUS, during a period in which the actions followed the traditional model and the FHP was just beginning. During this period, a small reduction in inequality of service provision could be seen.
Several questions need to be considered when evaluating a given utilization profile for healthcare services. The type of service sought, chosen from within a range that includes basic healthcare units, urgent services, emergency services, specialized outpatient services or clinics, and public or private hospitals, depends on its proximity and the access offered, as well as the social and cultural values associated with such consumption. The reasons for utilization and its frequency, the seeking of attendance at the right moment or late, and the preventive, attendance, laboratory or therapeutic procedures involved are also important aspects to be analyzed.14 The dynamics established between the clientele and the service, or between demand and offer, are reflected in the degree of problem-solving powers within the healthcare service and the degree of user satisfaction.
The literature relates the intensity and manner of utilizing healthcare services to the level of quality of life and the level of individuals' knowledge of healthcare and the network of services. Individuals' capacity to self-assess their state of health and also their expectations and needs for attendance, their self-care practices, the existence of alternative networks and the relative degree of autonomy that these have also collaborate in this determination.4 Prominent among the factors associated with the utilization of healthcare services are sex, age, socioeconomic conditions and the morbid condition. In general, healthcare services are utilized more by women, children and the elderly, the more wealthy and more educated, those who have healthcare plans, and those who present morbid conditions.9
Equity in healthcare, according to Braveman,1 consists of the minimization of existing disparities between groups that present differing degrees of social privilege. This implies that need must be considered in healthcare attendance. Part of this inequality in the utilization of healthcare services may be a representation of iniquity, especially in relation to socioeconomic conditions, that does not correspond to inequality in the need for healthcare attendance.13 Culyier3 made a distinction between two forms of equity: horizontal, which signifies equal attention to equal needs; and vertical, which signifies unequal attention for different needs.
By comparing two samples, one representing a population covered by the FHP and the other not covered by it, the present study had the objective of analyzing differences in the degree of service utilization (preventive or curative) and in the demand for attendance among those who report some episode of illness. The study also had the objective of analyzing the distribution of service utilization and demand for attendance, according to some indicators of socioeconomic conditions, and sex, age and morbidity (or the respective limitations of activities). For this, it was decided to perform a population-based health survey, since this is one of the best ways for making health diagnoses. This method has the capacity to reveal a more complete picture than an investigation built up from the records of services and institutions.2
This study formed part of a larger project with the name "Study of morbidity and utilization of healthcare services among the population covered by the Qualis project".* It was also a component of the national multicenter project "New healthcare attendance models: Evaluation of the healthcare program (FHP) in the municipality of São Paulo" which was developed between 1998 and 2002.**
This project was evaluated and approved by the Ethics Committee of Hospital das Clínicas, School of Medicine of the University of São Paulo.
This was a cross-sectional study, for which the data collection was done in two districts attended by the Qualis project (Nova Cachoeirinha and Curuçá districts) between January and March 2001. It covered a total of approximately 190,697 inhabitants. Sampling by clusters was performed. The population analyzed consisted of residents of the Curuçá and Nova Cachoeirinha districts, divided into four strata:
a) 33,949 residents of the part of the Nova Cachoeirinha district that was within the coverage area of the Qualis project units of the Galvão and Ilza Hutzler districts;
b) 76,638 residents of the Cachoeirinha district, outside of the area of coverage of the Qualis project;
c) 27,589 residents of the part of the Curuçá district that was within the coverage area of the Qualis project units of Silva Teles and Santa Rita;
d) 52,521 residents of the Curuçá district, outside of the area of coverage of the Qualis project.
For each stratum, the census sector and the home were adopted as the primary and secondary sampling units, respectively. All those living in the home that was drawn were included in the sample and were interviewed. For the strata b and d, 30 census sectors were drawn, with a probability proportional to the size of the district and, within these, 10 homes were drawn. For the strata a and c, the areas were divided into micro-areas that corresponded to the geographical spaces covered by a single community healthcare agent. For each micro-area (out of a total of 30), 10 homes were drawn. This procedure allowed 300 families to be identified in each stratum.
The present study analyzed the variables sex, age, income, schooling, work situation, pension coverage, healthcare service utilization over the 15 days prior to the interview, demand for attendance among those who reported they had had an episode of illness during the 15 days prior to the interview, and limitations on activities resulting from this episode (only the earliest episode mentioned was analyzed).
For the statistical analysis, the variables were grouped into the categories described in the Tables. The schooling level attained by the mother or guardian was attributed to children aged less than 15 years. The income classification was based on the minimum salary in force at that time (R$151.00). For the work situation and pension coverage, the classification was based on the information relating to the head of the family.
The data from the questionnaires was recorded in Epi-6 and transferred to Stata 7, for the statistical analysis appropriate for clusters.
Proportions appropriate for clusters were estimated, and Pearson's chi-squared test was utilized with correction using Satterwaite's approximation and transformation into F distribution, as recommended in the Stata 7 program. Through this, it was sought to analyze the differences in the distribution of the "factor" variables, or in other words, the sociodemographic and morbidity variables reported in the two samples. This same method was utilized in univariate analysis of the distribution of the "outcome" variables of service utilization or demand for attendance, according to the sociodemographic variables and reported morbid conditions reported, for each sample. Multivariate analysis of the "outcome" variables in each sample, adjusted for the sociodemographic variables and morbid conditions, was performed by means of Poisson's regression for clusters.6 The prevalence ratio, 95% confidence intervals and significance level of the associations were estimated. In the multivariate analysis, the sample in question was first selected (area covered by the FHP, or not covered by it), and then the outcome variable being studied was put into the model along with the potential confounding variables (sex, age, reported morbid conditions and indicators of socioeconomic conditions that presented p<0.20 in the univariate analysis).
Table 1 describes the distribution of the variables investigated, in the two samples. It can be seen that there were no significant differences between the samples with regard to sex (p=0.64), morbidity/limitation on activities (p=0.69), or pension coverage (p=0.30). There were no significant differences between the samples in relation to schooling (p=0.05), income (p=0.07) or work situation (p=0.07). Nonetheless, it can be seen that the area covered by the FHP presents lower proportions of individuals with higher levels of education (5.1%), individuals with per capita family income of more than five minimum salaries (4.8%) and employed individuals (42.0%), and a higher proportion of inactive individuals (25.7%). These same proportions in the area not covered by the FHP were 9.2%, 10.9%, 50.6% and 17.8%, respectively. There was, however, a significant difference in the age distribution (p=0.03), such that the area covered by the FHP had a greater proportion of elderly people (10.4%) than did the area not covered (6.4%). It can also be seen that there was a very low loss of information, which was greatest in relation to income (216/3,901 =5.54%).
In Table 2, it can be seen that the distribution of service utilization in the area covered by the FHP presented significant differences according to sex (females =16.4% and males =10.2%; p<0.001), age (for example, 10-19 years =4.6% and over 60 years =22.9%; for all categories, p<0.001), morbidity (p<0.001), and work situation (for example, employed individuals =11.8% and inactive individuals =18.4%; for all categories, p=0.04). In the area covered by the FHP, no significant differences were observed in relation to schooling (p=0.32), income (p=0.93), or pension coverage (p=0.44). In the area not covered by the FHP, the distribution of service utilization also presented significant differences according to sex (females =17.1% and males =13.1%; p=0.01), age (for example, 20-29 years =8.1% and 50-59 years =23.5%; for all categories, p<0.001) and morbidity (p<0.001). However, no significant differences were observed in relation to work situation (p=0.25), schooling (p=0.26), income (p=0.09) or pension coverage (p=0.81). Comparing the totals for the areas, the service utilization was not significantly different: 15.2% in the area not covered and 13.5% in the area covered (p=0.25).
In Table 3, it can be seen that the demand for attendance in the area covered by the FHP presented significant differences in relation to the degree of limitation on activities resulting from morbidity (for example, 49.75% for slight limitation and 72.76% for severe limitation; for all categories, p=0.04). In the area covered by the FHP, there were no significant differences in the distribution of the demand for attendance, in relation to sex (p=0.12), age (p=0.42), schooling (p=0.19), income (p=0.79), work situation (p=0.65) or pension coverage (p=0.48). In the area not covered by the FHP, significant differences were seen in the distribution of the demand for attendance with regard to schooling (for example, none =41.9% and college-level =85.2%; for all categories, p=0.02) and work situation (for example, inactive individuals =53.0% and informal/self-employed =77.6%; for all categories, p=0.03). On the other hand, no significant differences were seen regarding sex (p=0.13), age (p=0.10), degree of limitation (p=0.76), income (p=0.97) or pension coverage (p=0.61). Comparing the total for the areas, the demand for attendance was significantly different: 56.8% in the area covered and 69.4% in the area not covered (p=0.02).
Tables 4 and 5 describe the results from the multivariate analysis, i.e. associations between the variables of outcome and each factor investigated, after adjustment for potential confounding variables, within each area studied.
In Table 4, it can be seen in relation to service utilization that, in the area covered by the FHP, there were significantly lower prevalence ratios for the male sex (PR=0.77) and for the age range of 10 to 19 years (PR=0.36), and significantly greater prevalence ratios for those who presented morbidity with some degree of limitation. No significantly different prevalence ratios regarding schooling, income, work situation and pension coverage were observed. In the area not covered by the FHP, there was also a lower prevalence ratio for the male sex (PR=0.80) and a greater prevalence ratio for morbidity with some degree of limitation. It was, however, noted that with regard to age there were lower prevalence ratios for the categories of 10-19 years (PR=0.58) and 20-29 years (PR=0.44). For schooling there were greater prevalence ratios for the categories of 5th to 8th grade (PR=2.31), high school (PR=2.53) and college education (PR=2.99), with a significant linear trend (p<0.001). For income, there were greater prevalence ratios for the categories of 2-4.99 minimum salaries (PR=1.52) and >5 minimum salaries (PR=1.61), also with a significant linear trend (p=0.036).
In Table 5, with regard to the demand for attendance among those who reported episodes of illness, it can be seen that, in the area covered by the FHP, there was a significantly lower prevalence ratio for the ages of 10-19 years (PR=0.64), and a greater prevalence ratio for a severe degree of limitation (PR=1.38). No significantly different prevalence ratios were observed for sex, schooling, income, work situation and pension coverage. In the area not covered by the FHP, significantly lower prevalence ratios were observed for the ages of 10-19 years (PR=0.71), 20-29 years (PR=0.55), 30-39 years (PR=0.77) and 40-49 years (PR=0.75), and also for inactive individuals (PR=0.69). Higher prevalence ratios were observed for schooling, in the categories of 5th to 8th grade (PR=1.90), high school (PR=2.01) and college education (PR=2.38). These latter categories also presented a significant linear trend (p=0.001).
Various cross-sectional analyses are possible in examining the question of access to and utilization of healthcare services. The present study gave priority to two fundamental aspects of this: the utilization of services and demand for attendance among those who reported some episode of illness. Both aspects related to the 15 days preceding the interview. The utilization of services was considered to be any preventive or curative motive or type of service provided that would translate the user-provider relationship in a broader sense. The demand for attendance, among the individuals who reported some episode of illness, brought into evidence how this relationship was reconciled with specific situations. In such cases, certain needs were established, thereby generating the seeking of healthcare attendance, in which there were more intense and specific expectations that needed to be attended to.
In the attendance model implemented through the FHP, the organizing of the attendance and the types of professionals differed from those in the "traditional" basic attendance. This could modify the utilization patterns or demand for healthcare services. The actions of community healthcare agents, in particular, in forming the link between the home and the healthcare services, may contribute towards lower demand for services, without this necessarily signifying decreased access or lack of attendance. This is because the needs are possibly being attended to through the agent's intermediation, without the individual having to go to the healthcare unit.10
The fact that the samples from the areas covered and not covered by the FHP were very similar regarding sociodemographic variables, with the sole exception of the age distribution, has very important significance. The choice of the Curuçá and Nova Cachoeirinha districts for the original implementation of the FHP was made without significant sociodemographic bias. The areas studied reflect the general characteristics of the districts that they form part of, which occupy the 18th and 19th places in the ranking of Human Development Indices for the 31 administrative subdivisions of the city of São Paulo.7 The sampling methodology for the present study succeeded in obtaining two samples with a high degree of comparability. Thus, the internal validity (comparison between samples) and external validity (generalization to the population studied) are strengthened. Other positive points of the study were the low loss of information and the successful treatment of confounding effects through the multivariate analysis.
Although the intensity of service utilization reported did not differ between the samples, its distribution according to the sociodemographic variables presented important differences. According to age, while the area not covered by the FHP presented greater utilization among the youngest and oldest among the population, the area covered by the FHP presented lower utilization only for the 10-19 age group (i.e. adolescents). According to schooling and income, while the area not covered by the FHP showed growing utilization among the more privileged categories, the area covered by the FHP did not present any differences between the categories. These results may indicate that one of the impacts of the program could be service utilization that is less unequal and less dependent on socioeconomic conditions, in the area covered by the FHP.
With regard to the demand for attendance among individuals who reported some episode of illness, two matters need to be considered: the proportion of the demand was greater in the area not covered by the FHP, and the demand distribution according to socioeconomic variables differed between the two populations. With the presupposition that there is no repressed demand, it can be supposed that the lower degree of demand may have occurred because of the support from the PSF team, and in particular because of the actions of the community healthcare agents. This is coherent with the observation that, in the area covered by the FHP, the demand was greater among those who reported a severe degree of limitation, a difference that did not occur in the area not covered. Furthermore, the pattern of greater demand was repeated according to age in the area not covered, with greater demand among the youngest and oldest parts of the population, while in the area covered the demand was only lower among adolescents.
According to the indicators of socioeconomic conditions, there was growing demand from the categories with more privileged schooling in the area not covered, while such differences did not exist in the area covered. There were also differences according to work situation, with lower demand from inactive individuals in the area not covered, which did not occur in the area covered. Once again, the results seem to indicate that one of the impacts of the program has been a distribution of healthcare attendance that has lower dependence on the socioeconomic conditions of the area covered and, importantly, higher dependence on the degree of need (morbidity).
The results from the present study have confirmed the inequalities in the utilization of healthcare services described in the literature in relation to sex, age, schooling and income, and in the demand for attendance according to age, schooling and work situation. Nevertheless, these inequalities are smaller in the area covered by the FHP, which demonstrates coherence with the objective of program relating to the promotion of horizontal equity. Inequality in the utilization of services in the situation of morbidity reported within the last 15 days was confirmed, with greater inequality in the demand for services in the area covered and in accordance with the degree of severity. This is coherent with the objective relating to vertical equity.
These findings indicate that a certain level of access to and utilization of healthcare services among the population living in the poorer areas of the municipality of São Paulo does exist, even in the areas not covered by the FHP. However, in such areas, the presence of the FHP has been shown to be capable of diminishing the effect of the unequal social conditions (as measured via the variables selected), on the access and utilization profile that exists there, with improvement in the social equity in this respect. The present study was not aimed at measuring the social iniquity by reference to an ideal service consumption pattern, or a pattern that would be needed for there to be an impact on healthcare conditions. Rather, it has sought to identify the difference that the FHP, when effectively implemented (as in the cases studied here), may make to the access and utilization profile of healthcare services among a population with these living conditions. This type of study is of great importance for better comprehension of the results and impacts that are possible through the implementation of the FHP, which may be distinct, depending on the living conditions and access to healthcare service among the target population.
To Dr. Maria Cecília Goi Porto Alves, of the Endemic Disease Control Superintendency (SUCEN), for her assistance in the sampling process.
1. Braveman P. Monitoring equity in health: a policy-oriented approach in low- and middle-income countries. Geneva; WHO; 1998. (Equity Initiative Document, 3). [ Links ]
2. Cesar CLG, Figueiredo GM, Westphal MF, Cardoso MRA, Costa MZ, Gattás VL. Morbidade referida e utilização de serviços de saúde em localidades urbanas brasileiras: metodologia. Rev Saúde Pública 1996;30:153-60. [ Links ]
3. Culyier AJ. Health, health expenditures, and equity. In: Van Doorslaer E, Wagstaff A, Rutten F. Equity in finance and delivery of health care: an international perspective. Oxford: Oxford University Press; 1993. [ Links ]
5. Marques RM, Mendes A. Atenção Básica e Programa de Saúde da Família (PSF): novos rumos para a política de saúde e seu financiamento. Ciênc Saúde Coletiva 2003;8:403-15. [ Links ]
6. Mendoza-Sassi R, Béria JU, Barros AJD. Fatores associados à utilização de serviços ambulatoriais: estudo de base populacional. Rev Saúde Pública 2003;37:372-8. [ Links ]
7. Pedroso M. Desenvolvimento humano no município de São Paulo [dissertação de mestrado]. São Paulo: PUC/SP; 2003. [ Links ]
8. Santana ML, Carmagnani MI. Programa de saúde da família no Brasil: um enfoque sobre seus pressupostos básicos, operacionalização e vantagens. Saúde Soc 2001;10:33-54. [ Links ]
9. Sawyer DO, Leite IC, Alexandrino R. Perfis de utilização de serviços de saúde no Brasil. Ciênc Saúde Coletiva 2002;7:757-76. [ Links ]
10. Silva JA, Dalmaso ASW. Agentes comunitários de saúde: o ser, o saber, o fazer. Rio de Janeiro: Fiocruz; 2002. [ Links ]
11. Silva NN, Pedroso GC, Puccini RF, Furlani WJ. Desigualdades sociais e uso de serviços de saúde: evidências de análise estratificada. Rev Saúde Pública 2000;34:44-9. [ Links ]
12 Travassos C, Viacava F, Fernandes C, Almeida CM. Desigualdades geográficas e sociais na utilização de serviços de saúde no Brasil. Ciênc Saúde Coletiva 2000;5:133-49. [ Links ]
13. Van Doorslaer E, Wagstaff A, Rutten F. Equity in finance and delivery of health care: an international perspective. Oxford: Oxford University Press; 1993. [ Links ]
14. Zucchi P, Del Nero C, Malik AM. Gastos em saúde: os fatores que agem na demanda e na oferta dos serviços de saúde. Saúde Soc 2000;9:127-50. [ Links ]
Av. Dr. Arnaldo, 455 2º andar
01246-903 São Paulo, SP, Brasil
Received on 28/5/2004.
Approved on 30/7/2004.
Funded by Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp - Grant n. 1998/15246-3), and by the STD/AIDS/MS project (Grant n. 635/99).
Coordinated by the Department of Preventive Medicine of the School of Medicine,
University of São Paulo.
** Jatene AD, Malik AM, Goldbaum M, Novaes HMD, Silva JA, Giffoni RM. Novos modelos de assistência à saúde: avaliação do Programa de Saúde da Família/Qualis, no Município de São Paulo. São Paulo; 2002. Technical-scientific report presented to Fundação de Amparo à Pesquisa do Estado de São Paulo.