"The Use of Videotex for Health System Support in São Paulo"

The Use of Videotex for Health System Support in São Paulo, Brazil (*)

Ilana Fogelman & Eduardo O.C. Chaves

1. Introduction

To be useful, information must be collected and transmitted efficiently, so that it is in the hands of those who need it when they need it.

In the Secretariat of Health (SH) for the State of São Paulo (SSP), Brazil, Videotex (VT) is being used in order to make more efficient the process of data collection and transmission from primary health care (PHC) centers to regional and central organs and to permit data feedback from central to peripheric organs.

This paper describes a project which has been taking place in the past twelve months in the city of São Paulo (CSP), capital of SSP, and draws some provisional conclusions.

SSP, with 572 cities and a population of around 31.5 million people (one third of which in CSP alone), has a health system (HS) divided into 62 Regional Health Offices (ROs). Eight ROs are located in CSP itself. The SH of SSP co-ordinates and supervises the entire system.

2. Objectives of the Project

The project consists in the establishment of a VT-based communications network in CSP, in order to achieve the following objectives:

- To improve the information system for epidemiologic surveillance (ES) on communicable diseases (CD), by establishing electronic data collection and transmission;

- To improve the utilization of information, by offering immediate electronic feedback to those who provide or collect data.

- To disseminate technical and administrative information to health professionals at the periphery of the HS.

- To improve communications among all elements and levels of the HS, through the use Electronic Mail (E-M);

3. The ES System

VT was introduced primarily (but not exclusively) to improve the information system for ES on CD, in view of the major importance of CD as a public health problem in the State.

ES on CD in Brazil is based on compulsory notification of suspected and confirmed cases of CD and has two main objectives:

- To detect, as early as possible, unusual disease patterns and, consequently, to trigger prompt disease control measures (immediate action ES system).

- To provide information on long-term trends in patterns of incidence and prevalence of diseases, so as to support planning, policy-making and program evaluation (long-term action ES system).

Neither one of the two objectives was being satisfactorily achieved. Deficient and inefficient data collection and information flow were among the major reasons for that.

The core of the ES system on CD is the Emilio Ribas Hospital (HER), which is the reference hospital for CD in SSP. Because of this, it draws the largest portion of cases that require hospitalization.

In the traditional, pre-VT system, the service units had to report by mail to the Center for Health Information and Informatics (CIS) any suspected case of CD. CIS would alert by telephone or written notice the Center for Epidemiologic Surveillance (CVE), which is the organ in charge of epidemiologic control. CVE, by its turn, would alert (usually by telephone) HER and the RO responsible for the unit which reported the case.

However, written communication from the service units was slow and unreliable and telephone communication with ROs was an arduous task, because of insufficient and, therefore, always busy telephone lines. This hindered the effectiveness of the immediate action ES system on CD.

ES data to support long-term action were collected by PHC centers and public hospitals, which sent their reports on confirmed cases of CD to CIS by mail, on a weekly basis. Underreporting, slowness in the information flow and inexistence of feedback to those who submitted information decreased the usefulness of the system.

4. The VT Technology

It must be said that the VT system implanted in Brazil -- by the Telephone Company of SSP -- was derived from the French one, and so suffers from some of its basic flaws. One of these is that the information which is transmitted from the central computers to the terminals travels at the acceptable speed of around 120 characters per second, but the information which is transmitted from the terminals to the central computers travels at the rather slow speed of only 7.5 characters per second.

Despite this problem, VT was chosen as the technology for this project for these reasons:

- VT's technical features, such as the possibility of setting up pre-formatted questionnaires for data collection, its capacity of accessing data banks and a built-in E-M service.

- VT's user-friendliness, which requires no specialized personnel and very little training for its utilization.

- Lower implementation and operating costs, when compared to alternatives, given the fact that the VT network was already implanted and available in SSP.

5. The Pilot Project

Since no previous utilization of VT had taken place in the health sector in Brazil, a pilot project was proposed in order to evaluate VT's potential, suitability and acceptability. ES on CD in CSP was chosen as the primary goal for the experiment.

One hundred MSX microcomputers emulating VT terminals were rented from the state telephone company, during the first year, for a cost of approximately US$ 17 dollars a month (which includes the subscription fee to the VT network). All applications developed were stored in the computers of the telephone company. A VT page composer was also rented.

The terminals were distributed to users at the central, regional and peripheric levels of the HS: at the central level, to the organs which have responsibilities in the ES System (especially CIS, which co-ordinated the project, CVE, and HER and its reference Laboratory); at the regional level, to all the eight ROs of CSP; at the peripheric level, to all PHC centers of two ROs (around 70).

At the users' level, two professionals (with at least High School diplomas) of each unit were offered a two-hour training session on the operation of the system, and became multipliers.

At the central level, the implementation team was composed of two Medical Doctors (with no previous experience in VT), who are in charge of applications development, and two technicians (familiar with VT programming) who are in charge of implementing and maintaining the applications.

6. Applications Supported by the Network

The network was proposed as an optional information and communications tool. Although aimed primarily at ES on CD, it was in fact used for a number of other applications. After connecting themselves to the host computers, users can either access the E-M service or CIS's menu, where all the questionnaires for data collection and information to be disseminated are listed. The menu includes the foolowing items:

A) Support to Immediate Action ES on CD:

The service units notify CIS by E-M of every suspected case of CD. CIS tabulates the information and relays it to CVE, HER and the RO involved.

HER also utilizes VT's E-M to inform CIS and to alert CVE and the ROs about hospitalized cases of CD which come directly to HER, bypassing PHC centers, thus allowing control measures to be immediately triggered at all levels.

E-M also supports horizontal communication among ROs, which transmit notifications among themselves, on behalf of hospitals which do not possess a VT terminal.

B) Support to Long-term Action ES on CD:

Reporting, through VT, of confirmed cases of CD has been experimentally used to allow comparison with the traditional, mail-based weekly reporting system. A version of the official questionnaire was adapted to VT and PHC centers agreed to notify by both methods so that the necessary evaluation could take place.

C) Information Feedback:

CIS is responsible for storing data which are continuously collected by all health services in SSP. These data provide the basis for the organization of a series of "Feedback Bulletins" open for consultation through VT.

Epidemiologic data on the number of reported cases of CD are displayed in tables, histograms and maps. Detailed information on any disease can be obtained, as, for example, the number of cases occurring in each RO or in each neighborhood of the city, with weekly updating. Health units can, therefore, follow up the temporal and spatial evolution of any disease, as well as compare their situation with that of other units or with figures of previous years, which are also provided.

Feedback bulletins providing mortality and demographic data are also available. Further information which is not displayed can be requested by E-M.

D) Delivery of Laboratory Test Results:

HER is receiving by E-M the results of bacteriological tests performed in its supportive Laboratory. On an experimental basis, this application has been limited to the diagnosis of meningitis, due to the disease's high incidence at the present time and due to the urgent need for more timely delivery of test results in order to introduce prompt therapeutic measures.

E) Dissemination of Technical and Administrative Information:

Technical bulletins organized by CVE offer information on starting or current epidemics (meningitis, leptospirosis, dengue, AIDS), on new diseases (Brazilian purpuric fever), on ES procedures, on immunization campaigns, etc.

The Immunization Division of the SH collaborated with the setting up of a self-evaluation program on vaccines and with the selection of abstracts of scientific articles on immunization.

CIS's bulletin board, with announcements of meetings and congresses, extracts from official publications containing government acts and legislation of interest, service referral information, summary of important health-related news of magazines and newspapers, etc., is also available for consultation.

7. Discussion

7.1. Contributions of VT to ES on CD

E-M has been an effective means to improve communications among CIS, CVE, HER, ROs, and PHC centers in support of immediate action ES on CD. The reliability of this alerting system is assured, since all users at the central and regional levels check their E-M box at least once a day. These organs can rapidly and simultaneously contact each other for technical and administrative purposes. PHC centers can easily contact the central organs or their respective ROs to check receipt of notifications, request orientation on ES procedures, demand information on particular cases, etc.

The daily observation of the network's operation regarding notification sources has provided interesting details on the behavior of patients within the HS. In fact, it showed that, when patients have, or believe they have, an acute health problem, they bypass PHC centers altogether, seeking medical care directly in hospitals - especially in HER, in the case of CD. This was clearly evident during the occurrence of the meningitis epidemics in the city. A large number of the notifications that were received every day was generated by HER and not by the PHC centers in charge of the area where the patient lived. The PHC centers would mostly report, instead, cases of chronic CD (e.g., tuberculosis).

This factual observation led to the conclusion that, since patients with acute cases of CD bypass PHC centers, information feedback is an important contribution to peripheric units. If this downward information flow does not exist, PHC centers cannot be aware of disease incidence patterns in their region. On the other hand, it was also shown that, with respect to notifications originating in PHC centers, since they included mostly chronic cases of CD, the speed provided by VT was not so crucial.

7.2. Considerations on Acceptability and Utilization

VT's great user-friendliness allowed users with absolutely no knowledge of computers to easily master its operation. Its user-friendliness was fundamental to its high degree of acceptability and to an effective introduction of information technology in peripheric levels of the HS.

Although VT was generally accepted, its utilization patterns differed among users. E-M offered an easy way to evaluate participation in the project, by allowing CIS to measure users' promptness in answering a given message. In general, 60% of users would answer a message on the same day it was sent; 30% would have an irregular answering behavior, ranging from a prompt response to a five-day delay; 10% of the users (all of them at the peripheric level) would answer just eventually.

Even though this criterion only measures one feature of VT - its E-M facility - other statistics kept on data collection and access to data banks shows that the overall utilization pattern is basically the same.

An investigation into the causes of the low level of utilization in some PHC centers showed that the main causes were the fact that there was only one telephone line available in the unit, which was constantly busy for other purposes, and the fact that users were asked to do things also in the conventional way, to test the use of VT, which meant working double.

7.3. Considerations on VT's Performance

There was general agreement among users regarding the potential benefits of VT as an information and communications tool. They agreed on the fact that the project could and should be improved by exploring other routines (besides ES on CD), by including other users in the network and by offering a broader range of information.

Following users' suggestions, CIS implemented the other applications mentioned above. "Pages of Information" were organized in a hierarchical structure that could be easily implemented, modified, removed or updated.

The most useful application of the network was considered to be E-M, which demonstrated to be an effective tool for the improvement of communications in the HS, enhancing vertical and horizontal information flow, promoting interpersonal as well as mass communication. It improved interaction among different elements and levels of the HS by allowing a new range of interconnections to be established, by speeding up information flow, by overcoming the communication bottleneck due to insufficient and always busy telephone lines, by allowing an immediate forwarding of written messages, by transmitting information independently of the presence of a receiving party at the destination point, etc. Information could be sent to specific users, being protected by passwords, or could be of general interest, being displayed on bulletin boards. Bureaucratic tasks could be simplified, since messages written only once could be automatically forwarded to several different receiving parties.

VT also proved to be an efficient way to provide information feedback to data providers and collectors. Constantly updated data could be easily displayed as tables, histograms, maps or plain text. Its impact on motivating users to collect data remains to be fully evaluated.

The network also proved to be an efficient means to disseminate health information to geographically scattered users and to improve accessibility to technical information. It could be utilized as part of a distant continuing education program.

Users considered VT a more pleasant and simple way to collect data. A disadvantage was mentioned regarding the fact that once the questionnaires were filled out and automatically sent, users would have no further access to them, not being able to edit them.

7.4. Concluding General Considerations

Costs are certainly an important aspect in the utilization of computers in Brazil. However, the major obstacle towards a successful computerization of the HS has not been high costs, but the lack of appropriate human resources.

When considering a computerized communications network, the larger the number of users, the larger the benefit each user gets from the system. However, if the operation of such a network is to depend on specialized human resources, its impracticality increases considerably.

In this regard, VT proved itself, in the context of SSP, to be an especially suitable alternative to a standard, computer-based communications network, since (a) its initial cost was low (since there was a VT network already available) and (b) it had almost no human resources development implications (in terms of quantity, skills and qualifications), either for implementation or for operation.

The initial investment in hardware was virtually negligible, since the equipments were rented from the telephone company. Few specialists were required for its implementation and none for its operation. The training requirements were minimal, given VT's user-friendliness. The time that was needed to implement the procedures was also small.

VT, therefore, allowed a very early production of benefits to users. Other technologies could bring greater benefits, but only at a much higher overall cost, and after a much longer span of time.

Finally, a question could be raised as to the cost of data collection using VT. Given its slow data transmission speed, from terminal to central computers, if all data entry were to be done directly on VT, in the case of large amounts of data, the cost of communications could easily counterbalance the low cost of the initial implantation of the system.

In order to face this problem, and given the fact that the terminals used in SSP were regular MSX microcomputers, the possibility is under study of adding a tape recorder or even a disk drive to these terminals, writing some data entry program in dBASE II (available in SSP for these microcomputers), and only transmitting the collected data once a day.

(*) Paper presented at the Congress of Medical Informatics MEDINFO 89, in Singapore, December 1989

Last revised: May 02, 2004