Hospitalization flow in the public and private systems in the state of Sao Paulo, Brazil

OBJECTIVE To describe the migration flows of demand for public and private hospital care among the health regions of the state of Sao Paulo, Brazil. METHODS Study based on a database of hospitalizations in the public and private systems of the state of Sao Paulo, Southeastern Brazil, in 2006. We analyzed data from 17 health regions of the state, considering people hospitalized in their own health region and those who migrated outwards (emigration) or came from other regions (immigration). The index of migration effectiveness of patients from both systems was estimated. The coverage (hospitalization coefficient) was analyzed in relation to the number of inpatient beds per population and the indexes of migration effectiveness. RESULTS The index of migration effectiveness applied to the hospital care demand flow allowed characterizing health regions with flow balance, with high emigration of public and private patients, and with high attraction of public and private patients. CONCLUSIONS There are differences in hospital care access and opportunities among health regions in the state of Sao Paulo, Brazil.

According to Viana (2011), 10 social and health policies in Brazil can be divided into three periods. The period of liberal institutionalism (1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002) is characterized by an emphasis on market deregulation and self-regulation with the creation of the National Regulatory Agency for Private Health Insurance and Plans (ANS) and of the Brazilian Health Surveillance Agency (ANVISA), in addition to the key strategy of decentralization with emphasis on primary health care. The transition phase (2003)(2004)(2005)(2006), with positive results in both its balances of trade and payments, developed the emblematic program Bolsa Família, a direct money transfer from the state to people in poverty. The neodevelopmentalist institutionalism phase (2007-2010) is characterized by the emergence of the new middle class, which concentrates great purchasing power, and by regionalization in health, combined with the strong expansion of investments geared towards the construction of health facilities -clinics and hospitals -as well as the strengthening of the health care economic and industrial complex.
Studies on the social and economic development have been focusing on the role of the health care sector, which, in addition to the services available, mobilizes the industrial base: chemical, biotechnology and mechanical, electronics and materials. The service sector weighs the most in the health care economic and industrial complex. 2 Its organization follows spatial organization and configures hubs and regions.
Brazil has only recently started to plan health care in regional levels. Although stated in the 1988 constitutional prescription of the Brazilian Unified Health System (SUS), it advanced only when the Normas Operacionais de Assistência à Saúde (NOAS -Operational Health Assistance Regulation) were issued, in 2001 and 2002. The 2002 NOAS spurred the creation of regionalization master plans and the organization of health care networks on regional bases according to the needs of the population, defined by epidemiological parameters -and not by the existence and supply of services.  6 (2008) studied SUS health care in the state of Sao Paulo, focusing on the regional perspective in the context of the design of the State Health Plan. We used data from the Sistema de Informação Hospitalar (SIH -Hospital Information System) from 2000 and 2006 for hospital care, and from ANS to estimate the population covered by health plans. In that regard, they state: "Health managers are at an impasse: to consider or not the coverage of health plans in the SUS planning process? With the growth of the covered population and of the disparity of coverage among regions, particularly among municipalities, not considering this factor may contribute to the maintenance of inequalities in the access currently identified in the system".
"In the state of Sao Paulo, SUS beds registered in the Cadastro Nacional de Estabelecimentos de Saúde (CNES -National Registry of Health Facilities) are mostly philanthropic (46.9% of the beds); 40.2% are public and 12.9% are private beds, under contract with SUS. This feature requires great care in the policy of partnership with the philanthropic hospitals, for fear of jeopardizing care for SUS users, in some regions where there is a higher dependence of these hospital beds".
Regionalization is questioned by Machado. 4 Creating health care networks would guarantee fair and comprehensive care to all Brazilian citizens -assuming the articulation between municipalities that "export" and those that "import" patients, under the coordination of the state governments. However, the willingness for cooperation would be far from becoming a reality. Venâncio et al 9 examined management practices of regional referral in five regions of the state of Sao Paulo, assessing indicators, instruments used and the perception of regional and municipal managers. The formal mechanisms of regional referral were considered insufficient, as well as the instruments for their follow-up.
The regionalization of health has been a concern of health officials for many years, 1 about the articulation between patient exporter and importer municipalities. The presumed flow logic is that the larger the emigration of patients, the lower the ability to provide care for the population locally or regionally. On  hand, the higher the immigration of patients, the greater the attraction power exerted by the conditions of care supply. Each region has a hub municipality that concentrates the resources of secondary and tertiary-level health care and would receive the full flow of its referral area. In addition, there are diagnostic and treatment resources of intermediate level in satellite cities of the regional hub. Municipalities and importer regions are also, to some extent, exporters. The predominance of one or another type of flow must be clarified. To this end, it is necessary to consider the size of the referral population, the intensity of the flow in either direction, and the sum of both flows in relation to the total demand in the period.
Studying the hospitalization flow in public and private services can contribute to the knowledge of health needs and demands of the population on regionalized bases, and the adequacy of the structure of the regions and comparisons among them. An objective instrument for evaluating the intensity and direction of the flows is important to the regional structure of health care. The demand for diagnostic and treatment resources outside the place of residence allows inferring the degree of resolution of health care according to the existing care modules. Although the universality principle of SUS establishes the right to health care in any service of the system, it is important to know the origin of patients (municipality of residence) to plan assistance to local and regional demands. Concepts such as import and export, outward or inward referral and patient immigration or emigration are used in the literature with some impropriety.
This study aimed to describe the flow of demand for public and private hospital care among health regions in the state of Sao Paulo, Brazil.

METHODS
An exploratory study, based on a cross-section of the database built by Moreira d containing the records of all hospitalizations in Brazil by the public and private (health plans and paid by the patient) systems in 2006. Moreira's study aimed to analyze rehospitalizations in Brazil with data from SIH/SUS and the Comunicação de Internação Hospitalar (CIH -Communication of Hospitalizations) system by DATASUS. A database was created by a sophisticated process of record linkage (Reclink) and linkage with the variables name, sex, municipality of residence, and date of birth by means of pre-defined and tested algorithms. As in SIH, each event (hospitalization) may correspond to a partial billing, referring to part of the period or some of the procedures -a situation also accepted in the CIH related to the period -, three algorithms were set to identify the events the same hospitalization and to compose the record of its data. Thus, it was possible to identify the hospitalizations paid by SUS and those paid by the private system as well as their repetitions informed to the Brazilian Ministry of Health by SIH and CIH, respectively. 5 For this study, we selected all hospitalizations financed by SUS and by the private system of people residing in the state of Sao Paulo, Southeastern Brazil, including the ones in other states of the country, from January 1 to December 31, 2006. The hospitalizations were classified according to the municipality of residence and their location in one of the 17 health regions of the state. The estimated population, the number of beds and beds per 1,000 population of each Departamento Regional de Saúde (DRS -Regional Health Department), in 2006, were obtained from the State Health Plan. 8 Hospitalizations were classified according to each DRS (total hospitalizations) and registered as hospitalizations of patients residing in the same region or not, according to the patient's city of origin. This enabled estimating the number and percentage of patients assisted in their own region; the total number of hospitalizations in each region (column C, Tables 1 and 2) minus the total number of people hospitalized in their region of origin (column E, Tables 1 and 2), which provides the number of immigration cases, i.e., the number of cases coming from other regions; the total number of hospitalized residents (column D, Tables 1 and 2), minus the total number of people hospitalized in their region of origin (column E, Tables 1 and 2), provides the number of cases of emigration, in which the hospitalization occurred outside the region of the patient. The total number of people hospitalized in each DRS, in relation to the resident population per 1,000 inhabitants, shows the coefficient of total hospitalization of each region of the state.
The intensity of the immigration and emigration of patients was estimated by the index of migration effectiveness (ME), used by Caiado. 2 This indicator allows quantifying the intensity of patient inflow and outflow and the predominant direction. It is estimated by the net migration rate (immigration less emigration) and the crude migration (immigration plus emigration).

ME = (I -E) / (E + I)
Values close to 1 indicate a strong migratory attraction, close to -1 indicate areas of high emigration, and close to 0, areas with high migratory circulation.
We estimated correlation coefficients between the number of hospital beds and the coefficient of hospitalizations per 1,000 population. We also analyzed the index of migration effectiveness of immigration and emigration o patients from the public (SUS) and private (non-SUS) systems in each Regional Health Department.  Table 3). The hub cities have just over 15 million inhabitants, which means that most of the regions have a greater population in the satellite referral area than residing in the hubs -almost 10 times greater in Taubate and Sao Joao da Boa Vista and a little less than the double in the Greater Sao Paulo. The resulting services demand flows represented the dynamics of regionalization in the state.
The global rate of beds per population in the state was low (1.53 per 1,000 population), with wide variation between regions ( Table 3). The Regional Health Departments that presented the highest percentage of beds per inhabitants were Barretos (2.

DISCUSSION
The municipalities of the state of Sao Paulo had major differences regarding socioeconomic development and population density in 2009. 7 Of them, 11.3% had over 100,000 inhabitants and all were classified as high-wealth in the Índice Paulista de Responsabilidade Social (Sao Paulo Index of Social Responsibility); 61.7% had less than 20,000 inhabitants and 70.4% were classified as of low wealth. Although only 73 municipalities were included among the large ones, they had nearly 75.0% of the state population. This illustrates the great dependence of the smallest municipalities on the largest ones and, therefore, the importance of the issue of regionalization of health in the state of Sao Paulo. Larger, wealthier municipalities become hubs and attract the population of smaller, poorer municipalities. Thus, regionalization is the instrument to try to compensate for inequality and provide equal opportunities for disadvantaged populations.