Understanding complex trajectories in health and social care provision


Davina Allen,
Nursing, Health and Social Care Research Centre, School of Nursing and Midwifery Studies, University of Wales College of Medicine, 4th Floor, Eastgate House, Cardiff, CF24 0ABe-mail: allenda@cf.ac.uk


Ensuring collaboration between health and social care providers is a well-established policy concern in most developed countries. Thus far, however, this has proved to be a frustratingly elusive goal. Despite the growing body of empirical work devoted to this issue, social scientific theorising on the management of complex caring trajectories remains under-developed. This paper is an attempt to begin to address this gap in the literature. Drawing on Strauss et al.'s (1985) writings on illness trajectories and Elias's (1978) game model, we offer a framework – centred on the notion of a caring trajectory game – that can assist understanding of the linkages between individual trajectories of care and broader health and social care systems. It is only when we have developed a more theoretically sophisticated understanding of this relationship that we can begin to explain why trajectories of care take the course that they do. The framework arises from our analysis of eight ethnographic case studies of adults undergoing rehabilitation from a first acute stroke. In this paper we illustrate its utility by reference to one specific case: Edward.


Ensuring collaboration between health and social care providers is a well-established policy concern in most developed countries. For example, over the past 30 years in the United Kingdom (UK), successive governments have underlined the need for integrated working in the delivery of services. There is now a considerable body of empirical work analysing aspects of this issue: inter-professional and inter-agency collaboration (Bull and Roberts 2001, Hafsteindottir and Grypdonck 1997, Mistiaen et al. 1997), ‘system failure’ (Hart 2001, Victor et al. 2000), information transfer (Driscoll 2000, Luker et al. 2000), care management (Burton 2000b, Reed and Morgan 1999), interprofessional trust (Phillips et al. 1999, Phillips et al. 2000) and hospital discharge (Arts et al. 2000). With the notable exception of Strauss et al.'s study of the social organisation of medical work (Strauss et al. 1985), however, social scientific theorising on complex care management remains relatively underdeveloped (Parker et al. 2000). In this paper we suggest that Strauss et al.'s work can be fruitfully combined with a game model (Elias 1978) to produce a conceptual framework which can assist understanding of why integrated service provision is so difficult. We offer an approach – centred on the notion of a caring trajectory game (CTG) – that focuses on the relationship between trajectories of care and the micro-organisation of health and social services. This framework is based on the analysis of eight ethnographic case studies of adults who had suffered a first acute stroke and their associated network of care. For the purposes of this paper, however, we undertake an in-depth examination of a single case: Edward.

Study design and sample

The study was carried out between 1998 and 1999 in two Welsh health authorities and was designed to examine the fine-grained detail of service delivery to identify factors facilitating and/or impeding integrated care. A multi-case, cross-site comparative design was employed. People who had suffered a stroke were selected because they frequently have multiple and enduring health and social care needs (Burton 2000a). Four in depth ethnographic case studies were undertaken in each location (n = 8). Each case included the client, their family carer, surrounding web of care and the associated policy framework. The clients – three women and five men – were aged between 38 and 87 and all had multiple continuing health and social care needs. With the exception of Edward, who was discharged to a small residential facility to await relocation to a specialist service for young disabled people, all were discharged to their home. This required the negotiation of complex packages of care that met the needs of the client, their families and service providers and which balanced public, private and family care provision. In all cases equipment and adaptations to the home were required in order to support discharge. In one case re-housing was necessary.

Ethical approval for the study was obtained through the relevant research ethics committees in each site. Information sheets were prepared for the study participants. After initial access was agreed, the relevant service managers were contacted by letter and representatives from the research team attended multidisciplinary meetings to provide information and address questions about the study. Clients with potential to participate in the research were identified in consultation with acute sector staff. Once clients had been identified and signed consent had been obtained from them and their principal carer, access negotiations were taken forward with other case study participants on an individual basis. The main participants were identified through snowball sampling and audio tape-recorded interviews undertaken with them. Key events such as case conferences or home visits were observed and, where appropriate, tape-recorded. These data were supplemented by case note analysis. The local policy framework was established through the use of documentary analysis and interviews with service managers.

All data were transcribed in full and entered into a database. Computer-assisted qualitative data analysis software (Atlas/ti, Scolari 1997) was used to support data retrieval and analysis. Each team member considered the data independently and then met at intervals for ‘away days’ to discuss emergent findings. For each case, we mapped the networks through which care was delivered. An initial coding frame was developed which included both descriptive and analytic codes. Descriptive codes were used, for example, to identify specific categories of worker or service organisation. Analytic codes were developed in response to key themes in the data relating to roles and responsibilities, such as ‘flexible working’ or ‘boundary blurring’. This supported thematic analysis of the data for the purposes of producing a policy-oriented report for the funding body (Allen et al. 2000) and provided a starting point from which to develop more sociologically-oriented lines of analysis. For the purposes of this paper it allowed us to assemble the data to produce a chronological narrative account of each trajectory of care (see appendix, Allen et al. 2000). Comparative analysis of the case study narratives led to the generation of the cross-case themes which form the basis of the CTG (care trajectory game) framework discussed in this paper.

Theoretical framework: trajectories and games


The initial study design was informed by Strauss et al.'s (1985) illness trajectory concept, which is underpinned by theories of social interaction. Strauss et al. observe that technological specialisation and the evolution of complex bureaucratic health service structures have led to care becoming increasingly fragmented, especially in chronic illnesses. To study these processes the authors coin the term ‘illness trajectory’ to refer ‘not only to the physiological unfolding of a patient's disease but to the total organization of work done over that course, plus the impact on those involved with that work and its organization’ (1985: 8). They argue that:

[a] concept like trajectory is necessary for sociological understanding of illness management. [It is] above all a means for analytically ordering the immense variety of events that occur – at least with contemporary chronic illnesses – as patients, kin, and staffs seek to control and cope with those illnesses (1985: 9).

According to Strauss et al., healthcare work has two characteristic features. First, unexpected contingencies arise which stem not just from disease processes but from a whole range of organisational and technological sources. Second, it is people work, so that people react to and affect the work. These features (contingency and people) create the potential for trajectories of care to be complex and highly problematic:

[A] helpful image of what goes on with relatively problematic trajectories is this: efforts to keep the trajectory on a more or less controllable course look somewhat gyroscopic. Like that instrument, they do not necessarily spin upright but, meeting contingencies, they may swing off dead center – off course – for a while before getting righted again, but only perhaps to repeat going awry one or more times before the game is over. Sometimes, though, the trajectory game finishes with a total collapse of control, quite like the gyroscope falling to the ground (1985: 20).

Drawing on previous work on negotiated orders (Strauss et al. 1963, 1964) Strauss et al. (1985) argue that in problematic trajectories, ‘negotiation, persuasion, discussion, and teaching are the dominant modes of attaining maximum articulation’ (1985: 89). Such negotiation processes will take place within ‘a thick context of organizational possibilities, constraints, and contingencies’ and will therefore be shaped by a multitude of factors – interprofessional relationships, organisational processes and the different skills, influence and power of the negotiators – which will impact on the evolution of trajectories of care. We took these basic assumptions as the starting point for the research.

Trajectories as games

Whilst the Straussian framework proved invaluable in terms of designing the research and informing the processes of data generation, it also had several limitations for the purposes of our study. First, it is centred on the hospital context and does not include examination of trajectories which involve a progression from hospital to community care. Moreover, despite the originality of their work, there is a sense in which the Straussian framework remains implicitly predicated on a medical model. It is significant, for example, that throughout their work they refer to ‘illness trajectories’ rather than ‘trajectories of care’, which would be more apposite, given the inter-relationships between health and social care provision which are particularly relevant in cases of chronic illness.

Second, whilst they contrast routine and problematic trajectories, Strauss and colleagues stop short of explicitly conceptualising different types of trajectory. They do not indicate how potentially problematic trajectories might be identified or how disarticulated trajectories are recognised and who defines them as such. It is possible, for example, that a trajectory of care may be problematic from the patient's perspective, but not from the perspective of service providers. Because of the centrality of contingency in the Straussian framework, it would be illogical to attempt to identify prospectively a given trajectory as a prima facie ‘routine’ or ‘problematic’ case. Nevertheless, it is evident that some cases are more complex than others and certain features present higher risks of trajectory disarticulation.

Third, the cases we studied exhibited a dynamic which the Straussian framework did not readily explain. This is because whilst stressing the linkages of trajectories of care and the ‘thick context of organizational possibilities, constraints, and contingencies’ in which they are negotiated, Strauss et al. do not provide any basis for analysing the character of this relationship. Their case study examples vividly depict the patient's journey – the false starts, blind alleys and changes in direction – but the organisational context, interprofessional relationships and negotiation processes remain hidden from view1. As Strauss et al. make explicit, the aim of the illness trajectory concept was to shift attention away from professions and the division of labour and to focus on the client and the work involved, irrespective of who performs such tasks. One consequence of this stance, however, is that, despite the value of the illness trajectory concept for analytically ordering the events that occur over a sickness episode, it cannot explicate the linkages between such events and why it is that trajectories of care take the course that they do. Trajectories of care cannot be adequately conceptualised simply by reference to their constituent elements: if we are to increase our knowledge of their dynamic quality and the processes that give them their momentum, then we need to understand the relationships that link these elements together. In order to address these concerns we have looked to Elias's (1978) game theory framework. Elias's work helps us to begin to think about the features of complex trajectories of care and also provides insights into the ‘black box’ of Strauss et al.'s gyroscopic tilts.

Elias (1978) developed his game model in response to the dominance of scientific atomism in the social sciences. He argues that it is necessary to move beyond such reductionism towards frameworks that facilitate exploration of how the different components of social wholes are related to each other. Elias proposes a series of game models ranging from two-person to multi-tiered games. In this work the game metaphor is not intended to signify playfulness or any departure from the serious business of scholarship, but rather to provide a unit of analysis for understanding extended sequences of interaction in which multiple players exert mutual influence whilst recognising external constraints. For Elias, games are ‘interweaving processes with norms’ (1978: 80). Thus, like Strauss et al. (1985), he links cultural expectations and the meanings attached to actions by actors and introduces the prospect of individuals subscribing to different expectations dependent on their group membership and relationship to the game. However, Elias moves beyond Strauss and his colleagues’ interest in negotiation processes to focus on how social actions are intertwined, and to highlight the ways in which social networks become transformed over time.

Elias emphasises how people or groups which have functions for each other exercise constraint over one another. As he explains: ‘Their potential for withholding from each other what they require is usually uneven, which means that the constraining power of one side is greater than that of the other. Changes in the structure of societies, in the overall nexus of functional interdependencies, may induce one group to question another group's power of constraint’ (1978: 78). The important features of Elias's work for current purposes is that it points to the need to understand how these changes impact on all parties to the process. Elias explains that ‘[t]he more that the differential between A and B's strength decreases, the less power will either player have to force a particular tactic on the other. Both players will have correspondingly less chance to control the changing figuration of the game. In other words, to the extent that the inequality of the two players diminishes, there will result from the interweaving of the moves of two individual players a game process that neither of them has planned’ (1978: 82). One consequence of this, argues Elias, is that games can have unintended consequences which arise entirely as a result of the intermeshing of individual moves. The unplanned course of the game repeatedly influences the moves of each individual player and, as game complexity increases, the more certain it is that unintended consequences will be forthcoming.

Elias goes on to describe ever more complex game models and how increased game complexity affects individual players. He argues that in games involving large numbers of people there are pressures on the players to change their grouping and organisation. Large groups of players can disintegrate, splintering into a number of smaller groups. He also alerts us to the need to make clear in sociological analysis how players in complex games constantly struggle with the dilemma of, on the one hand, understanding the course of the game as the result of individual actions and, on the other, feeling that the game is a supra-personal entity. ‘For a long time it is especially difficult for players to comprehend that their inability to control the game derives from their mutual dependence and positioning as players, and from the tensions and conflicts inherent in this intertwining network’ (1978: 91). Additionally, Elias emphasises the importance of viewing game processes as part of a historical flow over time.

Taken together, the insights of Strauss et al. (1985) and Elias (1978) offer a framework for understanding both the relationships between players and the processes involved in complex trajectories of care. Building on this work, we have developed the concept of a care trajectory game (CTG) in order to encapsulate the useful insights of each perspective. If we are better to understand care trajectories, then we need to examine individual actions and decisions, the contexts in which these take place and the relationships between them. Analysing our cases through the dual prism of Strauss et al.'s illness trajectory and Elias's game framework reveals that CTGs have a number of dimensions which contribute to their level of complexity. In the second half of this paper, we explore the characteristics of CTGs utilising a framework developed inductively from the analysis of our case studies, applied to one of our cases, that of Edward. In using the metaphor of the game to analyse Edward's case it is not our intention to trivialise the experiences of those we studied, nor suggest that the concerns of this paper are in anyway frivolous. Our aim is to shed light on this key area of service provision. By suggesting an alternative metaphor with which to view this central health/social care concern our aim is to make a contribution to the production of new knowledge (Blumer 1969).

Case study: Edward


Edward Murphy was admitted to hospital with sudden loss of speech, then co-ordination. He had suffered a left frontal lobe infarction resulting in a right hemiplegia (paralysis) and dysphasia (difficulties in processing and formulating language). Married, with one son, Edward's wife, Rhondda, was pregnant. On admission Edward was completely dependent on nursing staff. He was immobile, incontinent and unable to comprehend any auditory or visual information. Two weeks later he began to develop some awareness and started to mobilise with the physiotherapists. However, although his dysphasia had improved, Edward had profound communication difficulties. Over the course of his CTG, both Edward and his wife suffered from depression and the multidisciplinary team often disagreed as they struggled to negotiate satisfactory arrangements for Edward and his family. During Edward's eight months in hospital a variety of care packages was explored, ranging from discharge to home through to placement with a foster family. He was eventually placed in a residential home, where he settled well, attending the day hospital for therapy once a week.

In many respects Edward's circumstances are unusual given his age and social circumstances, but we have selected his case for illustrative purposes because it best exemplifies the key dimensions of CTGs present in the cases we studied. In addition by focusing on the details of a single case, rather than developing a thematic analysis across all cases, we are able clearly to demonstrate the interaction of social process over time and how they are shaped by a multitude of structural, socio-political and organisational factors. This approach has been used to powerful effect elsewhere (Lawton 1998).

CTGs vary in the number of players they involve. In relatively simple CTGs the principal game players comprise the client and a single service provider. More complex CTGs, such as those studied for the purposes of this research project, involve many players, each with different biographies and knowledge. Players also have different levels of experience in playing the game and their power to shape its course varies enormously. Following Elias we predict that as the power differential between the players decreases game complexity increases, amplifying the probability of unexpected outcomes. This is highly significant given recent policy moves in the UK directed at equalising the asymmetries which have historically conditioned inter-professional and professional-user relationships in the public sector. The existence of divergent goals and cultural norms linked to different players also contributes to increased complexity and uncertainty in the course of any CTG.

The players involved in Edward's case were: Edward; his wife (Rhondda); Edward's son, parents, sister, and sister-in-law; consultant; speech and language therapist; senior occupational therapist; occupational therapist; social worker; relief social worker; two senior house officers; senior physiotherapist; physiotherapist; chiropodist; two residential home proprietors; day centre nursing staff; day centre doctor; dermatologist; senior staff nurse; deputy ward manager; ‘PP Homes’ manager; and a researcher. They were representatives from a complex field which encompassed an acute NHS Trust, local authority social services and the private care home sector. This meant that his CTG had to be played out across inter-agency boundaries with different cultures, finances, geographical boundaries, management structures, priorities, professional allegiances and planning cycles. This field is not wholly determined by the CTG players but clearly conditions their abilities to make certain moves.

Resource availability also shapes the complexity of CTGs. Lack of resources can stall a game or lead players into different moves in order to circumvent barriers to progress. The two sites in this study were marked by significant socio-economic differences. The cases in Area 1 lived in electoral wards with higher levels of socio-economic deprivation than those in Area 22. At the time of the study the allocation of health and social services budgets in Wales was based on a weighted per capita basis and therefore failed to fully recognise higher levels of disadvantage (An attempt to address the growing inequalities in health via a review of the funding mechanism is currently underway as the Welsh Assembly Government consult on the changes recommended by Professor Townsend in his report on the allocation of NHS resources in Wales [Townsend 2001]). Although Edward was fortunate enough to live in Area 2, his CTG was shaped by the limited support and facilities for young disabled people. This illuminates a further feature of CTGs which increase their complexity: the atypicality of a case. As Strauss et al. (1985) note, relatively straight-forward trajectories are those which follow a pattern which is well understood by the players concerned and conform to standard operating procedures. Health and social care provision is increasingly being driven by standardised care pathways, national guidelines and local protocols (Berg 1997a,b). CTG complexity will be shaped by the relative balance of routine and non-routine features which characterise a given case.

Edward's CTG was also fashioned in significant ways by disagreement over goals. On one side were Edward and the speech and language therapist who wanted a home discharge. On the other, his wife and the social worker claimed that with a young family to care for Rhondda would be unable to cope and the discharge would be unsafe. Initially a home discharge had been planned and arrangements were put in place to prepare for this, including a home assessment. However the ways in which individual players can shift the trajectory off course became clear during the home assessment when tensions between players became evident and the visit was terminated prematurely owing to the couple's distress.

These issues were then discussed at a case conference called by the speech and language therapist, which included all members of the multidisciplinary team, Rhondda and Edward's son and sister-in-law. Edward's mother and father were invited to the meeting by the speech and language therapist, but arrived late. The meeting began with the speech and language therapist explaining to the family that the purpose of the case conference was to ‘decide with the family and everybody where Edward is going to go’ following discharge from the unit. The speech and language therapist acknowledged Rhondda's concerns and argued that it was important to reach a decision about Edward's discharge destination so that time and effort was not wasted making arrangements if returning home was not a viable option. However, despite her even-handed presentation of the meeting's purpose, throughout the discussion the speech and language therapist (SALT) was at pains to express her preferred position or goal (a home discharge) and/or undermine Rhondda's (RM) justification for her decision not to have Edward home. This is illustrated in the following extract, in which the conversation focused on Edward's communication problems:

SALT to RM: It's the communication that bothers you most.

RM:     Mmm. Yeah, because it's no, you can't have a ( . . . ) he can't come up and tell me. You know? ‘I want this’ or you know? [ . . . ] He couldn't tell me ‘I'm tired’ or ‘I've got a pain here’. He can't describe things to me.

SALT:    What about then if you asked him questions based on what you think it might be?

RM:     Yeah, but I'm there forever aren't I ((Gives an example)) [ . . . ]

SALT:    So, you couldn't live with that level of pressure?

RM:     (.) See, I don't know. I don't know, like I can't say what he would be like at home?

(Case conference – tape recording)

Here then, the speech and language therapist locates responsibility for the CTG difficulties with Rhondda: it is her inability to cope with the pressure of managing Edward's communication difficulties that has placed the discharge arrangements in jeopardy. The discussion then focuses on community support, which the relief social worker (R.SW) explains is rather limited: a Day Centre is available some days a week but no home care. The scenario described by the social worker indicates a starker view of the reality of taking Edward home than that portrayed by the speech and language therapist and this prompts her to concede the ‘defeat’ of her preferred option. In so doing, however, it is Rhondda's inability to cope rather than the lack of support available to her that she constructs as the decisive factor:

R.SW: But erm (.) I think you've got to be aware that if we are going with the home situation, the responsibility is going to be levelled on your shoulders for most of the time. [ . . . ] So you've got to know what you’re taking on and, you know, what the situation is. [ . . . ] It is a permanent responsibility to know where we’re going with that. And I think it is necessary for us to work with that because, you know, for something to last, or be OK for a couple of months, isn't any good. Because we’re not talking about a couple of months.

SALT: Well, if that is the case, Rhondda, basically you’re saying that you can't have him home. I'm not trying to put words in your mouth but we have to be clear about what the options are.

(Case conference – tape recording)

The meeting is then refocused to consider what new arrangements need to be made in order to make progress in Edward's case. Nevertheless, despite the apparent agreement reached at the case conference meeting that Edward would be unable to return home, throughout the duration of his hospital admission, key members of the multidisciplinary team continued to press for this outcome. This might be seen as an illustration of the difficulty of any one player or set of players being able to control the overall shape of the game. The following discussion takes place two months after the case conference:

SALT:      I suggested to Raquel (relief social worker) that since his communication has improved a little, perhaps she would reconsider whether she (Rhondda) would have him at home. [ . . . ]

Senior Staff Nurse:  I've spoken to her and she said no way.

SALT:      And what's her reason this time?

Senior Staff Nurse: Still that she's got a lot on her plate. That she's now nearly, nearly due [ . . . ] and he's not going to have time to settle in.

(Multi-disciplinary team meeting – fieldnotes)

The tension between the SALT and Rhondda are very evident in this extract. Formally, CTGs are co-operative games, underpinned in the UK by a policy rhetoric which emphasises the centrality of individually-tailored packages of care, negotiated in partnership with users and their carers and, where appropriate, characterised by interprofessional and interagency collaboration and team working. However, our cases revealed inter and intra professional disagreements, disagreements between members of the multi-disciplinary team and families, and disagreements between different members of the same family including patients. Notice, in the above extract, how the SALT subtly undermines Rhondda's credibility through the implication that her ‘reasons’ not to have Edward home are situated and shifting, i.e. excuses, rather than being founded on substantive concerns. Edward's case was also shaped by the existence of a strained relationship between his wife and his own relatives: Rhondda's decision not to have Edward home was informed by the belief that she would receive no help from Edward's family:

RM: Well, like I say (.) well because as you say, [ . . . ] the care is not going to be there, it will be me (.) then, you know? Obviously his family is not here ((attending case conference)), which is a prime example. So, erm, you know?

(Case conference – tape recording)

As in the classic case of the prisoner's dilemma (Axelrod and Hamilton 1981)3, players have to make strategic decisions about whether to compete or co-operate with other players in order to expedite their game plan and this can result in regrouping and alliance formation. CTGs can be transformed into competitive games or at least contain competitive elements, thereby increasing their complexity. Such competition can be overt or covert and, as Elias (1978) has observed, it can result in different divisions and alignments within the overall game structure, which impact on its course and direction. All players are alert to the possibility that competitive sub-games may develop and attempt to guess how others are playing in order to anticipate their next move.

The disagreements over Edward's discharge arrangements resulted in the development of competitive elements in his CTG. For example, certain members of the multidisciplinary team expressed suspicion that Rhondda was attempting to use Edward's condition to assist her application for re-housing even though a decision appeared to have been taken that he was not going home:

Senior occupational therapist: His wife rang our main department this week. She wanted to have a copy of my home visit report, regarding housing (.) And she's seen me since but hasn't acknowledged this, so it's all a bit peculiar. Whether she's trying to go ahead with the housing situation regardless of where he's going. I'm not sure. [ . . . ]

Senior staff nurse: I think she is still using his case as the reason to get re-housed.

(Multidisciplinary team meeting – tape recording)

Irrespective of the veracity of these claims (which we were unable to establish), it is possible to see in this extract an example of the way in which an account of Edward's wife's motives are constructed which clearly comes into conflict with any notion of having Edward's best interests at heart. Rhondda's behaviour is formulated as both dishonest and governed by self-interest, as she ‘uses his case as the reason to get rehoused’. Thus a kind of moral blaming is taking place whereby her refusal to support the favoured outcome and care for her husband at home is linked to other proffered information, which is selectively interpreted, so as to discredit her character. This representation of Rhondda's motives has a clear function in terms of the competitive game; she is discredited as a player and her input into the game can justifiably be disregarded by those who remain committed to the goal of determining the ‘best’ outcome for Edward.

Elias predicts that as game complexity increases, players may fragment and regroup. This is evident in Edward's CTG which was shaped by a number of key co-operative alliances. Central to this was his relationship with the speech and language therapist (SALT). Owing to his communication difficulties the speech and language therapist took a lead role in his case and was a strong advocate for Edward throughout. In addition, when Rhondda indicated her reluctance to have Edward home, the speech and language therapist and other team members attempted to establish an alliance with Edward's parents in order to achieve their preferred objectives:

Senior occupational therapist: We wanted to get his parents involved. His wife was talking about sending him to London. His parents live in South Wales and I don't think they'd be happy with that. So, we want to give them the option. We don't want them to turn around at a later date and say ‘Nobody asked us what we wanted’. It’ll be good to get them involved anyhow because it'd be awful for him to feel abandoned.


In the event, the attempt to secure a co-operative relationship with Edward's parents was unsuccessful. The family arrived late for the crucial case conference, by which time all the key decisions had been made. Furthermore, they were unable to offer to care for Edward at home because they planned to be overseas for a prolonged period. It is revealing to note that Edward's parents’ actions – their failure to attend case conference meetings and their plans for overseas travel – do not lead to the same negative formulation of character that Rhondda receives. This suggests that the team's reactions to Rhondda are in part a reflection of the fact that her actions contravene cultural expectations about the relationship between husband and wife which require commitment for better for worse, in sickness and in health and highlight how, in any given CTG, a range of different social norms are available to players which can be brought into play as a warrant for action in a particular situation (Hughes and Griffiths 1997). CTG complexity will be enhanced if there are multiple conflicting cultural norms available in a given case.

In this case, realising that her decision had alienated her from certain members of the multidisciplinary team, Rhondda formed an alliance with the social worker, who was more sympathetic to her situation and in so doing brings into play another cultural norm: that parents should prioritise the needs of their children:

Relief social worker: Rhondda is aware that she's become the villain of the piece.

SALT: What, on the unit?

Relief social worker: Yes. She's not stupid. She's confided in me about having him back. It would be the easiest thing in the world for her to have him back, to go along with everything and pretend there's no problem. But she's being very wise.


It is helpful to remind ourselves of the contrast between this version of Rhondda's behaviour ‘she's being very wise’ and that produced by the hospital based team in the previous extract (‘she's using him to get rehoused’). The social worker's alliance with Rhondda continued to be a powerful influence on Edward's trajectory and served to bring the direction of Edward's care closer to what was desirable to Rhondda despite what was by now a clear challenge from the speech and language therapist (supported by some members of the hospital team) as to what might be the best outcome in this case.

The disagreements within the multidisciplinary team and between certain team members and Edward's wife resulted in a struggle to control the pace of the decision-making processes about Edward's discharge arrangements. This was evident during the case conference meeting. Ordinarily it is the social worker (R.SW) who would take the lead role in organising discharge, however, as Edward's self-appointed advocate, the speech and language therapist (SALT) adopted an assertive role by seizing the initiative when it became apparent that Rhondda did not wish to have Edward home:

SALT: So, erm, my next issue really is because of his communication (.) we need to get across to Edward that that is the situation before we think about offering him other situations, and see what there is. He's to be given the opportunity to say what he feels about that.

Rhondda: Yeah.

SALT: [ . . . ] I think that that is probably the next move. [ . . . ] But because of his understanding problem we still have to take time to get that across to him rather than whisk him away and put him somewhere where he's not really come to terms with the fact that he's not going home.

(Case conference – tape recording)

It is at this point that it emerged that the social worker and Rhondda have already arranged to visit alternative accommodation and, despite the speech and language therapist's (SALT's) efforts to control the pace of the game, the social worker (R.SW) offers a different plan, retaining control over the timing of events:

R.SW: Well, I've already spoken to Rhondda about Cheshire Homes which (.) I've made quite a lot of enquiries and to my mind it's the only option for him. [..] It's thin on the ground that cater for people in residential setting with disabilities . . . [ . . . ] Rhondda has already made arrangements to go this afternoon . . . 

SALT: Oh, right.

R.SW: . . . to have a look at it. So, we let you go ahead with that. Encourage you to go ahead and ask all the questions you want. And then come back to me and say what your impressions are, because, correct me if I'm wrong, say for instance that you hated it and you thought, ‘No, this is not the place for Edward.’ You know? We'd have to rethink things completely. So, at this stage, I don't think it would be, erm, (.) constructive to start the process of Edward understanding about the home situation.’

(Case conference – tape recording)

This example of developments away from the main field of play is one of many in our data and clearly illustrates that CTGs take place within a network of games. Furthermore, as Elias observes, as complexity increases the likelihood of players having access to all relevant information is compromised. We suggest that it is this type of incident that contributes to players’ perception that the game has taken on a course of its own which is beyond the control or comprehension of any one player.

It then emerged that the speech and language therapist (SALT) was about to go on holiday for two weeks. It is possible therefore that her position was informed by her desire to ensure that Edward was informed of his wife's decision before starting annual leave, providing another example of the way in which games interpenetrate one another. After some discussion, however, it was agreed that nothing more would be communicated to Edward until Rhondda had visited alternative accommodation and the speech and language therapist (SALT) had returned from her holiday. As Elias predicts, however, increased CTG complexity increases the possibility of unanticipated outcomes and, in Edward's case, this was evident in the actions of the speech and language therapist (SALT) who, despite the agreement reached at the case conference, informs Edward about his wife's decision not to have him home. Furthermore, she fails to communicate this information to Rhondda or other key members of the multidisciplinary team in the case, contributing further to the inability of any one team member to access all the knowledge and information available in order to judge the impact of any next move:

Senior staff nurse: Megan ((speech and language therapist, who is now on holiday)) told him [about his wife's decision not to have him home]. But unfortunately, nobody had phoned his wife to tell her that he knew they were no longer going to be together. And of course, then she came in on Tuesday not knowing what was wrong with him and he was all upset. [ . . . ] She went ballistic, that it was unprofessional, that he shouldn't have been told, if he was told, somebody should have called her and told her. And she's not happy with the whole situation at all.

(Multidisciplinary team meeting – tape recording)

The fragmented picture with which players in Edward's CTG were operating was also evident later in this case when it emerged that the social worker (R.SW) was also unaware of the speech and language therapist's (SALT's) actions. The following extract reports on a conversation between the researcher and the social worker about Edward's case:

R.SW: Since from the meeting we’ll have a fortnight while Megan (SALT) is away because nothing . . .  everything is in limbo until Megan comes back because nobody is even remotely thinking. Even if she'd gone there and thought it was wonderful, nobody was gonna tell him anything about it until Megan is back.

Senior occupational therapist: ((Peggy, who has overheard our conversation, interjects)): He has been told.

R.SW: He has?

Senior occupational therapist: Yeah, he has because Megan couldn't go away without telling him. He knew something was up. So she has told him that Rhondda cannot have him at home.

R.SW: Right.

Senior occupational therapist: There was no way that she could leave it because she came out of the meeting and he knew something was going on.

R.SW: Right. And so she's told him. Right, in that case . . . 

Senior occupational therapist: If you want to look in his medical notes, she's written a whole thing on it.

R.SW: Is he all right?

Senior occupational therapist: He's been OK.

Occupational therapist: He's indicated his own wishes as well in terms of where he wants to go. She's written it all down. [ . . . ]

R.SW:  Right. Well I'm glad you told me that. I was not aware of that. Erm, she wasn't either was she because (.) she was under the impression that he didn't know anything about this at the moment. So she needs to be told that because obviously it's going to present an awkwardness between them. He knows what she's thinking, but she doesn't know that he knows that she knows. [ . . . ]

Senior occupational therapist: Well, if you’re happy to tell her that's fine. [ . . . ]

Occupational therapist: I think that what it was, when they finished the meeting Rhondda walked straight past him. And he could see that she was in floods of tears. And obviously she didn't make any attempt to speak with him, so (.)

R.SW: Right. Right. So matters sort of came to a head. Well, anyway, that's life isn't it.

(Fieldnotes – tape recorded conversation)

Rhondda and the multidisciplinary team explored a number of options for Edward when it was decided that he would not be returning home. They finally settled on the ‘PP Home’, a specialist facility designed to encourage independent living. As Edward's discharge date approached, however, it emerged that the ‘PP Home’ would be unavailable as there had been a delay in vacating the place he had been assigned. Edward was aggrieved by this turn of events and alternative arrangements in a residential care home had to be organised temporarily. Within a week of his discharge, a ‘PP Home’ place then became available. Having settled into alternative accommodation and dissatisfied by the withdrawal of the original offer, Edward refused the place. His communication difficulties meant that he had been excluded from participating in many of the meetings to discuss his future but in this case the decision to remain was his own. The varying degrees of power or influence available to each player and the way in which participation in different stages of the game offers opportunities to enhance that influence over events are poignantly illustrated by Edward's difficulties. He was clearly unaware of an informal ‘rule of the game’ and his failure to adequately demonstrate enthusiasm for an independent living centre by taking a place immediately meant that he was unlikely to be given a second chance when another place became free:

SW: I think they've got the view that if he doesn't go there now, then they don't want to know him. [ . . . ] if he doesn't show the enthusiasm and commitment to it now, then that's your chance gone, sort of thing.


At the end of the study, Edward had settled well into his new accommodation from where he attended the day hospital for therapy once a week.


Combining the work of Elias (1978) with that of Strauss et al. (1985) in the notion of a CTG allows us to move away from competing mono-causal explanations of service failures, to focus on the dynamics of the system (including the roles played by patients and informal carers) and those system features that make managing complex care trajectories so difficult. Policy making in the UK is increasingly being influenced by a whole systems approach to health and social services provision (Department of Health 2000). For the most part, however, the perspective has been adopted in a relatively limited way, often for rhetorical rather than analytic purposes. In addition, the emphasis has tended to be on macro-level features of health and social services systems, rather than on the micro-organisational dynamics of the system in daily practice. Our work at this level has offered an opportunity to explore an often neglected dimension of the realities of planning and providing care in one context and to examine the interdependence of system components.

Viewed through the CTG framework, Edward's case has helped us to understand why trajectories of care take the shape that they do. We have seen how the moves of the various players were shaped by the other games in which they were engaged and how, in turn, these actions interacted to shape Edward's care trajectory in ways which could not have been predicted and which had unintended consequences. Central to the outworking of Edward's CTG were: the dearth of specialist facilities (resources) or community support for young adults who have suffered a stroke (non-routine case) and the disagreements that existed between the multiple players over CTG goals which led to alliance formation and the development of a competitive game (strategy). Our analysis demonstrates how these tensions were played out over a protracted period of negotiations (game networks) making it difficult for all players to keep track of the game's course (intelligence) and how the alliances between the hospital team led to a devaluing of the wishes of his wife once they reached the view that she was not putting his interests first (cultural norms and values). With this change in Rhondda's status a new team, constituted to exclude Rhondda, began to play on Edward's side. However the alliance between the social worker and Rhondda meant that her loss of power due to her compromised status was compensated for by the efforts, status and resources of another player.

Examined through a CTG framework, Edward's case raises broader issues in relation to current policy trends designed to promote integrated service provision. While the precise characteristics of a ‘joined-up’ or ‘integrated service’ are not made clear in the policy literature, it appears to comprise three key components: interprofessional and interagency collaboration; seamless services; and individually-tailored care. Implicit in much of the literature is the assumption that these elements are mutually supportive4. Our analysis suggests that such assumptions can be called into question.

The importance of the relationships between players in CTGs cannot be underestimated and there is a strong body of evidence which points to how differences in culture, goals, training and ethos have hamstrung inter-professional working. This has been taken seriously by UK policy-makers who have introduced a number of initiatives – such as inter-professional education, the creation of hybrid workers, and break-down of jurisdictional and disciplinary boundaries – designed to overcome professional ‘tribalism’ and promote integrated service provision. The assumption underpinning these trends is that for care to be of a high quality, multidisciplinary teams should be consensual (Dingwall 1980). Yet our evidence indicates that this may not necessarily be the case.

While conflicts and disagreements over goals were a feature of all the cases studied, Edward was the only example in which the multidisciplinary team was divided in such a fundamental way. His case demonstrates, as Elias predicts, that as the power differential between players is reduced it becomes less possible for one player to predict or control the outcome of action. Certainly the move to enforce shared responsibility for planning care can be seen as effecting a redistribution of power from health service players to social services staff and the additional role which is now to be played by patients and their carers makes planning even less predictable. Because many of the features of his case were unusual (his age and young family), there is a sense in which the complexity of Edward's case was accommodated by service providers, in as much as they were prepared to resist the normal pressures to effect a timely transfer of care. But this was not true of some of the other cases we studied, where the importance of acute bed utilisation as a measure of the effectiveness of interagency working acted as a powerful incentive to simplify case complexity in order to effect a timely transfer of care. Players were observed to be frequently orienting to this framework. The most common example in our study was for health and social care staff to form an alliance to press for a nursing home placement rather than a home discharge, as these are easier to organise. In the absence of resistance from the patient client and their family, a CTG could be transformed from a complex to a simple case to which standard operating procedures could be more easily applied.

Recent efforts designed to promote integrated working in the UK – such as charging social services departments for transfer of care failures (NHS 2002) and unified assessments – are likely further to increase these pressures and it is here where the interdependence of system components becomes potentially problematic. By acting as an incentive for health and social services staff to work collaboratively, these proposals have the potential to transform the CTGs’ network of social relationships in ways which disadvantage service users. By strengthening the alliances between health and social care players the relative power of the patient and their carers is weakened and this is likely to have an adverse effect on patient choice and hence, service quality. Furthermore, in the absence of additional resources, there is also a danger that pressures to simplify the game and achieve closure on the part of some players will only serve to increase CTG complexity at a later stage for the patient and their family, if future options have not been fully explored and the discharge arrangements break down.

Edward's discharge process might have been simpler and less protracted had the divisions between the players not existed, but would this have produced the best outcome for him and his family? Had the SALT's view predominated then Edward might have been discharged home without due consideration of the sustainability of these arrangements and the consequences for his wife. As Anspach (1993) has observed, the division of labour functions as an ecology of knowledge. In her study of a US neonatal intensive care unit, she cautions against a culture which allows only certain types of knowledge to be used to inform decision making and which, in the clinical setting she was studying, could impel physicians to continue supporting an infant's life long after this was appropriate. In developing her argument Anspach is referring to the dominance of some forms of knowledge over others, but her warning is equally applicable to the chronic illness context, if different voices are successfully silenced by the homogenising impetus of policy trends. While decision-making processes may be messier and more protracted, a plurality of perspectives ensures that a range of options is explored in order to reach outcomes which balance the needs of all those involved. There is, however, clearly a need for this plurality to be made more explicit, for players to be made aware of the unintended consequences of their moves, and for services to be organised so that time is allowed in which a range of options can be considered, rather than CTG complexity being compounded by competition to control the pace of the game. As we have argued, strokes cause ‘biographical disruption’ (Bury 1982); clients and their carers needed time to adjust to their altered circumstances and make the necessary modifications to their daily lives. There is clearly a need for the development of transitional services where, unlike the acute sector, the emphasis is not on trying to bring about CTG closure, but on the provision of an environment which facilitates accommodation and adjustment.


In this paper we have attempted to take a small step towards the development of a theoretical understanding of complex caring trajectories. Drawing on the work of Strauss et al. (1985) and Elias (1978), we have argued that it is insufficient for policy makers and service planners to focus attention on selected elements of the health/social care system without first identifying all system components and acknowledging and understanding the implications of their interdependence. The CTG framework outlined in this paper is offered as a structure for further research aimed at developing a typology of care trajectories which can be deployed for academic and practical purposes.

In this paper, analysing and explaining ‘the mechanics of this transformation of intentional human action into unintended patterns of social life’ (van Krieken 1998) in the case of Edward, has enabled us to identify some of the potential unintended consequences of current policy proposals in the UK. We have argued that rather than trying to eradicate complexity, the aim should be to develop service arrangements which are better able to accommodate it.


This study was funded in part by the Wales Office of Research and Development for Health and Social Care, National Assembly for Wales. The views expressed herein are those of the authors and do not represent the views of the Wales Office of Research and Development for Health and Social Care or the National Assembly for Wales. We would also like to acknowledge the valuable contribution of Lee Monaghan and De Murphy who were research assistants to this project. Earlier versions of this paper were presented at the British Sociological Association Medical Sociology Welsh Regional Group, July 2001, the British Sociological Association Medical Sociology Group Annual Conference, September 2001, and the Royal College of Nursing Research Society Annual Conference, April 2002. Thanks are also due to Ben Hannigan, Evelyn Parsons, David Hughes and two anonymous SHI referees who commented on an earlier draft of the paper.


  • 1

    As Mellinger (1994) has observed, although the negotiated order perspective has contributed a great deal to our understanding of organisations, relatively little attention has been focused on the character of real world negotiations.

  • 2

    The Welsh Office has calculated an index of socio-economic conditions for every electoral ward in Wales. These are based on eight factors: unemployment, the economically active population, low socio-economic groups in the population, population loss in the 20–59 age group, permanently sick in the population, over-crowding in housing, basic housing amenities, standard mortality rate. The latest version of these scores which is available was devised in 1993 and is based on 1991 District Electoral Ward. The cases in study Area 1 had higher socio-economic deprivation scores on this index than those recorded for our cases in Area 2.

  • 3

    ‘The “Prisoner's Dilemma” is used as the standard metaphor to conceptualise the conflict between mutual support and selfish exploitation among interacting non-relatives in biological communities. [ . . . ] The “Prisoner's Dilemma” is a simple two-person game where each player (for instance two prisoners accused of the same crime) can choose either to cooperate (C) or to defect (D = not cooperate)’ (Brembs 1996: 15).

  • 4

    See, for example: ‘The NHS and social services do not always work effectively together as partners in care, so denying patients access to seamless services that are tailored to their particular needs’ (DoH: 2000: 70).