ESRC L218252022
NHS Direct: Patient Empowerment or Dependency?
Full Report of Research Activities and Results
Background
The research problem emerged from the growing importance of a number of debates and the government’s desire to use an old and a new technology (telephone and Clinical Assessment Software – CAS) to develop new ways of accessing healthcare and the health service. Both developments highlighted concerns across a number of academic arenas; in particular, sociology, social policy, socio-technical studies, and health services research. Very briefly, some of these were:
- The importance of debates about modernity, reflexivity, risk and social change (Giddens 1994, Beck 1992 1996, Lash 1993, Lash et al 1996, Adam et al 2000)
- Social policy writers assertions that theories of reflexive modernisation were influencing the state and the way it was allocating welfare services (Jordan 1998, Culpitt 1999, Taylor-Gooby et al 1999).
- Research on how technology is actually used and managed by people in everyday life and how this is enabling and disabling, giving rise to new forms of being social (Whalen et al 1988, Garcia and Parmar 1999).
Although less important, the research was also influenced by literatures on the nature of work, the call centre as a labour site, and the growth of tele-working.
Much of this was reflected in the call of the Innovative Health Technologies Programme. In particular, issues of risk, reflexivity, trust, and the social management of health. From our perspective, these debates coalesce around reflexivity, access to healthcare and ideas of citizenship.
The reflexive modernity thesis suggests citizens and individuals have changed over the past thirty or so years. Whereas once life paths were structured around institutions such as class, the family or gender, today this is less so. Another key part of this thesis suggests we once trusted experts and expertise whereas now we are more sceptical. Supposedly, these changes mean we have more control over our lives. Today, we are better educated, more individualistic and reflexive in our dealings with society and in order to reflect this the state – especially in welfare services – needs to become more responsive to our individual needs, more flexible, and less paternal. It should aim to give citizens more responsibility and choice. If such views are accepted, then the health service (amongst others) needs to shift; it needs to become more heterogeneous, responsive and learning.
Unsurprisingly perhaps, such claims are disputed. Some challenge the thesis theoretically (Lash 1992). Others argue that citizens have not changed – at least in regard to the welfare state (Taylor-Gooby 1999, Lupton and Tulloch 2002). And still others suggest none of this has been empirically tested (Dingwall, 1999). Yet despite these challenges, there is a feeling that the reflexive modernity project is shaping government policy (Jordan 1998, Culpitt 1999).
Paralleling these debates was the formation of NHS Direct. This was part of the state’s desire to use innovative health technologies for ensuring greater and cheaper access to the NHS (Munro et al 1998). This nurse led, telephone advice service was to alleviate pressure on A&E and GP services by helping people to help themselves where possible, to facilitate self-learning and guide people to the most appropriate avenue for care; thereby matching need with resources more efficiently. Implicit within this, is the assumption that citizens can and will take responsibility for their healthcare, use the health service in an appropriate fashion and behave ‘rationally’. In short, there is the view that with easy access via technology, citizens can receive the appropriate level of care, engage in self-learning and use resources efficiently. Thus they will be the responsible, self-learning, empowered citizen of reflexive modernity.
Objectives
Our original proposal outlined three objectives:
(1)Meeting the Needs of Different Groups –
Our research aimed to understand how different groups used the system and whether or not it met user needs. This implies we were interested in the users of NHS Direct and not the non-users. The groups in our research were generally white, professional, and parents. NHS Direct’s own research on users suggests they are overwhelmingly between 30-60 years of age, white and middle class.
(2)What do people do with the advice given and is it seen as appropriate –
This objective was aimed at examining if people used the service reflexively, did they behave in the questioning manner outlined, did people make use of the advice given, when did they make use of it and when not, etc. If NHS Direct is meeting the needs of different groups then this may be further reflected in people adhering to the advice they are given and acting in the manner suggested by health professionals.
(3)Issues of Accountability, Trust and Confidence –
This objective is aimed at establishing how, if at all, trust, confidence and issues of accountability are addressed by NHS Direct. In order for the system to work issues such as these need to be successfully negotiated. Yet, supposedly, in a risk society and in a non-face to face setting, these issues become more difficult to establish whilst paradoxically becoming more important.
Methods
Access was negotiated to two separate organisational sites – one based in a mixed urban and rural area, the other in London. Once this was arranged and medical ethical committees satisfied, we began the research proper. Our methods included:
(1)A team member undergoing a two week training course and work experience as a frontline staff member using CAS
(2)Participant and non-participant observation by team members
(3)Interviewing 60 callers
(4)Interviewing 33 nursing and call handler staff
(5)Conversational Analysis (CA) of transcripts from 120 calls in conjunction with the NHS Direct record sheets of these calls
(6)Presenting workshops to staff and using these as feedback for the analysis of other data
Because of ethical considerations, it was necessary to gain caller consent to use their call and request an interview with them both verbally at the end of the call and in writing after a ‘cooling off’ period. In both instances we relied heavily on NHS Direct staff. All calls are recorded by NHS Direct but we had to ask nurses to seek informed consent for the call and for follow up interviews. It was agreed these questions were posed at the end of the call ensuring people in an emergency, pain or distress were not denied advice immediately. This left the research dependent upon nurse assistance and goodwill.
Consent for calls and interviews were gathered in three phases and the data was collected and analysed thereafter. After the first phase, we remedied two faults in the process – (1) nurses were selecting out certain types of calls for reasons they, not us, deemed appropriate; to remedy this we had to clarify our needs. And (2), nurses were asking all callers in the first week of data collection. This was generating a lot of work and hence we developed call sheets wherein once a quota of ‘call types’ was reached they should stop.
Ethical considerations meant we were dependent on the nurses and it must be assumed some calls, e.g. ones where the call broke down, were screened out as nurses may have felt unable or unwilling to ask for caller participation. However, given ethical issues, it is difficult to see how this could have been avoided. Despite this we fulfilled the interview and CA targets outlined in our proposal and undertook additional data collection which greatly enhanced our understanding of the processes.
Although complex this variety of methods enabled us to cut into our objectives in a range of ways. New data allowed us to develop new ways of thinking about data already collected and to generate new areas of importance for the project as a whole. For example, by carrying out participant observation and undergoing the training course as a call handler, the importance of CAS was realised. As a result, its centrality to the project’s objectives became clear early on. Similarly, comparing and contrasting the data from caller interviews with the CA of the calls, led us to rethink how people engaged with NHS Direct and professional expertise more generally. The use of call sheets from NHS Direct for the call transcripts allowed us to get a clearer sense of how and when nurses deviated from CAS. And nurse discussions of CAS and their use of it highlighted differences between ‘official’ and ‘real’ behaviour (Hanlon (b) et al, Greatbatch et al). Finally, we used the workshops (see ‘Activities’) as a means of garnering feedback on our understanding of what was happening in calls. In both sites, workshops were held wherein nurses listened to a call, read a call transcript and listened to our CA informed analysis of it. Nurses commented on this, oftentimes agreeing with and learning from our interpretations, but equally they gave medical and nursing accounts explaining why a nurse deviated from the protocols. This enriched our understanding of the call. Thus the workshops were a useful data gathering (as well as a useful training) exercise.
As indicated, our methods were flexible, developing as new issues emerged. It was not our intention to undertake certain methods, but as our reflexive understanding of NHS Direct developed, so did our methods. The addition of participant observation, the training course for handlers, collecting call sheets, and using workshops as a source of data gathering were important to getting a deeper understanding of how nurses use CAS and how this shapes the joint production of healthcare, reflexivity, learning, shared understandings, etc.
Results
A lot of material has emerged from the project and what follows is a brief summary of only the most important results. We have grouped them into four areas but each area contains a number of separate issues.
- SHARED UNDERSTANDINGS AND THE JOINT PRODUCTION OF THE SERVICE
Our original proposal argued healthcare is jointly produced between the patient’s ability to understand and describe their illness and the professional’s capacity to interpret this. In NHS Direct this problem is compounded by the fact the service is non-face to face. Hence the need to establish trust and develop shared understandings is paramount (Greatbatch et al). Our work shows that at least three types of misunderstandings may arise, those concerning – (1), the character of the NHS Direct service (2), patients’ symptoms, and (3), the advice or information that nurses offer.
Areas of Misunderstandings:
1. NHS Direct Service
- Callers’ expectations may be at odds with what is officially on offer and/or the expectations of particular nurses. Thus, for example, callers may expect a diagnosis, whereas nurses restrict themselves to triage (although there is often a fine line between allocation/triage and diagnosis/consultation). Or callers may contact NHS Direct with a range of aims that may not be immediately apparent to nurses: e.g. double-checking the explanations/actions of other medical professionals; seeking access to health services; trying to secure transport to medical facilities; attempting to shift responsibility for their own or third parties’ healthcare away from themselves to the NHS; and/or seeking reassurance.
2. Symptoms
Problems may arise regarding the character, location or effects of patients’ symptoms. For example, nurses and patients may experience difficulties achieving mutual understanding with respect to the sensations patients say they are experiencing. Such misunderstandings may arise because the descriptors people use for their symptoms do not have fixed meanings. Additionally, callers’ understandings of human anatomy may differ from those of nurses, and may vary from one caller to another.
- The possibility of misunderstandings between nurses and callers arising is perhaps increased in the context of telephone triage due to the fact that the nurses cannot see or touch patients. In face-to-face interaction medical professionals rely heavily on sight, touch and gesture. These resources are absent in the context of the telephone where professionals work on the basis of what they are told and, in some cases, paralinguistic phenomena such as tone of voice, breathing patterns, etc.
- Nurses must also evaluate the patients’ descriptions and understandings of their conditions. For example, nurses may believe a patient is exaggerating their level of pain or are characterising it incorrectly. (Note also that people may use the same descriptors - e.g. burning, sharp, stinging and the like - to describe different sensations.)
3. Advice/Information
- When nurses offer/relay advice or information the question arises as to whether patients understand the advice/information and its implications. In some cases patients respond in ways that display their understandings of advice/information. In others, however, their responses merely assert that they understand. Such assertions of understanding can mask interpretations of advice/information that may be problematic when viewed from the nurses/NHS Direct’s perspective.
Our analysis suggests that reaching shared understandings is not guaranteed. Rather, it requires work and flexibility because callers are varied in their knowledge, expectations, objectives, capacities, etc. These varied caller responses affect the nature of service delivery because NHS Direct nurses have to depart from organizational routines in order to deal with them (indeed, some callers are becoming familiar with NHS Direct’s routines and CAS protocols and pre-empting them). Thus NHS Direct personnel do not simply produce a service for the consumer. Service delivery is shaped and reshaped during their interactions with callers, who adopt different stances with regard to professional expertise and risk, and who display different understandings of the role of NHS Direct. This feature centralises flexibility.
B.TECHNOLOGY, STANDARDISATION AND PROFESSIONAL AUTONOMY
A key innovative technology within NHS Direct is the use of the computerised decision support software known as CAS. As with most technologies, how CAS is used differs from how its use is envisaged by its designers (Hanlon (b) et al, Greatbatch et al). This is because of A above - it is a response to the joint production of the service which is premised on departing from organisational routines. Thus although NHS Direct’s senior management want CAS to deliver standardised care to the generalised other and to predominate over the expertise and experience of nurses, nurses often diverge from the recommendations and routines prescribed by CAS. In so doing, they privilege their own expertise and deliver an individualised service (Greatbatch et al). For example, nurses are expected to adopt a uniform approach when using the algorithms, however, in practice they: re-order, conflate, decline to ask and supplement CAS’s algorithmic questions, initiate their own lines of symptom-based questioning independently of CAS, etc. Thus in theory, one of the strengths of CAS is it supposedly delivers a standard service regardless of time, space, staff, or caller. However, given the interaction and joint production between the caller and the health professional such uniformity is impossible to achieve. The nurse and other staff adapt the software and deviate from it in order to strengthen trust, reassure callers and develop shared understandings. They see CAS as being too inflexible and rigid within this interaction.
These processes highlight a number of issues e.g. trust, professional autonomy, individualised services, standardisation, etc. At the heart of these is a contradiction. The government has called for greater individualised health services and yet in this new area of healthcare there is a policy of using CAS to standardise provision. At the core of this standardisation is the assumption that the protocols in CAS can account for every eventuality (Greatbatch et al). However, as numerous studies on the social application of technology have demonstrated, such an outcome is simply not possible. This is especially true in an area such as healthcare where a caller may be asking an elderly relative or child to interpret what is happening to their body and then interpreting this interpretation for a nurse, who then has to interpret (again) the meaning. If a shared understanding (and thus trust and possible empowerment) is to emerge, the nurse has to retain greater flexibility than the system currently allows. It is only by doing so that the system can function. This innovation enables nurses to provide an individualised, tailored service and to deal with the variety of calls that come in from people with a diverse set of backgrounds, knowledge and understandings of their own bodies and healthcare requirements. In short (from the nurses’ view) it allows them to better meet caller needs.
Such flexibility runs into conflict with the government desire to control and manage risk. This desire drives the urge to standardise the service. Given media scares about the ‘risks and inconsistencies’ within NHS Direct, this desire is perhaps understandable if debilitating. The service needs to balance its trust in nurses to make independent decisions and the need to minimise risk. Indeed, perhaps the two are not in conflict – trusting nurses may be a way of minimising risk.