Journal of Medical Systems. Vol. 1, No. 4. 1977

Clinical Computer Information in the United Kingdom

Systems

John Anderson B.Se., M.A., M.D. , FRCP, FBCS

INTRODUCTION

During the past decade the National Health Service in the United Kingdom, under the direction of the Department of Health and Social Security, has made a limited invest- ment in the field of medical informatics in relation to patient care and a much larger development, as well as investment, in hospital district and regional care manage- ment. Perhaps this was to be expected in that the main thrust by the manufacturers of computers was to transfer software concepts and programs from the business to the health care field, often without much real study of the actual environment in which such systems were to operate. While such operations as far as administration is con- cerned have been useful to a limited degree, they have encouraged the notion of transfer of software and hardware without a thorough system analysis and design. This has been an unmitigated disaster in relation to the application of medical infor- matics to the personal and preventive care fields.

Progress has been made in developing relevant clinical information systems useful to the health care team, both in general practice and in hospitals, and this con- cerns applications for doctors and nurses and paramedical scientists (1). Such limited experiments have provided the basic learning period, during which much research has been done, and a little development of well-investigated areas. Rather less than acceptable was the effort put into research into model systems that might have a new and novel impact on the health care field, the emphasis being largely to automate ex- isting systems and procedures without considering the changing environment in which such systems were to be embedded. Little attention was also paid to the limitations of computer technology but in the first phase of research this is perhaps acceptable, in that it is necessary to promote and facilitate change, for pressures for change will stimulate technological developments. The information and communication revolu-

From the Department o f Medicine, King's College Hospital Medical School, London:

375

This journal is copyrighted by Plenum. Each article is available for $7-.50 from Plenum Publishing Corporation, 227 West 17th Street, New York, N.Y. 10011.

376 Anderson

tion is still gathering impetus in the health care field, as well as in society in general, but it has been slower than the experts have anticipated, largely because of inadequate tools to solve the information storage, retrieval, and analysis problems outside the business field. Lack of a thorough ongoing educational system to bring the uses of such advances to the notice of society in general and also to special areas, such as those involved in the health care fields, was evident. Rather, we have had the expert talking to the expert in a corner, while the rest of society has been wondering what the peculiar discussions would be likely to produce. Perhaps because of this and the saturation of people with reports of instant revolution andbreakthrough, we have had our senses blunted to real progress.

There have also been difficulties in relation to funding of projects, for as control and evaluation of the systems are centralized, everyone has to go through some in- teresting contortions in relation to evaluation, especially the thorny hurdle of cost-benefit analysis in a non-profit-making field. It would be correct to say that the approach of the Department of Health has been rather naive and lacked both defini- tion and thrust. To some extent also they have been caught up in the tide of the dis- aster of national productivity and their position has not been very secure. On the other hand, some of the difficulties have been of our own making. The special requirements of the different types of recording procedure, data storage, and analysis required in the medical field need a wide range of techniques and this was not appreciated when projects were designed. These requirements were not actually reflected in detailed ob- jectives. Given this, the problems of evaluation with the long lead time, especially for system implementation and development and the problems of user education, which were quietly ignored, combined to make a realization of adequate evaluation for- midable. What is now required is a rethink of its utility as now conceived. We cannot claim that evaluation of our motor car industry or in the aerospace field has been any more successful in the recent past, although the investment has been much greater.

There are special problems in relation to health care and medical records that have been known for some considerable time, but only by the few interested in the field. However, the pressure to ignore recording, storage, and communication procedures has brought the multiple uses of the medical record in particular out into the open, and the multiobjective, multitask, and procedure requirements of the documentation system have at last been revealed. Part of the medical record must in- evitably be patient-oriented and perform communication and clerical functions predominantly, hoping to guide the progress of care for the health team. Such records act as an aide-m6moire for those who are in the front line of care and undoubtedly im- prove the thrust of care and act as a communication means between those who must know what was done. Thus in different countries the medicolegal requirements will vary according to the nature of the law and the attitudes adopted by society and the legal system in relation to health care. What has been ignored so far has been the ability of such records to provide a system of evaluating both individual and collective health care. We are still struggling with the traditional importance given to death rounds and mortality data as the main index of care. Unless we change this basic in- terpretation of care we cannot raise our sights to get on top of the morbidity problem, which is the real medical problem.

What has been difficult to do is to research and develop new medical model

Clinical Computer Systems in the United Kingdom 377

record systems on a limited scale, discard what is obviously noncontributary, and en- courage the development of those models that look useful at present and in the future. Part of this problem relates to medical attitudes to certain aspects of quantitative rather than descriptive data. The medical profession usually has had little difficulty with computing its reward system but has been very reluctant to develop quantitative tools for use over the field of practice. Doctors are said to be nonnumeric person-s, full of sympathy for their fellow men, but this is one of the myths that must be discarded. At present in the United Kingdom our medical students are as able numerically as any student in science. We must encourage the expansion of this area of activity by the full use of new information tools that are available during their training.

Because information science is causing a new look at medical and care records, different types of record are required, not only to support the recording of normalized data, which have minimal error in relation to both noise and silence, but also to help create other types of record over the whole medical area, which provide adequate theoretical support for medical decision making and judgment. While the individual consultant and senior nurse deal with the problem of ensuring that the data gathered by others at the bedside are correct, the operational transformation is carried out in the oral tradition rather than by a written recording tradition. To change this mode of operation means a radical upheaval of attitudes and the direction of senior doctors down new pathways. They will now have not only to check that the input data are cor- rect but also to ensure that what has been recorded and stored is correct in the system. This means much more pressure toward the development of a more adequate medical language and translation from patient observation to precise statements in that new language. The medical vocabulary is large and dynamic and impinges on many scien- tific disciplines, as well as expressing its own special area. So far the problems of literature storage and retrieval, which are formidable in themselves, have made some progress, but we are far from a satisfactory solution. It is not surprising, therefore, that in the more complex field of medical records, progress is no faster and often more uncertain.

While the medical record will be processed in a literal sense, for assisting nursing and medical communication purposes, it needs to have an appropriate data structure designed for this purpose. There is also a massive need for data analysis to assist in the solution of the significant problems that occur during the formulation of diagnosis, in- vestigation, and treatment procedures. Here there is a requirement for a large and ac- curate data base and appropriate algorithms, such as those that are used in mul- tivariate analysis and other analytical techniques to assist in data evaluation. However, unless the data collection is reliable, valid, and normalized w i.e., collected within a frame of reference that is the same for all data gatherers m the records will not be worth the analytical effort to which they are submitted and the usefulness of such expensive analytical procedures to support medical decision will not be realized. What is really being said is that medicine has to come out of its individualist en- trepreneurial system and adopt cooperative ventures. This has to be inculcated not later in professional life but in medical school, and coordinated with clinical education and practice, so that doctors become more cooperative and interdependent. Only in this way are we going to lay the cornerstones for improvement in information process- ing and communication.

378 Anderson

P A T I E N T CARE I N F O R M A T I O N S Y S T E M S IN H O S P I T A L S

In considering clinical data processing in hospitals, some of the achievements of the London Hospital, the Birmingham Hospital (both teaching hospitals), and other hospitals in the United Kingdom, including those in Scotland, will be chosen for com- ment. This is not intended to be a complete survey but to raise important points about their contributions.

The London Hospital project (2) began in 1968-69 and has a Univac computer system with 62 visual display units, 1 situated in every ward of the 700-bed hospital. It also has a small number of printers in sites where hard copy is used. Apart from its patient administration and scheduling functions, the information system allows for requesting and reporting of patient investigations in the hematological, microbiological, and clinical chemistry departments. Thus it provides a clinical infor- mation link between the patients in the wards and laboratory investigations across the whole hospital. There is a paper backup ordering system when the computer system is not available, but during the ordinary working day the computer is used for this im- portant medical activity.

Naturally, it is a real-time system and is used by medical and nursing staff in the wards. After the initial special training for users that went on in the early years, they have come to recognize that it is part of the ordinary training of the hospital for all personnel to learn to use the computer, and this has been incorporated into the or- dinary nursing training system. Over the years what the doctors at the London Hospital have come to accept is that the computer is useful in meeting their requests, providing there is a rapid response at the terminal of less than 3 seconds. Indeed, they have been anxious to go on and expand the system further and are fighting hard to do this.

Some difficulties have come to light about the long lead time before systems are fully developed and implemented and they are several years behind their official schedule. This, however, isnot unusual as the lead times to implement large systems have been grossly underestimated in the past in all areas. Further problems are now being encountered due to the increasing sophistication of their system. As the system becomes more elaborate, programming and system maintenance requires the withdrawal of personnel from the main thrust of the project. For example, they have gone on to produce a drug information and advice system for all the wards, to advise doctors about drug complications and incompatabilities. The introduction of this system has been delayed because of the maintenance problem. It is necessary, therefore, if systems are to expand in this way, to appreciate the difficulties that arise. Perhaps this kind of development can be conducted in other ways by transferring systems developed elsewhere, which will minimize this, and further experiments to deal with this problem must be undertaken.

A similar program has been undertaken at the Birmingham teaching hospital where they have used Univac equipment. They have also had difficulty with meeting planned schedules and are at present just introducing a treatment drug system across the wards of the hospital. In other hospitals in the United Kingdom, less emphasis has been placed on inpatient clinical record problems and more on outpatient problems, especially in the General Hospital at Stoke-on-Trent. Other hospitals have concen-

Clinical Computer Systems in the United Kingdom 379

trated on treatment systems in particular, especially at the University Hospital at Ninewells near Dundee in Scotland, where they have had a long tradition in research work in this area. Other hospitals have used optical character reading to capture patient data, for both patient administration and clinical test requesting.

The medical record project at King's College Hospital was discontinued in 1975, but work on models of medical records has continued to develop the ideas resulting from the initial experiment and to increase our understanding of the difficulties that were met. Indeed, the insight gained by the experimental modeling has enabled some of the complicated interactions of medical records to be more fully understood. Thus we can plan and develop record systems on a more scientific basis and submit this to experimental evaluation, before launching whole hospital-wide systems. It is now possible to design summaries of medical records to serve different objectives for different users. Also, the different types of record to meet different objectives may in fact be covered by different types of systems, including special hardware and software designs. The systems that are needed for communication and medicolegal purposes are certainly not the same as those that are useful to support medical decision making. It may well be in the future that we will use distributed processing with different kinds of systems for different purposes, rather than try and embody allthe requirements in a single central large machine.

CLINICAL N U R S I N G R E C O R D S

It would be wrong not to record the important advances in nursing records concerning nursing orders about patients, for these have been the subject of experi- ment in at least three major hospital centers in the United Kingdom. An extensive nursing record was implemented in King's College Hospital in the initial part of the medical record project in the latter half of the 1960s and it was eminently successful in guiding the nursing aspects of patient care (1). A subsidiary project dealing with a pa- tient diary system was linked with this. This system was discontinued with the medical record system, although the computer was retained for patient administration and other administrative functions and is still in action nearly a decade later.

The basic ideas about nursing order systems have been taken up and developed by the Group at the Exeter Community Health Centres project, in the London and Bir- mingham projects, and in the teaching hospital at Ninewells in Scotland and elsewhere (3). Basically such systems allow the nurse in charge of patients in a ward to predict the nursing requirements for the next day and issue appropriate orders to staff. Nursing tasks can either be scheduled on a patient basis, where all that is required for an individual patient can be listed and a small team can take care of a group of patients, or they can be scheduled across patients so that a nurse can do a specific task for all patients. In reality the type of system used on wards depends a great deal on the number of nursing staff available and the existing work load. The computer system of course can suit its output to this and make the appropriate presentation.

At King's College Hospital we also experimented with a patient diary because of the importance of encouraging self-help in patients. This is very easily derived from the nursing order system and was a printout of what would happen to the individual

Anderson

S Y S T E M S IN H O S P I T A L S

[ng in hospitals, some of the achievements of "I0spital (both teaching hospitals), and other ng those in Scotland, will be chosen for com- te survey but to raise important points about

gan in 1968-69 and has a Univac computer ated in every ward of the 700-bed hospital. It tes where hard copy is used. Apart from its nctions, the information system allows for : investigations in the hematological, ,partments. Thus it provides a clinical infor- rds and laboratory investigations across the "dering system when the computer system is rking day the computer is used for this im-

is used by medical and nursing staff in the users that went on in the early years, they

he ordinary training of the hospital for all nd this has been incorporated into the or- ,~ years what the doctors at the London ~mputer is useful in meeting their requests, ,rminal of less than 3 seconds. Indeed, they system further and are fighting hard to do

9out the long lead time before systems are :y are several years behind their official :he lead times to implement large systems st in all areas. Further problems are now bhistication of their system. As the system and system maintenance requires the

tst of the project. For example, they have advice system for all the wards, to advise ompatabilities. The introduction of this maintenance problem. It is necessary,

y, to appreciate the difficulties that arise. :onducted in other ways by transferring inimize this, and further experiments to

:n at the Birmingham teaching hospital ey have also had difficulty with meeting �9 oducing a treatment drug system across I the United Kingdom, less emphasis has ,lems and more on outpatient problems, on-Trent. Other hospitals have concen-

Anderson

.~d that this was useful and is being oup. Most of the nursing systems are responding fairly immediately w, however, we are experimenting ower but is cheaper to operate as a nany of the nursing requirements. ~n to nursing work and determine leted and what may not have been he nursing staff. Undoubtedly this .ation is not only more reliable and ter. Thus nursing staff have more he ward and spend less time com- r. The acknowledgment system also ~g of nursing work loads. The data milar data obtained by Dr. Spencer trch at Baylor College of Medicine,

:ECORDS

=titioner clinical records and several .ystem. Initially there was a Guy's ztice and this seems to have had dif- y unit in the practice linked to the achieve too much in a short space of practice insofar as it assisted their group general practice record com-

en much more successful in recording 'ecord and archives of all the medical

:y, and clarity. It relies on a real-time :s in the surgeries and microfiche for surgery (office) and home visits. Such t consultation. The records are struc- :ed. The records maintain the existing t only in direct patient care but for are also used in long-term disorders veil as textual statements and free text :nt care in the hospital, as well as chives. The system prints out patient aonitors when renewal is required. The ~mprehensive. It is not possible to go ystem but what has been outlined has

:tives of such projects are fairly well sfer systems from other worlds such as

Clinical Computer Systems in the United Kingdom 381

the hospital into general practice. General practice has its own special record re- quirements and these need appropriate study. However, the Exeter project and several others that have been undertaken by general practitioners are providing a landmark in progress about the usefulness of such records, not only in assisting the general prac- titioner to care for his patients but also in collecting important data about the mor- bidity in the community health area. The major problem with such projects to date has been the large capital cost of the equipment that has to be used. However, with the microprocessor revolution an entirely new approach can be designed. It is expected that within the next decade there will be great progress in this area.

CLINICAL A S P E C T S OF P R E V E N T I V E M E D I C I N E

One of the great strengths of medical practice in the United Kingdom has been the preventive aspects of care, which were originally supervised by Departments of Public Health and, following the reorganization of the National Health Service, are now the responsibility of the district and area community physicians. Clinical records are maintained on all children from birth and are correlated with school medical records and those that continue into early adult life, the record system usually stop- ping at the age of 21. I will particularly comment about the project in the Leicester Health Area, led by Professor A. Buchan, one of the many systems that followed the pioneering work in West Sussex.

The system begins with the details of the birth record, both of births inside the Leicester Health Area, AHA, and of those outside where the birth is related to per- sons normally resident within the area. This information is based on records held by the registrar of births and deaths and this begins the patient's medical health record. Linked to this record are both the administrative and clinical arrangements made with the general practitioners by the AHA for immunization and vaccination procedures and their results. Consent forms are given to the parents and the results are recorded, and there is a follow-up by a health visitor if no return is made. As you may know, Leicester itself has a high percentage of recent immigrants. Health visitors are notified to visit the homes of those who do not appear for immunization at the general prac- titioner's office at the stated session. Thus follow-up and further education and new appointments are essential if a high rate of successful immunization is to be main- tained. Already in some areas of the United Kingdom immunization rates are so low that some diseases are reappearing. The recent publicity about whooping cough vac- cination has meant fewer immunizations and now whooping cough has reappeared in young children and has its usual mortality rate. Thus there is no room for complacen- cy; adequate systems of preventive records must be maintained. New technology is here but in many areas the will to use it is lacking.

Later records about handicap and disability are made, as are surveys of groups of children in the preschool period. School medical examination records are kept on the same system and reports are sent to the general practitioner when a significant abnor- mality is detected. Records of disabled patients continue at present after the school period and until they are 21 years of age.

This is a brief and rather inadequate survey of what is done but such records allow a monitoring of infectious disease; prevention and prompt attention is paid to

382 Anderson

infectious disease occurring in nonimmunized patients. The National Morbidity survey has already revealed that most children will

spend at least one night in a hospital before going to school, so the onset of illness in this period of life is still very significant. Such records also help to provide evidence for the public debate that has to go on about the safety of immunization with agents such as whooping cough and measles and what we must do for those who are the unfor- tunate sufferers from a rare complication. These are both political and ethical issues with which society must grapple but it can only do so if there are adequate records. In this way we can at least approach the problem with the hope of a sensible solution.

School health examinations are important and more effort is needed here to en- sure that both nutritional and social circumstances are not conflicting with natural development. There is still a great deal to be done but we will not see this unless we have adequate records, held by both community physicians and general practitioners. What is already apparent in some authorities with poor systems is that there is some considerable lack of interest. It is sad in this area where the United Kingdom has made fundamental contributions that we are now again seeing patients who have what is a preventable disease.

C O N C L U S I O N S

It is not possible in a brief survey of this nature to give anything but a hint of what has been achieved in the area of health care clinical records. Progress has been slower in the clinical area because of its very different nature from the business and other worlds in which computer systems have already proved useful. Initially it is essential to delineate needs and set objectives, which will become more realistic with practical experience. The long lead times and difficulties with projects and implementation only serve to emphasize the different and novel features that are required of such a system. One major flaw in the research and development has been the lack of any emphasis on the education of users, and of all health care workers in general, about information techniques that will become commonplace and accepted in the next decade. In many projects funding was denied for these purposes and yet the success or failure of a proj- ect may depend on such important areas as this. Already in the London Hospital education in medical information processing and in the uses of computers has become part of the routine training of every nurse. In London University some teaching and training in medical informatics is part of the medical course, although there is no separate examination and it may be taught either in the preclinical basic science area or in the clinical period.

In the introduction some general problems were illustrated reflecting the diverse definitions of medical records and the multiple purposes for which a "universal" record might be used. We have to move away from the area of high expectations and low or little achievement to the new areas where less ambitious but more effective systems will yield some useful and helpful results in selected areas. All too often com- puter systems are peddled by manufacturers as the panacea for all problems. The problems and paradoxes in health care records will remain with us for a long time, but

Clinical Computer Systems in the United Kingdom 383

much can be done to advance solutions with new hardware and software that are now being developed. The impact of the microprocessing revolution has still to be realized, but it is on its way.

The fundamental problem that is always raised is that doctors in general will not use such a system. Indeed, it has to be admitted that with the present emphasis in medical education on producing an individual entrepreneurial skilled doctor with high expertise, he naturally as a corollary must emphasize his individual qualities; it follows also that lack of cooperation and difficulties in relation to teaching method and poor communication will naturally be present. Any change of attitudes, and one is certainly required in medical data processing, has to begin at the grass roots and these lie in the area of medical education in our medical schools. Most medical schools in the United Kingdom have some form of computer education now. Slow but steady study is emphasizing the need for cooperative ventures and for a common medi6al language. We must now emphasize much more the interdependence that is required in medicine, rather than its individual entrepreneurial aspects.

Investment in research in this particular area has been very small, and in the United Kingdom, because of our present financial stringency, it is likely to remain so for some considerable time. Nevertheless, without research we perish and those in- terested will go ahead, although at a slower pace and perhaps in different directions from the one they might otherwise have taken.

Finally, I should like to pay tribute to the many medical and scientific personnel, especially those in medical data processing, who have cooperated in developing the systems that I have reported and who have battled with difficulties on all fronts and won through. They have survived change as well as trials and this record could not have been made without their effort. Long may it continue.

R E F E R E N C E S

1. Anderson, J., Chapter 16. Hospital Computer Systems (M. F. Collen, ed.) Wiley, New York, 1974. 2. Barber, B., Cohen, R. D., and Scholes, M., A review of the London Hospital computer project. Med

Informatics 1:61, 1976. 3. Ashford, J. R., and Pearson, N. G., A community based medical computing system. Community

Health 3:5, 1977. 4. Bradshaw Smith, J. H., A computer record-keeping system for general practise. Brit. Med. J. 1:1395,

1976. 5. Henney, C. R. Bosworth, R., Brown, N., and Crooks, J. 953 Medinfo 77. (E. B. Shires and E. Wolf,

eds.) North Holland Publishing Company, Amsterdam, 1977.