HEEDING, MEASURING, UTILISING: THE INFORMATICS TEMPLATE An explicit working definition for informatics. Barry T Jones BSc PhD Senior Lecturer Department of Psychology (Nurse Information Processing Group) University of Glasgow GLASGOW May 10th 1992 To appear in International Journal of Health Informatics page 1.
HEEDING, MEASURING, UTILISING: THE INFORMATICS TEMPLATE An explicit working definition for informatics. Designing courses in informatics for health professionals is not straightforward. The range in content is large and the range in commitmentto actively include informatics within professional education is no less. This is as true for Europe as for the US; as true for social work as for nursing and medicine; as true for basic education as for advanced; as true for practice as for management. One reason for this, perhaps, is that it is not clear what informatics should be. Consequently, it is often not clear what should be taught and it is often not clear why it should be taught. Refuge from this difficulty is frequently sought in early definitions (informatics has yet to reach its teens) but like early definitions everywhere they are notorious for their deceptive simplicity. As definitions evolve, however, they become more intricate and it becomes no easier to identify the skills that they represent and that newcomers should try acquire. I have tried to side-step the origins and evolution of the definition and simply equate informatics with informing since, curiously, I find I am much more comfortable with the process of informing than with the commodity information. More curious still, I find this observation is common. I propose a working definition of informatics which leans more on the notion of researchmindedness (informing) than on computing (information). It is a working definition that has its origins in a paper I presented at the Current Perspectives in Health Computing Conference in Brighton UK in May 1988 1. It will begin with an assertion and then I will explain what the assertion means. Assertion. Within any domain, informatics describes the wide range of activities that together attempt to solve the representation problem. This, of course, begs two questions: What is the representation problem? What sort of activities attempt to provide solutions? What is the representation problem? To answer this I will use a topical, illustrative example. (It is, in fact, topical both in terms of current activities within health care, itself, and also in terms of much of the content of the page 2.
current issue of this journal.) Consider the task of nursing workload/manpower management in a single hospital ward. It is not difficult to state the ward management goal: it is to get together the right number of nurses with the right mix of skills doing the right things so that care is delivered, at any time, appropriate to the particular needs of the patients in the ward. While this goal is simply stated, it is unfortunately, in practice, a difficult one to achieve since care demand is a constantly moving target and care supply is variously constrained. Of course, if a single omnipotent ward manager were permanently suspended above the ward with all-seeing eyes and total recall; with perfect intuition that impeccably knows and able to instantly command the deployment of staff following equally perfect judgements, then ward management would in practice present no problems. Regretfully, no such animal exists and it is this that creates the representation problem. It is a problem that, in the absence of the omnipotent, comes from the need to record (or represent) features of the ward and to subsequently use this record (or representation) to make decisions, in this case management decisions. If ward management requires the use of representation, what might this representation look like? There are two styles: the record or representation used might well be an informal one held within the ward manager s mind and management might subsequently be carried out along intuitive lines - or, on the other hand, it might be written down and the management style employed might be much more formal. Whatever the style, the representation problem is the same and (to put it in a more focussed fashion) it is the problem of knowing answers to the questions: 1. What features should be heeded (to inform)... 2. How should they be measured (to inform)... 3. How should the measures be utilised (to inform)......inform and, thereby, effect good management? Within the working defintion developed here, informatics simply describes the wide range of activities that attempts to provide answers to each of these three questions - or, in other words, to attempts to solve the (now more clearly focussed) representation problem. page 3.
What sort of activities attempt to provide solutions? Domain knowledge (nursing knowledge, in this particular example) is clearly crucially important at each of these three levels for the following reasons: 1. Some features of the ward must surely be more important to heed than others; 2. Some ways of measuring these particular features must surely be more appropriate than others; 3. Some ways of combining and utilising these measures must surely be more effective than others. It is through the application of domain knowledge (nursing knowledge, here) that each of these three questions would normally be answered. Often, however, domain knowledge and the received wisdom that accompanies it cannot guarantee to provide the best (or even a good) answer and procedures need to be adopted or developed to address this difficulty 2. This difficulty is no straw man. Take for example the current case: what ward features should be heeded to inform? If, for illustrative purposes, we consider the problem of representing only care need (ignoring care provision), then should attention be paid to (among many others) the ward speciality, to the individual patient, or to the particular shift or the day of week? If it were decided that individual patients should best be heeded, then should their sheer numbers be measured or some measure of their individual and specific needs? If the latter, then what particular measure of individual need should be used? Finally, how should individual measures of need be combined or aggregated up to ward, speciality or hospital level? Of course, there are many contending ward features, many contending methods of quantifying them and many contending ways of combining and utilising them for managing wards - indeed, the very wide range of local solutions testisfies to this! Within the working defintion advanced here, informatics describes the activities that both research and develop the choice and evaluation of (i) these features, (ii) their measures and (iii) the uses to which they are put inasmuch as they impact ward management - and that this is done in an effort to find (i) the most important, (ii) the most appropriate and (iii) the most effective ones. The informatics template, heeding-measuring-utilising, and the research, development and evaluation framework that it entails extends, of course, far beyond the task of managing wards: page 4.
What form should health education take in a particular ethnic minority community? How should hospital performance indicators be established and how should they be used? How should motivation to quit in problem drinkers be assessed by therapists. What should a ward nursing information system do and how can the best one be identified? What should an exit questionnaire contain and where/when should it be applied? What quality of community life is there for clients discharged from psychiatric hospitals? How can child abuse be best identified earlier? How can a good information strategy for a unit, a health authority a nation be devised? These represent just some of the areas of professional care within which measuring is of some consequence to the proper functioning of the services that have been created to meet the needs encountered there. The view developed in this paper is that measuring is at the very centre of informatics and for this reason it features centrally within the informatics template introduced here - taking its input from heeding and providing an output for utilising. Computing, information technology and informatics. What role have computers and information technology (IT) to play within the area this paper describes as informatics? This is an important question to ask for there are many who regard informatics as primarily computing or computer-use 3. If IT were used synonymously with computing, then I would argue that the role it has to play would be a fairly small (but not necessarily insignificant) one. More important, perhaps, not only would it be small, it would be inadvertently distracting - diverting attention, effort and resources away from what I see as the more fundamental issues in informatics (such as finding a good representation) towards less important ones (such as finding any old representation, but one that was simply more conveniently or more quickly collected and used... computerised). If IT were to refer to those skills, practices or techniques that underpin the informatics template heeding-measuring-utilising (of which computing, of course, might often have a part to play), then I would argue that IT would have a very big role, indeed. However, it is unfortunate that the meaning of any word is firmly set in terms of the way it is most commonly used and there is little doubt that for 99% of professional carers, the word IT simply means computing or computer-use. For this reason, therefore, just as I do not think informatics can page 5.
profitably be thought of as computing or computer-use, I do not think it is at all helpful to use the term IT in relation to informatics either (for it implies computing is fundamental to informatics and, as the informatics template is designed to illustrate, I do not think it is). Graves and Corcoran (and nursing informatics). While it is part of the defensible rhetoric of care educators, scholars, practitioners and managers alike to recognise that the term information in IT is more important than the technology (and the informatics template is meant to capture this), the paradox is that perhaps the most widely-quoted statement about the nature of contemporary informatics appears to advocate something close to the contrary 4 - identifying computer science and computing as a critical (nursing) informatics component, both as goal and driving force. Within Graves and Corcoran s most scholarly view of nursing informatics 4, computing (for them, the construction of expert systems and information systems) is seen as important both as an end in itself and as a means to force nursing to develop and organise a better data/ information/knowledge-base (they do this couched in terms of nursing diagnosis and minimum data sets). This looks attractive but the value of many important aspects of the expert/ information system departure within nursing is not beyond doubt 5,6 and, in terms of advancing a discipline (nursing), is building machinery for use in a particular area of nursing really the best method to encourage the advancement of that particular area? For, as ever, the danger is to mould the nature and direction of the discipline to suit the machinery rather than the reverse. Final note. The history of the scope of informatics clearly puts computing close to the centre of effort 4,7. But static definitions become sterile and can easily cease to serve crusades well - crusades such as how carers can become better informed to better care. The view promoted in this paper is that this crusade might be more appropriately served through activities that seek to provide and use better representations, better measures - activities that try solve the representation problem. Of course, some of these activities might well involve computers but computers are neither a necessary nor a sufficient component for an activity to be classified as informatics. The informatics template has been introduced to make this view both accessible and explicit and the template is proposed as a defintion that can be more readily used on the ground. It is not meant to be a static one. page 6.
REFERENCES 1. Jones BT, Hyslop A, Ritchie I and Welch J Bypassing problems with the establishment of nursing levels: modelling good management knowledge in an expert system based upon automated rule induction. In Starling et al (Eds) Pulling Together: Current Perspectives in Health Computing. Weybridge: BJHC Books 1988; 185-190 (Abridged version of presented paper). 2. Walsh M and Ford P Nursing: Rituals, research and rational actions London: Heinemann 1989 3. Ball MJ and Douglas JV Integrating nursing and informatics. In Ball MJ, Hannah KJ, Gerdin Jelger U, Peterson H (Eds) Nursing Informatics New York: Springer-Verlag 1988; Section 1 Unit 1 Chapter 2 4. Graves J and Corcoran S 1989 The study of nursing informatics IMAGE: Journal of Nursing Scholarship 21(4) 227-231 5. Jones BT 1992 Expert systems in nursing: the wide chasm between practice and promise. Information Technology in Nursing 4(2) in press 6. Hoy, D 1990 Computer assisted nursing care planning in the UK Information Technology in Nursing 2(2) 22-23 7. Gorn S Informatics (computer and information science): its ideology, methodology and sociology. In Machlup F, Mansfield U (Eds) The study of information: interdisciplinary messages New York: Wiley 1983; 121-139 page 7.