"The Challenge of Creating a Health Information System in the Context of Local Health Systems"

The Challenge of Creating a Health Information System in the Context of Local Health Systems: The Example of the State of São Paulo, Brazil (*)

Eduardo O.C. Chaves

Contents

I. Introduction

II. The Background

III. The Challenge

IV. The Solutions Proposed

V. Conclusion

I. Introduction

The task of formulating policies for the development of computerized information systems in support of local health systems in Latin America and the Caribbean is certainly not an easy one. Difficulties of all kinds suggest themselves promptly.

The operation of a health information system involves, among other things, the following aspects:

1. The first aspect to be considered is, naturally, that of information itself. In this case, we must deal with at least two issues:

A. On the one hand, we must face the problems related to the generation of information within the health system. The main problems here are in defining the layout of the system, the form of data collection, the flow of information, the type of processing it will undergo, the form of storage, the degree of security and confidentiality, the level of analysis, the means of dissemination of the report, etc.

B. On the other hand, we must also face the problems of access to information already existing outside the health system. The main problems here are in ascertaining what information is already available, how one can have access to it, what procedures must be followed in order to routinely access this information, etc.

C. Information is, in general, utilized within a health system mostly, on the one hand, to subsidize the tasks of planning policies and programs, and, on the other hand, to provide the grounds for the tasks of evaluation and control of the implementation of these policies and programs. Consequently, the people who should be most interested in health information are, on the one hand, central and regional planners, who must define policies, programs, goals and strategies, as well as, on the other hand, those who occupy posts of direction, coordination and supervision, and who must, therefore, be responsible for controlling and evaluating the execution of policies, programs, strategies. etc. There are two big problems here. One lies in succeeding that planners and those who occupy posts of direction, coordination and supervision agree as to what information is necessary for the tasks they are supposed to perform. People overestimate their information requirements, and the result normally is that the amount of information considered necessary ends up being enormous -- a fact which contributes to it not being gathered, or to it being gathered with a a rather low degree of trustworthiness and reliability. The second problem is in the fact that, especially in government organs, there is a relatively high degree of rotation in posts of direction, coordination and supervision, which, in their majority, are positions of trust. When the occupants of these posts change, the definition of what information is necessary also changes, and, consequently, are changed data entry forms, processing routines, reports, etc.

2. The second aspect of the issue which must be considered is that of the computational structure necessary for the collection, processing, storage and dissemination of the informations generated within the health system as well as for access to information available only outside the health system. The main problems here are in defining the model of data processing to be adopted, the equipments and software to be utilized, the norms for negotiating their acquisition, the appropriateness of developing the software internally, the degree of control and standardization to be enforced for hardware and software, the policies concerning data security and hardware and software protection, the line of action concerning maintenance and support, etc.

3. The third aspect to be considered is that of the human resources necessary to put in operation this entire structure, either in the aspects related to information itself or in the aspects related to its computerized processing, or, still, in the aspects related to the access, through computers, to information stored in distant data bases. The main problems here, as one can easily imagine, as those of recruitment, remuneration, training and creation of attractive working conditions which may allow retention of personnel.

It is true that each one of these aspects must be taken into consideration by private enterprises as well as public institutions. Not much (if anything) in these observations is peculiar to health systems. Public institutions, however, have a number of other additional complications, such as, for instance, requirements for the standardization of procedures and routines, guidelines and norms for the collection, treatment and storage of information, for the acquisition of computer equipment and other computational services, for the acquisition and development of software, and other controls and, frequently, barriers, of a bureaucratic nature.

II. Background

The State of São Paulo is the most populous and, from an economic point of view, the most important State in Brazil. Its economic growth, however, does not equally benefit all of its population, since there is a large number of people in the State who live in inadequate conditions, when not in absolute poverty.

The city of São Paulo, which is the capital of the State of São Paulo, has a projected population of 10.822.614 inhabitants for 1988, which is approximately one third of the State's total population. The capital city and 37 surrounding cities constitute the Greater São Paulo Metropolitan Region, with a total population of 16.454.422 people. Excluding the capital city, the Metropolitan Region has a population of 5.631.808 people.

The State as a whole has 572 cities, 534 of which, with a total population of 15.047.092, outside the Metropolitan Region. The projected population of the entire State for 1988 is, therefore, 31.501.504 inhabitants.

The public Health System of the State of São Paulo is constituted by health units (hospitals, ambulatories, primary health care centers, laboratories, etc.) which belong to the government in any of its three levels: federal, state and municipal. Today, various kinds of agreements regulate the responsibilities of the three levels, so that they become a real system, and not simply an aggregate of units.

Administratively, the system today is divided into 62 Regional Health Offices, which are grouped under five Co-ordinations. One such Co-ordination is responsible for the 15 Regional Health Offices in the Greater São Paulo Metropolitan Region, eight of which are located in the city of São Paulo itself. The Health Secretariat of the State of São Paulo co-ordinates and supervises the entire system.

To complete the picture, the public Health System hires services from the private Health System, and offers them freely to the population. In some areas (e.g., Hospital in-patient care), the amount of services hired from the private hospitals is about four times greater than the amount of services rendered by the public hospitals (80% to 20%).

III. The Challenge

1. In terms of the numbers

Until the beginning of last year, the three levels of government -- federal, state and municipal (or local) -- operated independently in the State, as far as the health sector was concerned.

The federal level, to make matters worse, had two arms in the State. On the one hand, the Ministry of Health acted mostly in the area of defining policy for what has been traditionally understood as "public health". This policy was then complemented, detailed and executed by the State. On the other hand, the Ministry of Social Security and Assistance had, as one of its tasks in the State, to offer medical and hospital assistance (in-patient, out-patient, laboratory support, etc.) to all those affiliated with the National Social Security system (i.e., those who have registered employment and their dependents). To tell the entire truth, the federal level still has a third arm in the State, through the Teaching Hospital of the one federal School of Medicine existing in the State, the Escola Paulista de Medicina, located in the capital. The Federal University of São Carlos -- the only federal university in the State -- does not have a School of Medicine.

Since the federal level itself could (or would) not offer services of its own in sufficient quantities to face the demand, it had to hire these services from the private sector. With time, the hiring of services became the norm, not the exception, the result being that the greatest majority of services rendered by the federal level were services hired from the private sector.

As far as the state level is concerned, the State Secretariat of Health had to complement, detail and execute the policies of the (federal) Ministry of Health as well as offer primary health care in its own units. The State also had responsibility over some 35 hospitals, which, originally, were specialized hospitals (for Mental Disorders, Tuberculosis and Hansen's disease), but which were gradually becoming general hospitals.

The five Teaching Hospitals (two in São Paulo, one in Ribeiråo Preto, one in Campinas and one in Botucatu) of the four Schools of Medicine (São Paulo, Ribeiråo Preto, Campinas and Botucatu) of the three State Universities (Universidade de São Paulo - USP, Universidade Estadual de Campinas - UNICAMP, and Universidade Estadual Paulista - UNESP ) were also under control of the State, but not of the State Secretariat of Health. The state Universities in São Paulo are not under the control of the Secretariat of Education, as one might expect, and as is the case, mutatis mutandis, at the federal level, where the federal Universities respond to the Ministry of Education. In São Paulo, the Universities either have responded directly to the State Governor or, as is now the case, respond to the State Secretariat of Science and Technology).

At municipal or local levels, the services under control of the local health authority depended on the size and the initiative of local government. Some cities, mostly the larger ones, have rather complex health systems, with primary health care units, secondary hospitals, support laboratories, etc. A couple of cities even have Schools of Medicine maintained by local government, which use a local hospital as Teaching Hospital (e.g., Jundiaí, Taubaté). In most of the smaller cities, however, all public health services were under the control either of the State or of the federal government.

Computing all three levels (federal, state and municipal) the public Health System in the State of São Paulo has, roughly, 2,500 primary health care units in the State (excluding Hospitals and Laboratories), so divided:

- Federal level: 100

- State level: 1,500

- Municipal levels: 900

The federal government has five Hospitals in the State, all of them in the Greater São Paulo Metropolitan Region. The State has around 36 hospitals, 35 Psychiatric Ambulatories and 83 Laboratories. The municipal or local levels support a few general hospitals.

The private sector has around 2,500 units in the State, most of which render services to the public sector through contract with the federal government.

2. In terms of the organization of the system

With the creation, in June of 1987, of the Unified and Decentralized System of Health (SUDS) in the State, all the health units which belonged to the federal level were transferred to the control of the State Secretariat of Health, i.e., came under the administration of the State level.

Roughly at the same time, special agreements between the State Secretariat of Health and the Universities of the State brought the services of medical assistance provided by the Teaching Hospitals under partial control of the Secretariat of Health.

At the same time, the State began transferring to local governments the control and administration of the primary health care units which it had and which were being transferred to it by the federal government. To date, 479 cities have already signed agreements with the State government, whereby the State releases control of the primary health care units in the geographical area of that city to local government and assumes the responsibility of transferring funds for the maintenance (and even expansion) of the services.

The end result of this process is that local government will be responsible for primary health care, and the state government will be responsible for secondary and tertiary health care. The federal government will only have responsibilities in fixing general guidelines, which will be supplemented, detailed and operationalized at the state level.

3. In terms of the multiplicity of information systems

As far as information systems are concerned, this whole picture comes down to the following.

Prior to the creation of the Unified and Decentralized System of Health (SUDS) the State Secretariat of Health had a number of information systems under the co-ordination of the Center of Health Information and Informatics (CIS) (Centro de InformaçÆes de Saúde -- CIS), created in 1977, in order to centralize the process of collection, processing and dissemination of information within the Secretariat. At the beginning, the CIS would process its information manually. When, around the beginning of this decade, it felt the need for automated data processing, it did not have any alternative but to resort to the highly centralized State Data Processing Company (Companhia de Processamento de Dados do Estado de São Paulo --PRODESP), which only used batch processing in mainframes at that time.

The federal level had its own health information systems in the State, some of them processed manually, some of them processed by the Data Processing Company of the Ministry of Social Security and Assistance. The Ministry of Health by and large did not have any health information system in the State, although it did make en effort to develop a data base of health units and services.

The larger cities had their own health information systems. The smaller ones would often use the systems provided by the State Secretariat of Health.

The main problems of this setup, as one can easily guess, was the segmentation and compartmentalization of information, which often caused its duplication, as well as the lack of trustworthiness, reliability and updating in the information which was gathered. The State Secretariat of Health, the two federal ministries, each tried to implement its own health information systems, the result being that the State ended up having several disintegrated small information systems.

Beginning in 1986, before the creation of the Unified and Decentralized Health System, the State Secretariat of Health started to use microcomputers and to decentralize the process of data entry and processing to its Regional Offices for some of its health information systems.

IV. The Solutions Proposed

1. In terms of the general organization of an information system

With the present administration, a clear policy on health information was defined, which was coherent with the principles of the Unified and Decentralized System of Health to be implanted. This policy was built upon five pillars.

In the first place, there was the need to begin by the organization and implantation in the State of a true State System of Health Information, unified, integrated and interinstitutional, under the coordination of the Health Information System (CIS) of the State Secretariat of Health.

This system eliminated duplications and started feeding information to the information systems of the federal Ministries, of the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística -- IBGE), to the overall information system of the State, co-ordinated by the State Foundation for Data Analysis (Fundaçåo Sistema Estadual de Análise de Dados -- SEADE), which belongs to the State, to the Latin American and Caribbean Center for Information on the Health Sciences (Centro Latino Americano e do Caribe de Informaçåo em Ciências da Saúde -- BIREME, institution which belongs to the Pan American Health Organization and is dedicated to collecting, processing and disseminating bibliographical information in the health area.

The CIS is also directly connected with the World Health Organization through its computers, through the international node of the Public Data Network (RENPAC) of the Brazilian Telecommunications Company (EMBRATEL).

This information system comprehends the information generated by health units and services of all three levels of government, as well as by private health units and services which are hired by the Unified and Decentralized System of Health.

The second pillar on which the policy of health informatics is built is the development and refinement, under the control of the CIS, of the sub-systems of information which are required for the Unified and Decentralized System of Health. These are the priorities in the area of health information. Included here are the data base of health units and services, an information sub-system on ambulatory care, an information sub-system on hospital care -- all for use by central and regional offices -- and an information sub-system for patient registration and movement -- for use by local health units.

The third pillar is the development of human resources in the area. This involves hiring, forming and training specialized personnel in the area of health information and in the area of informatics applied to health. The task here is enormous, and it is impossible to overestimate the difficulties, in function of the large numbers involved.

The fourth pillar of the policy of health information is the acquisition of data processing and telecommunications equipments, to allow the various units of the Unified and Decentralized System of Health to transmit data and otherwise to communicate among themselves. It is a heavy investment, but much has already been done.

The plan is to consolidate a network of computing equipment. In a first moment, all the microcomputers located in central and regional organs will have access to the central computers, localized in the Center of Health Information and Informatics (CIS). In a second moment, the Hospitals, Specialized Ambulatories, Laboratories and Blood Banks will be connected. In a third moment, the municipal Secretariats of Health of the largest cities, municipal consortia, and the largest Primary Health Care Centers will join the network. Finally, every health unit in the State will be connected.

According to this model, the data produced by the health units will be collected and, initially, sent to the Regional Offices for data entry and processing. After the data of all the units which belong to a Regional Office are consolidated, they are transmitted to the central computers of the Center of Health Information and Informatics (CIS), where they will receive final treatment, for the whole State. In the central computers these data will be ordered in data bases for free consultation by the health system and by any interested user with access to the Public Data Network (RENPAC). Of course, a major user of this information will be the organs responsible for planning, supervision and control.

Finally, as the fifth pillar, the dissemination of information will receive special attention, with the use of modern methods of computer-assisted publishing.

Through the computer network, all the main organs of the Unified and Decentralized System of Health will have instantaneous access to information. Also, through a Videotex system, already implanted, all the organs will have access to messages, general information, summaries of articles, etc.

Finally, through a computerized PBX and telephone call distribution network, which will operate 24 hours a day, seven days a week, anyone will access to a telephone will be able to ask for information through a four-digit telephone number.

2. In terms of specific information systems

In this section I will discuss four specific information systems which were developed or were in the process of being developed. Since the fourth of these systems is the one that applies, in a more directly form, to local health systems, it will receive more attention.

A. Development of a unified and decentralized information system on health units and services

Prior to the creation of the Unified and Decentralized System of Health, each level of Government had its own database on health units and services. As one can imagine, these databases had different designs and layouts and the information contained in them overlapped and was not consistent.

One of the first tasks to be faced by the Center of Health Information and Informatics (CIS), after the creation of the new health system, was to devise and develop an information system on health units and services which was comprehensive and that contained all the information which was required by federal, state, regional and local organs.

The system had to be integrated, in the sense that it would no longer be admissible that the different levels of government kept their own databases: the database was to be one, maintained in the central computers of the Center of Health Information and Informatics (CIS). However, each regional office would receive a copy of the segment of the database that corresponded to the units and services under its jurisdiction, and would be responsible for verifying and updating the information on these units and services, to enter it into diskettes, and to send it to the central computers.

B. Development of a unified and decentralized ambulatory care information system

Up to the moment, the Secretariat of Health has several rather specific and segmented information systems, to provide information on the production of health units and services, on payment of hired services, on budget execution, on morbidity, etc.

These systems are rather limited, since each of them deals with a very particular part of a whole. Each one was developed at a different time, with different tools, and it is virtually impossible to add up the information they provide to obtain an integrated and coherent view of how things stand in the health sector, with respect to services produced, money spent, morbidity, etc. This means that it is virtually impossible to really evaluate the system.

To replace this rather chaotic situation, two systems were conceived: an Ambulatory Care Information System and a Hospital Care Information System. In this section we will describe the first, and in the next the other one of these systems.

The Ambulatory Care Information System was conceived and designed from the point of view of the user who wants to evaluate ambulatory care within the health system, either globally (for the whole State) or partially (from the point of view of the regional administrator or supervisor or from the point of view of the director of a local service). Specific health indicators were defined, and then the information requirements for these indicators were determined. On the basis of the information required for these health indicators, forms were devised which would collect the data.

The end result will be one single, integrated system, which will collect information on the production of health units and services, on payment of hired services, on budget execution, on morbidity, on so provide the elements for the utilization of health indicators, which will permit the constant monitoring and evaluation of ambulatory care in the health system.

C. Development of a unified and decentralized hospital care information system

The Hospital Care Information System is the other side of the same coin. It aims at defining and measuring the production and performance of the hospitals which take part in the Unified and Decentralized System of Health (either those own by the federal, state or municipal government or those private hospitals hired by the State).

The system will allow the formulation and execution of sectorial policies with respect to the necessity of building more hospitals or hiring services, and will provide elements to the hospital administrator which will make planning, budgeting and quality control possible. The system will also give the researcher a number of measurements which will help in the explanation of a large observed variability in the utilization of resources, the production of services, and the cost relationship.

The selection of indicators will consider the variability of types of hospitals, institutional "case mixes", regional variations, etc.

D. Development of a unified and decentralized system of patient registration and control of patient care and movement

a. Objective of the project

This project has as its objective the development of a Unified and Decentralized System of Computerized Registration and Control of Patient Care and Movement for the State of São Paulo.

The system will allow the creation of a data base structure, implemented through a computer network, which will make possible to register a patient, when s/he first comes to a health unit of the Unified and Decentralized System of Health of São Paulo (SUDS-SP), and then, if this is the case, follow her/him, as s/he goes through the various levels in the hierarchy of the health system. The patient's medical record will be kept in a regional data base, consultations at the various levels of the system will also be scheduled through the computerized system, and the patient's history, as s/he moves through the various levels of primary, secondary and tertiary care, will be recorded. A summary of transactions will be sent to a central data base, stored in the central computers of the State Secretariat of Health, making the process of data collection and dissemination of information much more agile.

It must be noted that the manual method today used by the State Secretariat of Health is inadequate and inefficient to provide reliable information about the care given to the patients and about their movement through the many units of the system. The movement, today, is also often horizontal, as the patient moves from unit to unit, at the same hierarchical level, without finding a resolution for her/his problem. One unit is often unaware that the patient has already been to another unit, since there is no centralized registration system and no record of the care given the patient at the various places s/he goes to. The present method is also inadequate to provide support for an efficient system of information on morbidity, epidemiology, performance statistics, etc.

b. Objectives of the system to be developed

The main objectives of the system to developed are:

-- To create a relational system for collection and treatment of information on all the patients cared for by the state system of health, which will be at the same time unified (at the central level) and decentralized (at the regional levels);

-- To implant, at the regional levels, decentralized data bases, which will have information on the medical records of all patients, the care that they have received and their movement through the various hierarchical levels of the health system;

-- To implant, at the central level, a centralized data base, which will have information on morbidity, epidemiology and the performance of the health units of the system.

-- To implant procedures for the analysis of health conditions in the state, which will relate the information collected by the regional data bases on morbidity and epidemiology with other demographical and bio-statistical data about the population of the state.

-- To implant procedures for the analysis of the efficiency of the health system, which will relate the information collected by the regional data bases on patient care and movement with data about the installed capacity of the state health system.

c. Methodology to be used in the project

The system will be developed by the Center for Health Information and Informatics (CIS) of the State Secretariat of Health, with the human, material and financial resources of the State Secretariat of Health and of outside financial sources.

The initial design of the system was made by the staff of CIS and is being perfected through discussions with chosen experts on the operation of the state health system.

A project team will be hired to develop all the software that will be needed by the system, both at the regional and at the central levels.

The model of distributed data processing was chosen, in which each basic health unit will have an IBM PC compatible microcomputer and will enter and process its data locally. Periodically, the local information will be consolidated, analyzed and sent to regional nodes (which can be a Regional Health Office or the main city for a consortium of municipal health systems, or, in the case of larger cities, the Municipal Secretariat of Health). Periodically, these regional nodes will consolidate and analyze the information and send it to the central computers of CIS.

Project development will be coordinated by a central team, subordinated to the Directory of CIS.

The system will be initially implemented and tested in the region of Penapolis, in the Northwest of the State. This region has several peculiarities which make it an excellent testing field for the system. First, the region is a microcosm of the State, and is characterized by a consortium of seven small towns (around 5,000 people in average) which organized themselves around the city of Penapolis (around 50,000 people). The consortium has sixteen primary health units and two secondary units for referral. For tertiary level referral, the consortium has to resort to a Hospital outside its limits, in the city of Aracatuba (which is the seat of the Regional Health Office which has jurisdiction over Penapolis and its seven satellite towns).

V. Conclusion

The creation of a health information system in a developing country is a tremendous challenge. The easiest element to cope with in this context is the problem of determining which equipments may be required to store and process the information and of acquiring them. The main difficulties are, on the one hand, to organize the information system itself and to conceive and design the specific sub-systems which constitute it, and, on the other hand, to educate the health personnel as to the necessity of gathering and transmitting information in a reliable and trustworthy manner, to train them to operate the specific systems, and to make administrators and supervisors aware of what the health information system can provide them and conscious of its importance and benefits.

(*) Paper presented at the "Taller sobre Formulación de Políticas de Informática para Apoyar los Sistemas de Salud en la America Latina y el Caribe", Washington, DC, USA, November 7 to 11, 1988

Last revised: May 02, 2004