Circulatory system disease G Ischaemic heart disease G3 Acute myocardial infarction G30 Anterior myocardial infarct G301 T476 Hit by boat water skiing U1287 Bitten or struck by crocodile or alligator, occurrence on farm T5408 Hit by aircraft, without accident to aircraft, member of ground crew or airline employee injured T55z1 spacecraft accident NOS, member of ground crew injured Data Recording in Primary Care Field Studies Patient Records Enhancement Project Lesley Axelrod Human Centred Technology Lab Interactive Systems Group, School of Informatics University of Sussex, Falmer, Sussex, BN1 9QH UK l.axelrod@sussex.ac.uk Flis Henwood Social Informatics Research Unit School of Applied Social Science, University of Brighton Falmer, Brighton, Sussex BN1 9PH UK Geraldine Fitzpatrick Human-Computer Interaction Group Technical University of Vienna Karlsplatz 13, 1040 Vienna, Austria Jackie Cassell Helen Smith Amanda Nicholson Brighton and Sussex Medical School Mayfield House, Falmer, Brighton, Sussex BN1 9PH UK Greta Rait Department of Primary Care and Population Health, University College London Medical School, London NW3 2NF UK Abstract—This position paper describes the Human-Computer Interaction (HCI) field studies component of the multidisciplinary Patient Records Enhancement Project (PREP). PREP seeks to understand variability of data found in primary care electronic records, in particular the balance between coded data and doctor’s ‘free text’ notes. HCI fieldwork will establish variables that affect recording practices. In field studies we observe and record data recording practices in general practice (GP) surgeries, interview staff, video consultations with real patients and video consultations with standardized patients (played by medical actors). By standardizing patients we can compare the impact of other variables: different doctors, in different surgeries, using different e-health systems. Our early findings suggest that variability is due to a complex web of reasons, driven by personal, contextual and organizational processes. Findings from thematic analysis will result in design implications for studies by epidemiologists and public health researchers, design of NHS training and work processes, and design of electronic health record interfaces. Keywords: electronic health record; primary care; field study; human-computer interaction. I. INTRODUCTION The electronic health record is at the centre of pervasive health care management. There is a rapid increase in use of pervasive information technologies in healthcare settings, which is set to transform the sector. HCI research has a critical and growing role to play in understanding the challenges and opportunities about how such technologies function in practice, in order to inform design of systems that are effective, not only in budgetary terms but also in provision of care. Electronic health records are well established in National Health Service (NHS) primary care settings in the UK with General Practitioners (GPs) and their surgery staff using a range of approved ‘systems of choice’ [3] to record patient consultations and other patient related interactions, such as incoming letters, test results, prescriptions, etc. One consultation might consist of several entries relating to different elements. Any one entry might consist of coded information or free-text notes, typed in by GPs and other surgery staff, or a mix of both coded and free text data. Read codes (named after James Read [4]) are arranged in hierarchical families, developed to reflect medical thinking. The logic of this coding system is rather lost in the smaller branches – for example there are a plethora of possible accident codes that are rarely used. In practice, a level of coding with four digits is often used. There are codes for diagnoses, symptoms, prescriptions and tests. It is quite possible for doctors to select different levels of codes, or different branches of codes, when referring to the same thing. Figure 1. Examples of hierarchical Read codes. Figure 2. Examples of rarely used accident codes. Doctor’s notes made in free-text are ‘messy’, and lack ‘normal’ sentence structure, e.g. often lacking full stops and Funded by The Wellcome Trust (086105/Z/08/Z)) PervasiveHealth 2011, May 23-26, Dublin, Republic of Ireland Copyright © 2012 ICST DOI 10.4108/icst.pervasivehealth.2011.246073