Progress with measurement, . information and decision-making in critical care medicine An intelligent way forward E. R. Carson, J. J. L. Chelsom and R. Summers Centre for Measurement and Information in Medicine and Department of Systems Science, City University, Northampton Square, London EC1V 0HB, UK This paper focuses upon the need to provide information to the clinician in the critical care unit in order to enhance the decision-making capability. The role of intelligent instrumentation is highlighted, indicating its function in converting data into information which is then interpreted in the clinical context. Examples drawn from the interpretation of blood gas results and the provision of ventilation management are presented. The role of intelligent instrumentation in the wider role of patient management is discussed. Keywords: Intelligent measurement, intelligent instrumentation, critical care medicine, medical decision- support system 1. Introduction Advances in measurement technology have resulted in substantial increases in the amount of data available to clinicians in critical care medicine. More physiological variables can be measured more frequently, highlighting the need for an increased capability for their processing, storage and presentation. The increase in measurement technology over the last 20 years can be observed by looking at the number of data items measurable in a typ- ical Critical Care Unit. Fig 1 shows this increase in data complexity in terms of both the number of raw data items measurable and the extra data which can be de- rived from the raw data set. Today it is not uncommon to find an individual patient record chart which contains over 100 separate data items. When this is combined with frequency of measurement of up to 96 per day, it is possible for the data rate to reach over 4000 data items per day per patient. Not only is there an increase in the number of meas- urements but also the number of alarms - both limit alarms and equipment malfunction alarms - has in- creased substantially. If these results of more powerful measurement technology are to be used to best clinical effect, there is the need to enhance the capability of converting data into information, so that as much sup- 150 100 NUMBER OF _ DATA ITEMS MEASURABLE -- 50 0 ~ ~ ~ ~ W p~A RAMETERS 1970 I I 1975 1980 1985 TIME Fig 1 Illustration of data complexity (adapted from Price and Mason, 1986) 104 Measurement Vol 9 No 3, JuI-Sep 1991