An introduction to data analysis
Having sketched out the context for my work I will now introduce the first data extract (Extract 1), some ethnographic fieldnotes from my observation of a nurse in her clinic for patients with coronary heart disease (CHD). Regular review of patients with chronic diseases such as CHD has been identified as a crucial element of high quality care (Wagner, Austin, & Von Korff 1996) and the registration and recall of these patients is made easier by the EPR. Like all of the nurse-led chronic disease consultations I observed, the reviews involved the systematic completion of a computer template (electronic form) identifying tasks to be done and data fields to complete.
Extract 1. A morning in the Coronary Heart Disease (CHD) clinic
We were between patients and there was a 20 minute gap as a patient hadn't shown for his appointment. The nurse started to check some cholesterol results on the computer, referring to an in-house guideline printed on a laminated sheet. Suddenly the screen froze. The system had crashed.
She jumped from her chair and rushed out into the corridor where she was met by a secretary who had also left her desk in a panic because the usual IT person was not in today. The nurse returned and said she couldn't get on with what she wanted to do. I followed the secretary downstairs to the reception area.
The tiny office next to reception was soon full. The secretary was on the phone talking hurriedly to the IT supplier, and two of the GPs were kneeling on the floor around the server, bums in the air, fiddling with buttons, while an alarm sounded. Another GP looked on from the sidelines joking about the reliability of IT. One GP stayed in his room and didn't join this impromptu meeting round the server. The receptionists kept themselves to themselves but one of them asked me quietly 'Does this never happen in your place?'
I overheard the secretary saying 'One of our doctors thinks it's the UBS' only to be corrected by the doctor whispering 'not the U B S, the U P S.' I discovered this meant the uninterruptable power supply, which struck me as a misnomer; it was certainly causing plenty of interruption. Chaos really.
The receptionists were a bit stuck. Patients kept arriving but they didn't know who to expect and couldn't 'arrive' them (meaning mark an A next to their name on the appointments list to indicate that they were waiting). The waiting room was filling up.
After a few minutes, some lights started flashing on what may have been the UPS and there was a visible collective sigh of relief amongst the GPs. The secretary was still talking to the IT supplier but the GPs returned to their rooms to resume surgery.
I went back to the nurse's room. The screen said 'connecting' but did not appear to be connecting in any meaningful way. The nurse was flustered now and went downstairs to try to find out who her next patient was. As she followed the patient up the stairs I heard her warning the patient 'We've got a problem today 'cos the computer has crashed and isn't working.'
The patient sat down. The nurse began by saying 'I'll have to do it a little out of order because I've no computer.' She grabbed a yellow post-it note and wrote the patient's name at the top. The patient gave her a urine sample for testing. The nurse said it was fine and scribbled on the post-it note. She leaned over the corner of her desk towards the patient asking 'Do you know which medicines you are on from a cardiac point of view?' A familiar opening which I had by now come to recognise, although this time I could not help noticing that for the first time it was the patient rather than the computer screen to whom the question was directed. The patient - smartly dressed and well-spoken - put her handbag on her knee and said politely 'I'm prepared for all eventualities, my dear' as she produced a list of her repeat medications and handed it to the nurse. Reading down the list the nurse said 'So...from a cardiac point of view you're on...nicorandil, isosorbide mononitrate, atorvastatin, diltiazem. Are you on aspirin?' The patient said 'they' had stopped it because she bruises too easily, and then added that one of her medications had recently been increased during a hospital admission. The nurse handed the list back and turned to the computer, then typed a few keystrokes to see if the computer was working but it just bleeped and remained frozen. There was no further discussion about the medication or the admission.
The nurse took the patient's blood pressure, there was a brief discussion about exercise then the nurse announced 'This is so confusing not having the computer...uuuuuhm... (long pause)...diet... do you have a balanced diet?' Then 'What I think I had better do is your blood test, and just hope we are back on line after that. It just goes to show how we rely on computers.' She kept checking and rechecking the computer. Blood sample taken, she returned to her desk saying 'let's see if we have any joy (types keystrokes) OOOooh that looks encouraging.' She leaned towards the computer and said to it 'c'mon you can do it.' She typed in a password but nothing happened. 'Oh that looked so promising. Oh that is such a shame. We're so close. I'll just go downstairs and see if it is just me.' The nurse left the room and I chatted with the patient until the nurse returned about 5 minutes later.
After 25 minutes of downtime the computer came back to life. The nurse turned to it and said 'Let's see if we've got anything from your recent hospital admission' and opened up a hospital letter. She read it quietly and said to the patient 'That doesn't say anything about you increasing the medication.' The patient replied 'they did' to which the nurse responded 'I'm not disbelieving you' then turned to the computer again and sighed 'it's gone again'. The patient looked down at her repeat medication list on her lap and said that it was the nicorandil which was increased. The nurse responded 'Sadly our return to the computer was only temporary so I can't do anything at the moment. I'll go and have a chat with Dr Vaughan as the cardiologists haven't organised any follow up. So since they increased the nicorandil how much have you been using your spray?' Patient replied: 'Ooooo a lot less, only a third'
The nurse apologised saying 'I'm sorry it's been such a higgledy-piggledy consultation' and left the room again to speak to the patient's GP, returning with the advice that she should stay on the same dose of medication as it was the maximum dose and seemed to be helping. She made a note on her post-it note 'nicorandil |30mg'.
At the close of the consultation the nurse apologised again 'I'm sorry. It was a bit of a come and go consultation' to which the patient replied 'WELL DONE' then added gently '...you can go off computers.'
The nurse was running 30 minutes late by the time she was ready to see her next patient.
It may seem surprising to begin with a narrative account of one of the few occasions in my data collection when the EPR became temporarily unavailable. However, this brief experience reveals the extent to which the EPR has become embedded in practice and is a useful point of departure for highlighting some important methodological and analytic concerns.
The nurse's interaction with the patient conveys a strong sense that the cardiovascular check should be an orderly affair, and that the order of prompts and fields inscribed in the computer template is the 'right' order of conducting the clinic, warranting apology if things have to be done 'a little out of order'. She leaves the room twice, and the consultation becomes - in her words - 'higgledy-piggledy' and 'come and go'. The nurse's reference in her opening question to 'the cardiac point of view' was a common way of framing the scope of the chronic disease consultation. This, and other similar formulations (e.g. 'SO, ... cardiac review') establish what Goodwin refers to as 'figure' and 'ground' - what is more (and less) relevant to the professional activity taking place (Goodwin 1994). Implicit is an assumption that the discernment by patients (and nurses) of symptoms of one chronic disease from another is unproblematic; this was taken for granted even in the face of the most complex multimorbidity (Swinglehurst et al 2012).
Although I was not surprised that it was disruptive and stressful when usual routines broke down (especially with a researcher observing), this incident brought home to me the extent to which nursing care had become interwoven with technology use. The difficulty was not merely that the nurse could not access the patient's medical notes. The notes - it turned out - were less reliable than the patient's account, at least with respect to her medication. But without the template, the nurse found it difficult to 'go on' - 'Sadly our return to the computer was only temporary so I can't do anything at the moment'. Neither does it seem likely that this senior, well qualified nurse cannot do a cardiovascular check without the prompts before her eyes. It seems much more likely that it is because her embodied practices have become so finely tuned to incorporate the technology that to conduct the clinic without it has become almost impossible (Swinglehurst et al 2012).
Garfinkel, in an early seminal text on medical records identified the handling of emerging local contingencies, answering the immediate question of 'what to do next?' as one of the main concerns of clinical work (Garfinkel 1967). As an observer in this clinic, my sense was that the consultation could not progress without nurse, patient and (working) template all co-present, and that it was often the template which prompted 'what to do next'.
This recursive and inextricable relationship between the EPR and contingent social practices was a significant early insight in this research project, and one which informed my decision to use a linguistic ethnographic approach. The EPR is not simply a piece of technical kit reducible to a collection of hardware and software sitting on the clinician's desk, but an example of a 'social substance' definable only in terms of the properties of a social world (Harre 2002). Whereas most previous studies of healthcare interaction either fail to incorporate the EPR at all, or 'contain' it by focusing on the 'computer' as if it were a separate entity, I was keen to develop an approach which engaged with the EPR as integral to interaction and social practice. If, for example, I had simply wished to consider the impact of the computer on the nurse-patient interaction, the observations detailed above may have been overlooked, as the computer was 'out of use' and the institutional context might be deemed relatively unimportant. However, by orienting to the EPR as part of a complex of interconnected social practices and institutional context, linguistic ethnography opens up richer appreciations of what is actually going on.
These ethnographic fieldnotes offer a glimpse into the organisationwide nature of the disruption, a consequence of what Iedema has referred to as the wide 'organisational reach' of the EPR (Iedema 2003) and its inter-connectedness. As Susan Leigh Star has noted, when the server is down 'the normally invisible quality of working infrastructure becomes visible’ becoming 'the basis for a much more detailed understanding of the relational nature of infrastructure' (Star 1999: 382). What I was witnessing was a critical incident in which the usual activities of the clinic were (largely) suspended. It led me to the uncomfortable conclusion that in extreme - and thankfully rare - situations, the EPR becomes the patient, at least inasmuch as attending to it and correcting its ills becomes the highest and most urgent priority amongst clinical staff.
I hope this brief account already conveys a sense of the way in which the EPR contributes to the regimentation of the clinic, and how - even in its absence - it infiltrates the discourse of the consultation. Having taken a look through what Erickson calls the 'social telescope', I will now zoom in with a 'social microscope' to look in detail at some interactions (Erickson 2004). Again, I will use this as a way of highlighting some of the analytic challenges presented by the EPR, as well as building on the notion of the EPR as 'regimenting' practice.