Regionalization of public hospital care in the metropolitan area of Vitória , Espírito Santo , Brazil

This article analyzes the process of formation of the regional health care network in the metropolitan area of Vitória, Espírito Santo, Brazil (RMGV-ES), with specific focus on public hospital care, trying to characterize the current stage of structuring inter-hospital network in the State of Espírito Santo (Brazil), particularly in the Grande Vitória area. Hospitalizations were analyzed performed in hospitals of the Brazilian National Health System and the displacements of the users of these services among the municipalities that make up the RMGV. The methods adopted were the processing and analysis of data from the Sistema de Internações Hospitalares do Sistema Único de Saúde (SIH-SUS), systematized by the Departamento de Informática do Sistema Único de Saúde – Datasus. The RMGV recorded significant population flows in the use of public health services, as shown by the data collected. The largest shifts were observed between men, children and adolescents and for surgical admissions. Considering regionalization as a new principle and guideline of the health decentralization process, this article has identified changes in the process that previously pointed to the municipalization and now seeks to strengthen a regional and intermunicipal perspective for assistance to the population.


Foreword
In view of the recent dynamic for formatting networks in health regional attention, which has been established throughout Brazil in line with the Health Management Pact (Ministry of Health) guidelines, this study analyzes the regionalization of public hospital care in the Metropolitan Region of Vitória-ES (RMGV-ES).The survey defined as analytical objective, to scale the current stage of structuring that intercity network in the state of Espírito Santo, particularly in close proximity to Vitória -(Grande Vitória). 1he cut-set was the hospital care in the provider network of the Brazilian National Health System (SUS) services, by analyzing data on hospital admissions carried out and especially the displacement made by the users among the counties which make up the RMGV.The research took as presupposition the exhaustion of municipalization model of public health policies (Barreto Jr., 2008;Dowbor, 2009), a process that has led to the formation of regional service networks.The new format of such a public policy has required the regulation of inter-municipal agreements, to negotiate the financial transfers which are made from the Ministry of Health to the cities and the accommodation of demands for health services which go beyond the county's territory.These agreements formed the core of decentralization and regionalization of public health policies in Brazil and made the central theme in the agenda of a Brazilian state reform, linked from the 1980s and in the following, due to two factors.On the one hand, the characteristic centralization of bureaucratic authoritarianism of the post-1964 military regime converted the decentralization into a regulatory principle of changes to the democratic sectors; and on the other hand, decentralization has also become an important part for liberal thought of opposition to that regime.The theoretical debate on public and social policies in Brazil shows that between the 1990s and the 2000s, there was a shift in the main focus of this discussion: while in the 1990s, themes such as decentralization and reform of the state were predominant -through which one tried to establish a more efficient and effective state model for management and implementation of public policies -in the years 2000 and following, the question was, at least partially offset culminating in the attitude of public managers to prioritize the regionalization of services in the health agenda, as a way to organize the model of care (primary, secondary and tertiary) and ensure access to health services.
This rationality was included in the process of restructuring and reform of social policies in Brazil as an action that potentially increases the effectiveness of health policy and favors it from acting as a component of reducing socioeconomic inequalities policies.As a counter argument, the neoliberal rationality is taken to which state action works in reverse, producing inefficiency and patronage which burden, heavily, the citizen who sought any help.Several authors (Barreto Jr., 2008;Cohn, 2009aCohn, , 2009bCohn, , 2009c;;Behring, 2003) discuss the transformation process which was submitted to health policy that after the promulgation of the Constitution, still needed another decade of regulations (ordinances, Operational Standards of the Ministry of Health, among others) to acquire the contours currently presented.However, studies investigating how these changes countered and transformed the county health systems are scarce, particularly in the aspects related to the installation of hospital care networks, object of this article.The city is a potentially privileged lócus for the clash of political forces in the broad sense, given that, in its territory, there is a space to house the existing consensus and dissentions in society.But how is this logic reflected in the policies regionalization?With public management innovations opened with the adoption of health policies decentralization where the local sphere gains strength, powers and resources are expected, in conducting a policy relating to the interest of the whole community.Just as it is expected that the health gain importance on the regional agenda, not only of the health systems components management, but also, especially, in the clash of local political forces.
The process of municipalization and later regionalization, of such social policies placed the mayors in the center of decisions on the political model that would be offered to the population of each locality.Throughout the health policy process reform, municipalities were placed at the forefront of a situation that requires agile interventions that supersedes the old routines of urban cosmetics, as managing large infrastructure projects, social policies and the promotion of local economic activities.Even under the coordination of the Ministry of Health, the SUS assigns a relative autonomy to the municipal manager as to define which actions and health programs will be developed in the county, recommending the public control through municipal councils and, more recently, the regionalization of attention to health.Thus, the mayors have autonomy and make decisions based on their public commitments, and the same applies to the hues adopted in the public policies formulation.
This is what happened in RMGV where bottlenecks persist related to the population access to services and health conditions, as revealed by statistical data which will be presented later in this article.In the region in question, as in most Brazilian urban areas, the population does not present a homogeneous socioeconomic profile among its different cities and micro-regions.This heterogeneity is reflected in the existence of poverty and increased vulnerability to risk areas, caused by the socioeconomic conditions, unequal access to health services, low levels of education and income.This state of affairs is relevant to public interventions agenda that seek to mitigate these differences and ensure minimum conditions of health and quality of life for all population segments and areas of such region, according to the principles of the Brazilian National Health System.
As it will be seen throughout this article, the Metropolitan Region of Grande Vitória, taking into account its complexity, population size and economic and political relevance in the state of Espírito Santo, witnessed major population flows to the use of public health services, as indicated by collected data.According to Barreto Jr., Ferreira e Silva (2008, p. 7), "access conditions reflect characteristics of services supply which facilitate or hinder the ability of individuals of a given population to use health services according to their needs ".Thus, the demand and use of services are conditioned by a number of factors that, according to Andersen (apud Travassos;Martins, 2004, p 191.): Can be grouped into: predisposing factors, enabling factors and health needs.In the first group, predisposing demand, variables are related to socio-demographic characteristics such as age and gender.In the second, consumption capacity or use coalesce factors for coverage of public and private services and its services network available.And in the third, the need is influenced by the existence of self-perception of health problems or pre-existing diagnoses which may lead the individual to seek health care.
Considering regionalization as a new principle and guideline of health decentralization process, this article will seek to identify the changes in such process that previously pointed out to decentralization, strengthening of regional and inter-municipal perspective for health care.

Materials and methods
The study was conducted by processing and analyzing data from the National Hospital Admissions System of the Brazilian National Health System (SIH-SUS), particularly the records generated from Hospital Admissions Authorizations (AIH), systematized by the Computing Department of the Brazilian National Health System (DATASUS).This system records data on all the hospital admissions financed by SUS, in public hospitals (federal, state and local) and private (for-profit or charitable purposes).
To operationalize the data processing, the computer program TaBWin was installed locally, developed by the Ministry of Health, free and open for dissemination, allowing broad autonomy for operation of databanks and making various intersections, which are not enforceable by the internet.To enable the use of TaBWin, it was downloaded from Datasus2 website, from the reduced files of AIH belonging to the state of Espírito Santo, in the period from 2011 to 2013.A cohort study of this period was made, in order to achieve a significant quantity of such hospital admissions that supports the required breakdowns for the study.The AIH reduced files are monthly and bring complete information on the physical network (originating from the National Health Facilities Register -CNES) and on production (hospital admissions SIH-SUS), carried out under the Brazilian National Health System.The study does not cover, therefore, admissions made in private hospitals which are not contracted by the system, for which there is no available data.
In data processing, we selected the number of hospital admissions (excluding commitments for extension of such admissions), crossing them between the counties' hospital admissions performance against the patient's home, which was feasible using the TabWin software.Thus, it was possible to map the flow of inter-municipal hospital admissions and quantify: i) the attraction that such counties have on residents of other cities; and ii) the movements between cities required to effect such hospital admissions.For a more accurate analysis of these movements, the following attributes were selected: age, gender and complexity of hospital admission.

Outcomes and discussion
The Metropolitan Region of Grande Vitória (RMGV), in the state of Espírito Santo, is comprised of 1,884,096 inhabitants3 in the counties of Cariacica (378 915 inhabitants), Fundão (19,585), Guarapari (118,056), Serra (476 428), Viana (73 318) , Vila Velha (465 690) and Vitória (352 104).Given its complexity, population size and economic and political relevance in the state, such intercity flows of hospital admissions showed a typical dynamics of large settlements.In a temporal cohort of three years which was observed in this study (2011)(2012)(2013), the hospitals located in RMGV held 298,963 admissions, none of them in the cities of Fundão and Viana, in which there were no hospitals associated with the Brazilian National Health System.Admissions are so distributed, according to Thus, it is observed that more than 92 thousand inhabitants, statement of populations of Fundão and Viana, do not have any hospital special agreement with SUS within their territory and the service always occur in extra-municipal limits, which needs to be organized by primary care and the secondary and tertiary reference system.When compared to population the distribution and of hospital admissions, it stands to capital Vitória, which concentrates 18.7% of the population and over 53% of hospital admissions.This appeal, as discussed later, is due to the presence of the Hospital das Clínicas da Universidade Federal do Espírito Santo (UFES) in the city, a university teaching establishment managed by the Federal Government, aimed at regional and specialized care of a significant portion of the population of the Espirito Santo .The cities of Serra, Cariacica and Guarapari focus hospital admissions in lower proportion to population.
The data in Table 2 allow us to analyze the distribution of hospital admissions according to the nature of the service provider, network management important indicator of health care.The state and philanthropic providers were responsible for the largest shares of hospital admissions in the RMGV, and one should also note that almost 10% of admissions were made by the federal service.The philanthropic network is also the largest provider of hospital services in Brazil and in the Southeast, where it is responsible for most of the halves of SUS hospital care, which reiterates the importance of Hospitals and charitable hospitals in the public system.The proportions of such hospital admissions, carried out in state and federal hospitals in the region have higher levels to those observed for the total of Brazil (22.1% and 4%, respectively) and the Southeast (20.9% and 3.7% , respectively), 4 suggesting some atypical features of RMGV, opposite the process of health care decentralization, triggered from the decade of 1990.The significant participation of the federal service in the region is due to the Hospital das Clínicas of the UFES in Vitória, a university teaching hospital, which offers general and specialized care, and turns into the main regional state health service of Espírito Santo.
As the displacement of residents to obtain hospital admissions outside their county of residence, main object hereof, the Table 3 allows us to identify, from the cross between the county hospital admission data, city of residence, the population proportion that did not get hospital care in their own city, needing to travel between the cities of RMGV.Vitória and Vila Velha were the ones that had the highest proportions of citizens' care within the limits of their own territory, especially the first, capital of the state of the Espírito Santo.These are the richest towns in the region, concentrating public and private health services, education and culture, among others.Moreover, according to data from SUS, Vitória was the city that attracted the highest proportion of residents for hospital admission of all RMGV, followed by Vila Velha.Even in the case of these cities, in the three years 2011-2013, 48,480 admissions of residents located outside the Vila Velha metropolitan area were made, certainly by regional and specialized nature of their SUS hospitals.It has been pointed out earlier that the cities of Fundão and Viana have no hospitals providing services at SUS.Among the other cities, the following are highlighted: among the residents of Cariacica, almost 80% were admitted to Vitória and Vila Velha.Two thirds of the population of Serra, who required hospital admis-sion, was met outside the city, a situation that occurs to almost half of the residents of Guarapari.

Table 3 -Hospital Admissions performed by SUS, by county of admission according to the county of residence -Metropolitan region of Vitória -ES (2011-2013)
This analytical approach, which aims to understand the formation of the regional network of hospital care in the RMGV, should also reflect the data in Table 4, allowing us to know two important indicators of this dynamics.The scale shows the evasion and invasion rates of hospital admissions.Conceptually, the expressed evasion rate, for a given municipality, what proportion of its residents who were treated at hospitals located outside its ter-ritory.This indicator is intended to examine the need for intercity travel to obtain hospital admissions and is of great importance in the planning of health care.In line with the data analyzed above, the highest evasion rates were found in Viana and (1) Refers to the number of hospital admissions for residents of the reference city, carried out in hospitals outside that city.
(2) Refers to the number of hospital admissions in the reference city, in patients living outside the county.* Referring exclusively to the admissions of 257.450RMGV inhabitants made inside RMGV hospitals.
Further in accordance with Table 4, the invasion rates are shown.This indicator expresses, as part of the hospital admissions made in a given municipality, the proportion of care offered to residents of other cities.It is an important benchmark in the SUS, as the transfers from the Ministry of Health for each of the municipalities are subjected to a quantity limit determined by the ministry itself, which will be paid per admissions a year.They are called AIH limits.Due to this limitation, the relevance of the ministerial regulations which in recent years promoted the renegotiation of these limits are shown, which in turn proposed the Agreed and Integrated Programming (PPI) and the Regionalization Director Plan. 5 The great reference centers are of course those that attract the largest population groups, which were found in the cities of Vitória and Vila Velha, which showed, respectively, 66.1% and 48.7% invasion rates.
In order to analyze more accurately the phenomenon of regional health care in the RMGV, tables 4, 5 and 6 show the evasion rates by gender, age and complexity of care (basic and surgical clinics).
The intercity travel for hospital care, with separate analysis for hospital admissions for men and women (whose data are presented in Table 4), are resulting from the provision of services in the municipalities, the different gender needs as health, and factors associated with local economic dynamics.The male admissions evasion rate was higher than women in all locations of RMGV.In the city of Cariacica, known as the RMGV commuter town, male evasion was significantly higher than the female, as shown in Table 4.The hospital network of Cariacica, convened with SUS, recorded movement in only two hospitals: a psychiatric care hospital called Hospital Adalto Botelho, which serves mainly residents of other localities; and the Hospital Meridional, responsible for general care, with hospital beds for medical and surgical care.Over 90% of men of Cariacica had hospital admissions in the cities of Vitória and Vila Velha when needed it.Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH/SUS).
(1) Refers to the number of hospital admissions of residents of the reference city, men and women, performed in hospitals outside the municipality.* Referring exclusively to the admissions of 257.450RMGV inhabitants made inside RMGV hospitals.
The observation of these aggregate data by age groups (such analysis defined the following groups: those under 14 years old; 15-49; 50 and over -see Table 5) further helps to understand the hospital evasion rate phenomenon.Except in the city-poles -Vila Velha and Vitória -the most significant evasion rates were recorded in hospital admissions of people up to 14 years old, all at levels above 70%.These percentages indicate the concentration of pediatric hospital beds and signal potential bottlenecks for children and adolescents admissions in RMGV.Quite significant rates were also checked for people above 50 years old.The demographic transition and the consequent population aging will certainly make this problem even more acute and will require a new planning of the general public policy and specifically in health.
Finally, the SUS inpatients evasion data aggregated from basic clinics (obstetrics, medical and pediatric clinics) and surgical greatly reveal and clarify the rates for understanding the logic of RMGV inter-city.The basic clinical evasion rate is significant in Cariacica (69.5%) and Serra (58.7%), but in all the municipalities of such region, it is lower than the evasion rate of surgical clinic.More than 90% of admissions to surgical clinic in RMGV occurred in the cities of Vitória and Vila Velha, which attests to the concentration of this type of hospital bed in the city-pole of the region.In Guarapari and Cariacica, evasion rates for surgical clinics were higher than 90%.
As the data indicate, the inter dynamic exerted impacts on local and intercity public health policies.It made that logics for services demand would end without notice of municipal boundaries, the lack of hospitals in all cities of RMGV or of basic / specialized service required.Common in large urban areas, the study of this phenomenon helps to understand the public policies managers to adopt mitigation measures of dynamics and which effectively ensure a public health services offer in the scope of SUS.

Final Considerations
About a decade ago, the signs of exhaustion of such municipalization model of public health policies had appeared.Recommended by the Federal Constitution of 1988, this model was important in the first SUS structuring decade.But the dynamics for care and the population's real needs, with regard to hospital care, forced the inclusion of the regionalization issues in the organization of the public policy agenda.According to Barreto Jr. (2005b), we observed that the process of regulating the Brazilian health system -between the enactment of the Federal Constitution in 1988 and the late 1990s -was strongly oriented towards decentralization policy through municipalization of actions and health services.Successful in principle, the process of decentralization began to show some limits and restrictions at the end of the decade, especially in what concerns the organization of attention of  (1) Refers to the number of hospital admissions of residents of the reference city, by type of clinic, performed in hospitals outside the municipality.
(2) The Basic Clinic includes Obstetrics, Internal Medicine and Pediatrics.
regional, intercity reference and high complexity services.Monika Dowbor also addresses aspects of the exhaustion of such municipalization model of health policies, emphasizing that: In summary, the implementation of SUS, which is processed primarily by the municipalization of primary care strongly marked by the adoption of federal programs of that level of attention, was conducted at the federal level by MS in the negotiation process with the intergovernmental negotiating body in which organizations of municipal and state health secretaries were present.[...] The bargaining game included the feasible design and financial incentives for adoption of basic care by municipalities who did, massively adopting federal programs that level of attention.If the demand for municipalization of SUS encompassed various actors of such sector, the adoption and widespread access to federal programs on primary care did not absorb much energy of the participants in the National Health Conference and the National Health Council.[...] The two conferences held in the 2000s (ninth and tenth) evaluated the Pasf and the Pacs as an alternative to primary care and detailed as to be more carefully with the recommendations, without, however, refute or endorse decisively strategy.The attention given by the National Board of Health to the programs is almost nonexistent when analyzing the main resolutions in the range from 1992 to 2007: the question of PSF only appears once, referring to the CNS support for implementation of the PSF and Pacs in the Federal District (CNS 2007).
For the traditional private sector, in turn, provider of hospital services, the discussion of primary care did not represent possibilities of action a priori.(Dowbor, 2009, p. 212-214).
Among the main aspects and innovations arising from the decentralization of public policies, the systems regionalization is emphasized as a hierarchical strategy of health services and the search for greater equity, which is based on configuration and functional systems of assistance to health, through the organization of state territories in regions / microregions and assistance modules; the conformation of such hierarchically services organized networks; the establishment of mechanisms and reference flows and county inter-counter, aiming to ensure comprehensive care and the population's access to health services and actions according to their needs.
What we found in this article was the formation of a network determined by the supply of health services concentrated in the city-poles of Vitória and Vila Velha, and not guided by a rationality that distributes hospitals in a more equitably way between the municipalities of RMGV.This hypothesis is supported by the fact that the larger displacement is concentrated in cities with large population and scarcity of public services.The logic of organization of health services does not necessarily require the existence of hospitals in all cities, provided that the regional references are accessible and clearly defined.But in the case of RMGV, there is a marked imbalance in the likely avoidance of hospital admissions for children, adolescents and people over 50 years old.For the first group, part of hospital admissions requires more sophisticated equipment, such as cancer treatments or transplants, and the offer could be better distributed.In the latter group, the transformation in morbidity profiles caused by population aging will require increased public availability of hospital beds to meet the expected increase in demand.And, in all situations, the primary health care structure should be organized to serve as the health system gateway, reference the service levels and mitigate the inter-city inequalities.

Table 1 -Hospital admissions carried out by SUS, according to the place of admissions and resident population -the metropolitan area of Vitória -ES (2011-2013)
Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH / SUS); IBGE.

Table 2 -Hospital admissions performed by SUS, by place of admission, according to the nature of service provider-the metropolitan area of Vitória -ES (2011-2013)
Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH / SUS).

Table 4 -Evasion Rate and Hospital Admission Invasion Rate performed by SUS, according to the county -me- tropolitan region of Vitória -ES (2011-2013)*
Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH/SUS).

Table 6 -Evasion Rate of Hospital Admissions performed by SUS, by Age Group, by municipality -the Metro- politan Area of Vitória -ES (2011-2013)*
Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH/SUS).(1)Refers to the number of hospital admissions of residents of the reference city, by aged group, performed in hospitals outside the municipality.* Referring exclusively to the admissions of 257.450RMGV inhabitants made inside RMGV hospitals.

Table 7 -Evasion Rate of Hospital Admissions performed by SUS, by type of clinic, according to the city -Me- tropolitan region of Vitória -ES (2011-2013)
Source: Ministry of Health / Computing Department of the Brazilian National Health System -Datasus.Hospital Information System SUS (SIH/SUS).