Regulation of access to hospital beds in emergency care and the development of integrated health services

Background: Comprehensive health care delivery is a major challenge for the Brazilian Unified Health System. Inadequate coordination among different health care levels and facilities produces poor access to health care and low degree of integration across levels of care, that is worse in emergency care delivery. Health policy has focused on health care networks development and regulation of access to health care provision. In the city of Rio de Janeiro, local health policy has prioritized the development of an Emergency Care Network. This research aimed at assessing the access to hospital beds and diagnostic resources, fundamental to continuity of care this network. Methods: Triangulation technique was employed, collecting of data from different sources and analysis resorting to multiple methods. Analysis of data from primary and secondary databases was performed. Interviews were conducted with managers and healthcare professionals involved in guaranteeing continuity of care in Emergency Care Network. Data analysis resorted to descriptive statistics and the interviews along with the data were interpreted under the conceptual framework of Integrated Health Care. Results: Poor access to hospital beds constituted a major obstacle to improvements in continuity of care for patients in Emergency Care Network. An insufficient number of hospital beds is the primary issue. Other aspects of hospital care contribute to aggravate this deficiency: the predominance of general hospitals, the existence of small and underequipped hospitals, prolonged lengths of stay and significant discrepancies of bed occupancy rates among specialized and general hospitals. Hospitals further face disrupted internal workflows due to inadequacy and limited availability of human and physical resources. Moreover, weak cooperation between federal, state and local government leads to persistence of a fragmented network with little or absent interaction among health care facilities and professionals from different government spheres. Conclusion: To reach a better degree of

have within its organization a group of workers appointed to oversee and control access to hospital's resources, called Núcleo Interno de Regulação (NIR), or Internal Regulatory Committee. These committees constitute the channel through which hospitals interact with other health facilities.
A triangulation technique was employed encompassing collection of data from different sources and analysis resorting to multiple methods, in order to fill in potential data gaps, to yield cross verification of data and to provide a more comprehensive understanding of a complex and little explored phenomena 30 .
Data was obtained by means of interviews and consultation of databases. Thirty-six interviews were held engaging managers and health care professionals. Among the managers there were one from federal administration, two state health managers and three from the municipal health department. Medical regulators, that is physicians performing regulatory activities, were interviewed during visits to the Centrais de Regulação, one at the State Access Regulation Center and three at the Municipal Access Regulation Center. Members of hospitals' Internal Regulatory Committee were interviewed in 5 different hospitals: two municipal hospitals, one with and one without an emergency department, a state and a federal hospital with emergency departments and a teaching hospital without an emergency department. Physicians engaged in regulatory activities were also interviewed at the UPA, which amount to 30 in the city of Rio de Janeiro. A total of 16 interviews were conducted, being half of them in municipal Emergency Care Units and the other half in Emergency Care Units managed by the state government. Five Regional Emergency Centers were visited and interviews with physicians involved in regulatory activities were performed.
Access to hospital beds was assessed through data obtained from the two systems used in CRMRJ to perform the allocation and distribution of hospital beds: the Sistema de Regulação (SISREG), or National Regulatory System, and the Plataforma SMS-Rio, or Municipal Regulatory System. Access to these systems was officially authorized by the municipal health department. Health facilities place hospital bed requests through these systems via online access. The request form filled in on SISREG website contains the patient's name, medical diagnosis, short clinical report and type of hospital bed requested. Upon the acceptance of the request for a hospital bed, health care facilities get a response through SISREG and e-mail. Medical regulators at CRMRJ work primarily on Plataforma SMS-Rio because of its more suitable resources to manage and sort out data from hospital bed request and municipal hospitals occupancy rates in order to provide the most efficient allocation. From SISREG it was possible to verify whether or not access to a hospital bed had been granted, consulting each form request at a time. Only from Plataforma SMS-Rio it was possible to obtain the total number of hospital beds to which access was granted. Unanswered requests placed on SISREG were not available at Plataforma SMS-Rio.
The data imported corresponded to the period from September 2016 to August 2017.
During this period, 56,907 hospital bed requests were placed on SISREG. During the same period, 10,750 hospital beds were regulated through the Plataforma SMS-Rio. A data analysis was carried out with the construction of the frequency distribution of the hospital bed requests according to several variables such as: type of the requesting health care unit (UPA, CER, hospital), administrative sphere of the requesting unit (federal, state, municipal), type of hospital bed requested (Internal Medicine ward, Surgery ward, ICU), type and administrative sphere of the health unit to which the hospital bed belonged.
From the imported data, two datasets were generated. A first dataset included all adult ICU bed requests placed on SISREG, either by a state UPA, a municipal UPA or a municipal CER. The second dataset included all the requests for admission to an Internal Medicine war bed placed by the same units. The UPA-CER ICU database encompassed 5,969 requests and the UPA-CER Internal Medicine dataset was comprised of 20,317 requests.
For the ICU dataset a random sampling resulted in 590 requests to be analyzed and for the Internal Medicine database the same procedure resulted in 1001 requests to analyze.
Each of these requests was consulted on SISREG through their request number in order to examine whether the outcome of the request had been access to a hospital bed or not. In order to estimate hospital performance two indicators were selected: mean length of stay and hospital bed occupancy rate. The data on length of stay was readily available at the database from the Health Informatics Department (DATASUS). Conversely, hospital bed occupancy rates were not available and had to be calculated. According to official norms 31 bed occupancy rate is calculated using information gathered locally at each hospital. Since this was not feasible, an alternative to the official formula was employed to reach an approximation. The rate of hospital bed utilization was calculated by means of the following formula: total hospital admission multiplied by the mean length of stay divided by the total hospital bed number multiplied by the total number of days, for a given period of time. The total number of hospital admissions was obtained through

Results And Discussion
Over the last decade state and municipal government have made substantial investments in the development and improvement of the Emergency Care Network in the city of Rio de Janeiro, with a strong emphasis on boosting regulatory mechanisms besides the expansion of Emergency Care Units 4 . The creation and reorganization of the Centrais de Regulação on municipal and state level, empowered with authority to decide over the allocation of resources was a central managerial decision. The number of managers and health professionals overseeing regulatory processes increased. Information technology was intensively used in order to facilitate and broaden the regulation of access to health care resources, like hospital beds and consultation with specialists. A transition from regulation based on personal relationships or contacts to a more institutionalized regulatory practice was firmly pursued, expecting to challenge the power hospitals have over their beds.
Much effort was employed to eliminate redundant and parallel procedures, to centralize work and information flows, to curtail personal interferences, to enhance expertise and to advance a paradigm shift towards cooperation and integration.
From the interviews and datasets analyzed the hospital sector and the local health system segmentation stood out as major issues that undermine the effectiveness of regulatory mechanisms and forestall the development of integrated health care services.

Hospital sector
Access to hospital beds emerged as a central issue for the adequate performance of the Emergency Care Network. In all interviews conducted this problem was highlighted as a major obstacle. Among the many problems elicited, an undersized hospital capacity, with an insufficient number of hospital beds, was the most pressing issue. However, undefinition or restrictions of the type of care provided by hospitals, low hospital performance and management problems, especially concerning human resources and infrastructure, were also pointed out as relevant aspects.
The biggest problem was the insufficient number of hospital beds, especially to fulfill the demand from the emergency departments. In the twelve-month period examined in this research, 24,564 requests were placed for hospital beds in Internal Medicine wards via SISREG. This number encompasses requests placed by municipal, state and federal health care facilities in the city of Rio de Janeiro. During the same period the CRMRJ was able to obtain a bed to only 3,239 of the requests placed for Internal Medicine wards, resulting in 86.8% of unanswered requests. For Intensive Care Unit beds the situation was even more dire, 8,310 requests were placed in SISREG, but only 936 requests obtained a bed via CRMRJ, leaving 88.7% of the requests unmet. (Table 1) A national research that assessed the implementation of the UPA, indicated that access to hospital beds was the major constraint to these units' performance. Difficulties to get patients admitted to hospitals were persistent and led to distortions of the work performed and of the care provided by these units 17 .
The network of public hospitals located in the city of Rio In a research that sought to define the number of public Adult ICU beds required to meet the existing demand, an analysis restricted to the ICU beds regulated by CER-RJ identified that, to guarantee a stable system and a maximum waiting time of 6 hours to reach an ICU bed, the existing number of ICU beds should be doubled, if maintained the average length of stay in intensive care units 39 .
A specific example of this scenario of insufficiency is the acute myocardium infarction (AMI) care pathway. In several situations it is necessary to perform primary or rescue coronary angiography. Interviews reported that access to this intervention is restricted to specific cases and frequently take various days to occur. Specialized Care Institutes and Hospitals in which this intervention can occur, perform the procedure, but do not admit these patients to a hospital bed for immediate follow-up. Therefore, patients that succeed in undergoing coronary angiography have to return to the UPA or CER, and are later discharged from these same pre-hospital units. If in the one hand this arrangement has somewhat increased access coronary angiography, which has an important impact in decreasing AMI morbimortality, on the other hand there are problems in providing postprocedure care in pre-hospital setting, that is in UPA and CER, for these units are neither trained or fit to do so, since they lack the resources and knowledge do deal with potential complications.
A research conducted in the city of São Paulo demonstrated that the organization of a specific pathway to provide health services for AMI patients had a significant impact on disease morbimortality 40 . Thus, the arrangements established fore AMI care in the city of Rio de Janeiro as outlined above are unsuitable for the serious public health care issue that this disease represents. Even if the provision of coronary angiography in a prehospital setting is acknowledged as an increase in access to health care, its availability is unpredictable and its performance conditions are far from adequate. Proper management of AMI cases, because of its high prevalence and high morbidity and mortality, requires integration between different levels of the health care system. The absence of such integration considerably contributes to increase morbidity and mortality 40 .
The health facilities that more frequently granted CRMRJ access to their hospital beds during the period of this research are listed in table 2. Pediatric, psychiatric and OB-GYN beds were not included. The top five hospitals that made their hospital beds available to CRMRJ accounted for 75,5% of all the requests that were met. It is noteworthy that these hospitals did not have emergency departments. Municipal hospitals with emergency departments made hospital beds available in varying degrees, from 2 beds yielded by MH13 to 225 beds yielded by MH6. Federal administration made available only 53 beds to CRMRJ in 1 year. Among the Teaching Hospitals, only TH3 made 12 beds available. It is noteworthy that a pre-hospital unit, has ICU beds registered in the National Registry, that were made available to CRMRJ.
( Table 2) Brazil has an obsolete hospital system, in which predominate general purpose hospitals, of small and medium size, that provides acute, elective and long-term care, indistinctly.
Patients with different health needs are all counting on an indisputably small hospital capacity of 1.56 beds per 1,000 inhabitants offered by SUS 35 .
Among the hospitals in the municipality of Rio de Janeiro, there is a similar situation.
Frequently the same hospital offers emergency care, elective surgical procedures and ambulatory care simultaneously. These hospitals constantly face the challenge to reconcile the tasks of an acute care hospital with the modus operandi and structure of a specialized hospital. In this situation, the same hospital bed is requested by an overcrowded emergency department and also by the long waiting list to elective admission, mostly to perform surgical procedures.
In 2012, the city administration created the CER in order to address overcrowding in the emergency departments of municipal hospitals. These units became responsible for all medical emergencies while the emergency departments of their neighbor acute hospitals focused on accident and trauma emergencies. These units were constructed in separate buildings and had their own health care staff, which represented an expansion of human and physical resources for emergency care.
A selection of 10 diagnostic and therapeutic resources available in public hospitals in the city of Rio de Janeiro is displayed in table 3. The number of Internal Medicine ward, Surgery ward and ICU beds and the existence of emergency departments is also presented for each hospital. Since all facilities possessed electrocardiogram devices and X-Ray equipment to perform simple radiographs these were not included. This table shows that a higher availability of medical technology is found precisely in hospitals that have a larger bed capacity and also emergency departments. (Table 3) Comparing data from the tables 2 and 3 it is striking that, among the five hospitals that more frequently made beds available to CRMRJ, only 2 have a high availability of medical Most municipal hospitals count on a good array of medical technology including CT scans, but many of them preferably admit to their beds patients waiting in their own emergency departments. The same happens with the two state hospitals that have emergency departments, SH2 and SH4.
While the hospital beds needed are expected to provide care to a large variety of health issues and problems, it is not unusual to find hospital beds whose availability is restricted to certain types of diseases. In some cases, this results from the restricted availability of medical technology as referred to before, but in other cases, especially in Teaching and Federal hospital, chief of services, without acknowledging the population health needs, decide alone the scope of care their service is providing.
In addition to the difficulties outlined thus far, a few hospitals exhibit a mismatch between their prevailing type of hospital beds, mostly surgical beds, and their predominant workload, mostly clinical emergencies, which often results in emergency department overcrowding.
A study on the definition of the functional typology of Federal Hospitals in the city of Rio de Janeiro observed that this process is barely rational and systemic and that change is largely driven by the urge to solve problems instead of deriving from planned action to render the hospital more efficient and integrated to the network of health care services.
Each hospital incorporates technology and seeks to answer to health needs independently, ascribing their behavior to an allegedly disorganization of the local health system that hampers the possibility of mutual adjustment in order to reach integrated care. The discourses of resistance to change perceived in these hospitals demonstrate that the power dimension within these organizations needs to be considered in order to shape strategies to revise the functional typology of these hospitals 41 .
Another study also corroborates the existence of difficulties for the integration of hospitals in the RUE, evidencing that this process requires the definition of pacts and care flows, adequate functioning of other health services in the network, especially Primary Care Services, efficient regulation of access to services, and engagement of hospital managers and professionals 42 .
Regarding hospitals' performance, a heterogeneous dynamic has been observed with significant variations in the average length of stay and hospital bed occupancy rate. Table   4 presents the total number of hospital beds and hospital admissions, the total number of Internal Medicine beds and hospital admissions due to clinical diagnosis. The table also shows the average length of stay and the hospital bed occupancy rate.
( 93% and FH5 86%. Occupancy rates higher than 85% increase risk to patients and in acute care hospitals rates higher than 90% denote a bed shortage crisis 35 . Conversely, most hospitals that do not have emergency departments exhibit occupancy rates below 75%, which suggests underuse of beds and resources. Teaching hospitals and Federal hospitals without emergency departments display predominantly this pattern.
The average length of stay is long for most hospitals, a previously known problem.
Considering all the hospitals listed in the table 4, the average length of stay is of 8 days, consistent with the OECD average in 2015 44 . However, when examining hospital admission due to clinical diagnosis separately the average LOS increases by 4 days, reaching 12.6 days. Calculating the average length of stay of hospitalizations due to clinical diagnosis for each administrative level, the results are the following: 9.5 days for municipal hospitals, 7.75 days for municipal hospitals, 15.4 days for federal hospitals, and 12.5 days for teaching hospitals. Considering only the hospitals that have emergency services (7 municipal, 2 state and 3 federal) this average drops to 10.5 days.
The hospital sector performance was considered problematic and represented an additional element to accentuate the shortage of beds according to some UPA coordinators who also worked or had worked in the hospital sector. Long average lengths of stay are explained by several limitations of hospital operation.
Among them we have: reduced operation at night and on weekends, primarily due to reduced availability of consulting physicians and diagnostic services. The availability of diagnostic services may actually be a critical issue for hospital efficiency. A study that sought to analyze the reasons for delayed hospital discharge of patients in internal medicine wards of teaching hospitals identified as one of the main reasons for this problem the long delays to perform diagnostic tests or to release tests reports 45 . Other factors related to delayed hospital discharged included: a more time-consuming decisionmaking process due to longer discussion over clinical case in teaching hospitals as well as difficulties in obtaining specialists opinion on cases 45 . However, the most common scenario is the absence of a daily ward routine performed by a single consultant. Instead, the care plan is devised each day by physicians and health teams working shifts, who may be unexperienced or unqualified, resulting in longer lengths of stay, misdiagnosis and ineffective treatment.
Besides all the hardships previously outlined, existing hospital beds can become unfit for admission due to physicians and nurse shortages or else due to neglected maintenance: infrastructure in need of repair and damaged equipment.
Within the municipal administration, it is already acknowledged that, in a scenario of chronic scarcity and prolonged financial constraints, the improvement of working processes is fundamental to increase access to existing resources. In this sense, the reduction of the average length of stay is one of the goals to be implemented, through optimization of health care provision.
The emergence of NIR promoted the expansion of the regulatory culture that succeed, little by little, in confronting hospitals' self-centered behavior in relation to becoming part of the health care network. It also contributed in making hospitals recognize their performance and efficiency problems. The progressive organization of the network and the growth of the regulatory culture forced hospitals and their teams to start perceiving themselves as a part of the system, having to share responsibilities.
The idea that chiefs of services have unquestionable control over admission to the wards under their responsibility is still appreciated by many, which lead hospitals to neglect their expected and potential role in an integrated health services network when organizing its operation.
Despite have being drafted in norms form many years now, only recently NIRs have been enacted in the city of Rio de Janeiro. Thus, these committees still have little power to oversee access to hospitals' resources. Support from the board of directors is critical for the NIR to perform its tasks. The extent to which they succeed in attending requests placed by CRMRJ, for hospital beds and diagnostic tests, depends on how hospitals perceive themselves within the network and how high-level managers from Health Department understand the importance of regulatory mechanism and the role of the hospital for accomplishing the goal of an integrated health system. Often even with the support of the board of directors, NIRs persist requiring authorization from chiefs of services to grant CRMRJ access to hospital beds.
In the process of building and structuring SUS, hospital sector was continually neglected.
The elevated investment costs required to improve hospital sector, coupled with the process of expansion of outpatient care, underpinned by an overcritical discourse to the hospital-centered model, culminated in a problematic scenario in which the Brazilian hospital sector is predominantly small in size and inefficient in the provision of care 37 .
National studies on the implementation of prehospital care in Brazil that analyzed the SAMU and the UPA pointed out serious issues concerning lack of support from hospitals and emergency departments, in terms of access to diagnostic tests and hospital admission, that significantly impacted the mission and performance of these facilities to deliver prehospital emergency care 17,18 .
In the case of Rio de Janeiro, as mentioned, there is an obsolete and fragmented hospital network, with serious management problems within its hospitals. In spite of the managerial technologies incorporated, especially with regard to regulation of access to care, the hospital network remained largely unchanged before the expansion of access to health care was made possible by improvements in the structuring of the RUE. Over the last 10 years state health policy has strongly focused on the expansion of the Emergency Care Network, with the creation of numerous UPA. Nonetheless, the creation of Brain Institute, Children's Hospital and Orthopedics Hospital express investments in the provision of inpatient and outpatient Specialized Care. In addition, a foundation to manage human resources for health care was created to overcome legal constraints to employ public servants, which allowed an increase in the number of health professionals, mainly in Specialized State Institutes for Endocrinology and Cardiology.
The public hospital, notably in the example of Rio de Janeiro, continues to be a locus of technological concentration, whose importance has not declined with the incorporation of technological advances in outpatient care, since access to medical technology in an outpatient setting is still very fragile in the municipality. Outpatient access to diagnostic tests, specialists' referrals and pharmaceuticals is insufficient and incompatible with the trend of progressive reduction of hospital beds observed in Brazil 35 .
Other factors render the inadequacy of the hospital sector and the gaps in the provision of care in the system as a whole even more acute. Population ageing constitutes a major challenge for health systems, which in Brazil's case is greatly accentuated by a situation of significant social vulnerability, derived from profound socioeconomic inequalities.
The hospital network of the city of Rio de Janeiro, considered all the 30 hospitals, of table 1, is comprised predominantly of large and medium sized hospitals. There are 14 largesize hospitals (more than 150 beds), of which 6 have emergency departments, and 12 medium-size hospitals (between 50 and 150 beds), 5 with emergency departments. The predominance of large-size hospitals in the network is a positive characteristic, especially when observed that these hospitals cluster most of the medical technology, medical specialists and pharmaceuticals available in SUS. However, this favorable aspect is overshadowed by the significant number of such hospitals that provide emergency care.
These hospitals face the challenge of reconciling the demand arising from overcrowded emergency departments and the requests of access to hospital beds and diagnostic tests placed by primary care services, prehospital services and small-size hospitals, via CRMRJ or CER-RJ. This dilemma remains unsolved and hinders the integration of these hospitals with the network health services. One of the managers interviewed in this study calls for an urgent remodeling of the hospital sector in order to better organize care, obtain the best performance from the existing hospitals and accommodating the emerging and growing demand for long-term care.
Hospital sector is complex and expensive. Costs to invest and remodel are high, as well as to maintain and improve. The recent successful development of RUE made explicit the need to expand the size of the hospital sector. However, increase of hospital services capacity could be attained through improvements in hospital performance and operation, which in face of prolonged financial constraints, might be the only path available to strengthen hospital sector within SUS.

Health System Segmentation
In addition to the analysis of the hospital sector presented thus far, the segmentation of the local health care system and the overlapping of services stand out as issues that further aggravates the situation.
In the city of Rio de Janeiro, this overlapping has its origins in the first half of the twentieth century, when the public health system was comprised of the municipal hospitals (Rio de Janeiro was the federal district at the time), open to all citizens, and of the Social Security Institutes for Medical Assistance that provided medical care only to those that were regularly employed. Later theses institutes were incorporated by the federal administration, being run by the federal government since then. Over time state administration also took over some hospitals when the city ceased to be the federal district.
Tense and conflictive federative relations between state and municipal administrations in the city of Rio de Janeiro is an old and well-known problem, mainly derived from both the presence of state and federal health facilities in the municipal territory and from the burden derived from providing emergency care to the population of bordering cities, without the regional political action expected from state administration 46  The segmentation of the network is elicited in table 5. In the sample of ICU bed requests, from the 50 requests placed by state UPA through CRMRJ that were granted access to a hospital bed, 78% were beds in state hospitals, while only 8% were beds in municipal hospitals. From the 185 requests placed by either a municipal UPA or CER through CRMRJ that were granted access to a hospital bed, 91% were beds in municipal hospitals, while only 1.6% were beds in state hospitals. In the sample of Internal Medicine ward beds requests (table 5), from the 16 requests placed by state UPA that were granted access to a hospital bed, 43% were beds in state hospitals and another 43% were beds in municipal hospitals. From the 345 Internal Medicine ward beds requests placed by municipal UPA or CER that were granted access to a hospital bed, 95% were for municipal units. CRMRJ had no access to beds in state hospitals. It can be noticed that state UPA had access to both state and municipal hospitals, while municipal UPAs had access to only 3 state hospital beds. (Table 5) The federal hospital system in the city of Rio de Janeiro, which harbors a large amount of the medical technology available in the local health system, does not operates with effective regulation of access to its beds and barely acknowledges its crucial role in the Emergency Care Network. Over the last few years, the federal hospital system has been continuously compelled to become integrated to the health care network. However, the federal administration has largely retained its autonomy in the management of its hospital beds. This autonomy of the federal hospitals in relation to the network is also seen in the interaction among the federal hospitals.

Conclusions
The recent developments in the regulation of access to health care provision that took place in the Emergency Care Network of the city of Rio de Janeiro secured improvements through the implementation of centralized pathways to promote transparent and fair distribution of scarce resources and utilization of scientific and negotiated criteria to allocate procedures and hospital beds. However, the regulatory mechanisms to organize access to health care seem rather to seek an optimization of the existing resources, falling short from inducing a remodeling of services. This research succeeded in displaying that the limited accomplishments of the experience examined can be ascribed to a combination of two prominent elements: insufficient funding, human and physical resources coupled with conflicting and fragmented institutional and governmental interplay.
Funding and infrastructure available to meeting the health system principles and people's health needs never did suffice in Brazilian Unified Health System. As outlined in this study, the hospital sector depicts well the scenario in which outright lack of resources or shortage of health professionals is greatly aggravated by an incompetent and fragmented managing of the existing health care resources and facilities. In Rio de Janeiro, as in most of Brazil, public hospital beds are far from being sufficient. Moreover, the average hospital is an old, small to medium size, general-purpose hospital expected to offer acute, elective and chronic care simultaneously. In the city of Rio de Janeiro, the majority of hospitals was built during the 70s and has never undergone any restoration. Approximately half of them account for most of the high-technology diagnostic and therapeutic resources.
The challenging and troublesome interplay between the governmental spheres that put the local health system to effect, due to the different levels of expertise displayed by each executive level, was also demonstrated. Furthermore, the interaction among managers and authorities is frequently tarnished by political interests, hindering efforts to administer cooperatively responsibilities, resources and power, paramount to achieving  All participants that agreed to participate signed an Informed Consent Form prior to being interviewed.