Design and methods

The agreed aim for the CPCEC project was to develop a more detailed understanding of what happens in dementia-care settings; what goes on when a care worker and a resident with dementia interact, and what can be observed and understood of the experiences that people with dementia have of their care and the settings in which it is provided. In doing this, the aim was to build a picture of dementia care that would assist critical thinking about care practice and, in particular, to identify those aspects of care which could be improved. The project was designed to focus upon patterns of interaction in care and associated daily practices. The rationale was that the features of caring relations that were repeatedly witnessed were likely to be defining qualities or characteristics of dementia care.

The study sought to make a contribution to the small but growing body of evidence on communication in dementia-care settings (e.g. Goldsmith Reference Goldsmith1996; Hamilton Reference Hamilton1994; Killick and Allan Reference Killick and Allan2001). Drawing from authors that have argued it is both valid and helpful to consider various types of expression and behaviour by people with dementia as forms of communication (Hallberg, Norberg and Eriksson Reference Hallberg, Norberg and Eriksson1990; Hallberg et al. Reference Hallberg, Edberg, Nordmark, Johnsson and Norberg1993; Kitwood Reference Kitwood1997), we sought to test the specific hypothesis that, ‘irrespective of the severity or nature of impairment, a person with dementia will seek out and establish a means of self-expression and thereby make every effort to maintain a relationship with the world they inhabit’.Footnote ²

The research process and sampling

The project began with the development and ratification of a detailed ethics protocol, a full account of which is published elsewhere (Vass et al. Reference Vass, Minardi, Ward, Aggarwal, Cybyk and Garfield2003). The project was divided into three phases, each of which examined patterns of communication in different settings; first in residential care, then in different types of day care, and finally in home care. At the time of the research, Jewish Care owned and ran 11 care homes registered to provide residential, dementia and/or nursing care. The homes varied in size from 40 to 120 residents, and approximately 75 per cent of the residents were funded through the statutory health-care and social services agencies.

The sampling frame was a list of the 594 staff employed in both care and non-care capacities across the 11 homes (with ethnicity and gender recorded). From this list, 123 staff were randomly selected, including 28 men (23%) and 95 women (77%). The sample included care assistants, night care-workers, group leaders, qualified nurses, home managers, assistant managers, and various non-care workers such as domestic staff, administrators, maintenance workers, laundry assistants and dining-room co-ordinators. They were invited to a general meeting to discuss the project, and offered the opportunity of a one-to-one meeting with a researcher at which they could raise questions or concerns. The research team also gave presentations about the research at each of the homes included in the study. Following the consent gathering process, 32 care staff and six non-care staff agreed to participate in the project. All were interviewed using a semi-structured interview schedule, with an abridged version for non-care staff.

Turning to the residents, a sample frame was identified of 17 men and 78 women with a formal diagnosis of Alzheimer's type dementia among the residents of Jewish Care's facilities. From this pool, 28 (five men and 23 women) were randomly selected for the study. Once selected, the research team initially approached the next-of-kin (or guardian) of each resident to seek consent for the resident's participation. In so doing, the next-of-kin were asked to participate by giving an interview. Once consent was received from the next-of-kin, a member of the research team approached the resident, providing information on the project and asking for their consent to participate. It had previously been agreed that if either the next-of-kin or resident declined to give consent, the resident in question would not be included in the project. After consents had been obtained, 17 residents (two men and 15 women) were recruited. The five study homes (and eight care units) were those in which the sampled residents lived. Two were dual-registered, two registered as care homes and one as a care home with nursing. The five ranged from 40-beds settings to self-contained units with only nine residents. Some of the care homes we visited were purpose-built while others had been converted from other uses.

Data collection

The residents were filmed in communal day rooms where also researchers observed and recorded interactions and daily events. Each resident was filmed over four-and-a-half days between 07.00 and 19.00 hours, with at least one researcher always present. The cameras were fixed to the walls of the day rooms of each unit. Throughout the design period and during the fieldwork, the team had to balance ethical considerations with the ambition to generate an accurate picture of residential dementia care. In addition to the continuous monitoring of the wellbeing of the participants, it was agreed that no filming would be undertaken in private spaces, such as residents' rooms, or of the more intimate forms of care, such as dressing, bathing and assistance with using the toilet. The data presented here are therefore based exclusively upon observations and filming in communal areas of the homes, and the measures and timings that follow omit those periods when residents received care elsewhere in the homes.

To provide an overall picture of life in the residential care facilities for both staff and residents, the video data were supplemented with diary observations of the routines and day-to-day activities on each of the residential units. Eleven volumes of diary observations were compiled and they have nearly 1,400 entries. Access was also provided to the documentation held by the care homes on each of the sampled residents. Where possible, efforts were made to interview or to conduct shorter exchanges with the participating residents. In summary, the data gathered for the first phase of the CPCEC project included:

1. 1. 32 semi-structured, in-depth interviews with randomly-selected care staff, and six interviews with non-care staff.

2. 2. 18 semi-structured interviews with next-of-kin.

3. 3. Interviews and/or a series of shorter exchanges with 17 people with dementia.

4. 4. Over 300 videotapes (equivalent to over 1,000 hours of recording).

5. 5. 11 volumes (more than 2,000 pages) of observation diaries.

6. 6. Records of the structure and content of 17 files and care plans.

Analysis

The interview data were coded using NUDIST software and the themes that emerged were cross-referenced with other sources of data including the textual analysis of care-home documents. A comparative analysis of the interviews conducted with staff, next-of-kin and residents was undertaken. Findings from the observation and video data were used to confirm the interview evidence and to pinpoint areas of dissonance. Approximately 230 tapes (excluding those from pilot studies) were amassed during the fieldwork, from which archive at least three were systematically selected for each residential unit. This ensured that different times of day and each participating resident were represented on at least one tape. In all, 30 tapes with over 100-hours of footage were scanned using event sampling to note and time each discrete interactional episode.

To analyse the video data, the team (including representatives of Jewish Care) used a modified version of the ‘Quality of Interactions Schedule’ [QUIS] (Dean, Proudfoot and Lindesay Reference Dean, Proudfoot and Lindesay1993). Diary entries were also coded according to QUIS headings, but one problem was encountered: one of the most prevalent ‘activity states’ that was observed was social isolation, but it was not a given category. An additional category, ‘neutral-neutral’, was introduced for periods where no observable communication took place.Footnote ³ High levels of concordance between coders were achieved but some difficulties remained with the coding tool. The differentiation of ‘neutral-neutral’ periods from discrete episodes of interaction was problematic. Another major concern was that there was a great deal of interaction (especially at the non-verbal level) that we found difficult to record and interpret. Throughout the ‘neutral-neutral’ periods, people were moving, gesturing, nodding their heads, observing and acknowledging one another, listening and reading. In sum, we realised our own shortcomings in recording communication and social interaction and the limitations of pre-defined categories. It was clear that an exclusive focus on the (more easily interpreted) contact between staff and residents would fail to capture the essence of what had been observed and recorded.

Two stop-watches were used to time precisely the durations of the residents' direct communications and of the spoken inputs (or talk) that formed elements of these encounters. A tally was also kept of the frequency of the interaction episodes. A second analysis of the video data was undertaken by one member of the research team. This scanned the entire stock of video tapes using event sampling and each episode of interaction was classified into one of three categories:

Once classified, episodes illustrative of each heading were transcribed and examined in detail for both verbal and non-verbal content. The observational diaries and interviews were also used to support more detailed scrutiny of the filmed self-expressions and interactions of the residents. Given the individual and diverse forms of the recorded expressions, analysis and interpretation of video data was at various levels for different residents, but the findings were systematically organised to identify the patterns in the self-expressions, the interactions and in the responses to the (social) environment.

The findings

This section reports the findings from the video data and observational diaries, which are illustrated with extracts from the interviews. Most residents spent the best part of the day in the communal areas of each unit, where the filming took place. As the carer interviews revealed, however, certain episodes of personal care (not recorded) could present crucial opportunities for sustained interaction between staff and residents. Many carers specifically cited bathing as being amongst the most important occasions for individual forms of communication:

We identified the core patterns of interaction in residential care. The analysis of the video data revealed that people with dementia on average spent 10 per cent of the day in direct communication with others. The timings revealed much individual variation in the overall levels of interaction and the relative frequency of different types. Regular visits from friends and relatives, as well as close friendships with other residents, raised an individual's level of interaction. However, those who required assistance to mobilise were observed to move around the care home less and to exercise less choice over where and with whom they sat.

One project participant was a member of a small network of friends who were observed in the group management of impairment. Mobile members of the group assisted those who were in wheelchairs; those with better memories reminded and reassured others who had forgotten family visits or the time of day; and two group members clarified and repeated the comments of a third who had suffered a stroke, to facilitate her participation in conversations. Strategic positioning also played a part in influencing levels of daily interaction. One project participant sat next to a doorway that linked two communal rooms, which enabled her to engage care workers repeatedly in brief exchanges as they passed by. We also observed that some residents congregated around the foyer and main entrance door of their home, which meant they encountered people arriving and leaving and enjoyed higher levels of communication than those who sat in the day rooms.

Direct contact with care workers occupied on average 2.5 per cent of the day with little variation. The analysis of the video data revealed a low frequency of interactions between care staff and residents. Over the 12-hour observation period, the average number of discrete episodes of interaction was 17.7 per resident, of which 12.9 were initiated by care staff and 3.1 lasted no more than five seconds. On average, 2.3 of these encounters were wholly silent, with no verbal input from either worker or resident. Thus, most contacts were initiated by care staff but they tended to be brief and some were entirely silent. Of all carer-initiated interactions, 77 per cent were coded as task-based, 15 per cent were social or relationship oriented, and eight per cent combined a task-based encounter with social or relationship-oriented content. By timing precisely the duration of the spoken content in these encounters, it was found that only approximately one-third of the contact time between care staff and residents involved verbal exchanges. Thus, silence is the dominant mode of caring encounters.

The residents appeared to exercise little choice in the timing and provision of care and the observed care episodes unfolded at a pace set by the staff. Overall, the video evidence pointed to a standard form of engagement between staff and residents, despite marked variations in the residents' communicative and cognitive capacities. A core pattern of encounters, organised according to a daily sequence of tasks, lies at the heart of dementia care. Elsewhere we have described this pattern as ‘the mantra of dementia care: out-of-bed – wash – dress – feed – toilet – back-to-bed’ (Ward et al. Reference Ward, Vass, Aggarwal, Garfield and Cybyk2005). The video films revealed a common format and a predictable sequence to most of the encounters. Both residents and carers had a tightly-defined role to play in these encounters from which they rarely departed. The analysis has concentrated on the most representative and common attributes of communication between staff and residents, but the videos also captured less frequent encounters, during which moments of tenderness and warmth were communicated verbally and non-verbally, and which often led to clear signs of wellbeing for the resident. There were also occasions when workers successfully engaged residents in sustained interactions and managed to overcome a host of communication challenges. In our opinion, there is much potential benefit to be derived from close scrutiny of these exceptional encounters.

‘Care-speak’ or how to get things done in dementia care

The residents of the studied care homes had variable levels of impairment and related needs. According to our interviews and conversations with them, many craved the opportunity to be socially engaged. Despite this, a focus upon the efficient completion of a daily sequence of care tasks was managed and maintained by the staff and wider organisation. Accordingly a great deal of time, effort and skill was invested in maintaining a certain order of relations in care. How did staff maintain order and fix relations in these brief encounters? Despite the varied nature of the tasks that the workers were required to undertake, it was evident that interactions with residents had four main components:

• • An opening.

• • A signal of intent.

• • Accomplishment of the task, and

• • A closing remark or gesture.

Our observations found that it was not uncommon for some or all of the sequence to be performed non-verbally using gesture, guiding touching and physical handling. When a resident complied with a task and showed knowledge of the routine, often the carer's speech was minimal. The single most frequent predictor of greater levels of spoken input was when a resident failed to comply with a task. When undertaking tasks, the care workers were observed to employ a distinct style and pattern of speech, which we labelled ‘care-speak’. Over-riding differences in gender, ethnicity and age, care-speak was used by workers in all the studied homes. It is characterised by a series of directives, framed by narration of the task at hand and punctuated by words of encouragement. During these episodes, verbal input by the residents was rarely required and not often elicited. An example of the directive speech of one care worker follows:

When speech featured in task-based encounters, this form predominated. Such guiding narration of the task at hand might be considered good practice, not least when working with individuals with cognitive impairment. When, however, other forms of engagement were absent, there are clear implications for the social experience of the people with dementia. The speech form constrains input by the residents when tasks are undertaken, the result being that during the very limited periods of engagement between staff and residents, there was little opportunity for a resident to make an active contribution. Care-speak also ensured that there was little space during the contact for talk about non-task topics. This standard form of communication limited the opportunity for personal inputs by care workers. In effect, from the standpoint of the person with dementia, one encounter with a care worker was much the same as the next. Despite the variable size and level-of-care provided by the eight care units in the study, we found that the daily levels of direct contact between staff and residents differed minimally from one setting to the next.Footnote ⁴

In order to ensure compliance during the provision of care, staff drew upon what seemed a well-rehearsed tactical repertoire. The video analysis revealed four types of response, although from an observational standpoint one cannot be certain what factors influenced the choice of response:

• • ‘Desistance’ (i.e. desisting from or postponing a task). This was the least frequently observed but arguably the option that most readily recognises the will and agency of a resident.

• • Verbal persuasion. This was common in all the homes.

• • Coalition, when a second and sometimes even a third carer would contribute to the persuasion.

• • Enforcement using authoritative commands or by physical presence, for example when keeping residents in position during the provision of care.

The challenge of dementia care

The carers spoke of the emotional dimension to their work and the personal costs it entailed. Many highlighted the tension between the task-based, instrumental forms of care they felt under pressure to provide and their wish to spend time building relationships with the residents. There were clear indications that carers felt their style and patterns of working were not fully under their own control but shaped by the wider organisation and care system. They said that workload pressures precluded sustained interactions with individual residents. The challenges associated with communication, the difficulties arising from miscommunication, as well as the many levels at which communication can and does take place, were outlined by carers in a manner that highlighted the specialised nature of their work and skills. If it is accepted that dementia care is the ‘main business’ of care homes, then there is a strong case for promoting and supporting these skills and prioritising their development through training and supervision.

The interviewees reported having to manage their emotions by undertaking a form of emotional labour; they also spoke of investing their emotions in the residents for whom they cared, which of course entailed negative consequences when a resident died or her or his health deteriorated.Footnote ⁵ The care workers also described communication at the level of the emotions in dementia-care settings. When usual channels of communication were no longer an option, care workers signalled the importance of responding emotionally to the residents. As one care worker put it:

Several interviewees had the perception that people with dementia are attuned to the emotions of others and respond to the moods and feelings displayed by those around them. One care worker explained very well how this realisation affected their care practice:

Implicit in these responses is a sense that in dementia care there is a need to present a stable and positive affect, echoing the ever-smiling persona adopted by the flight attendants that was described by Hochschild's (Reference Hochschild1983) study of emotional labour. The care workers’ accounts appeared, however, to depart from the usual understanding of emotion as a (private) psychological state. Instead, in many of the interviews we undertook, emphasis was placed upon the sociality of emotions. In other words, emotions were presented as shared and having currency in dementia-care settings, as forming the substance of relationships and being a vital means of maintaining a social presence by or for the person with dementia.

The order of care

Care work was structured to maintain order at several levels. Some carers told us that they intervened when residents argued and, in this way, maintained emotional and interactional order in the communal area. One care worker lucidly explained the rationale:

Notions of the good and bad resident were also signalled during the interviews and were differentiated by the comparative levels of attention they requested. Labelling residents as ‘demanding’ or ‘attention-seeking’ revealed the carer's perspective on the legitimacy of a resident's needs and demands. We observed that these notions carried through into everyday practice, there being instances of care workers refusing requests for assistance or attention on the grounds that they were out-of-turn or excessive.

Bodily order and the erasure of need

The outward presentation of the residents' bodies was also a key site for the maintenance of order. Echoing Lee-Treweek's (Reference Lee-Treweek1997) notion of the ‘lounge-standard resident’ (see also Reed-Danahay Reference Reed-Danahay2001), the interviews with the care workers underlined the extent to which appearance and presentation served as a barometer of the provision and quality of care. One carer described her response to the changed appearance of a resident she had cared for who had subsequently moved to another unit:

This theme of the well-turned-out resident as an indicator of not only the good quality of the care but also of wellbeing was echoed during the interviews with the residents' relatives. One signalled the importance of smartness and presentation in judgements of care. Speaking of her father, one woman said ‘he was more unkempt at [a previous home] than he is here. He's nicely dressed here and he looks comfortable, so I wouldn't want to change it. I don't think there's any need to change [move him to another home]’. Another relative revealed the significance she attached to presentation even though the resident had indicated a desire to be elsewhere:

It is important not to downplay the emphasis placed by relatives upon how maintaining a clean and smart appearance supported an expected form of (embodied) biographical continuity. Our view, however, is that the absence of effective communication led to the residents' bodies acquiring a symbolic value in care homes. These clean, well-groomed and smartly-dressed ‘bodies’ not only act as indicators of a certain quality of care, they are the objective of the care. Many of the other individual needs of residents seem to have been erased and substituted by the value attached to physical and social order. We found this unquestioned drive to maintain order at every level of dementia care.

Environmental order and panoptic control

The descriptions by the care workers of the maintenance of order corresponded with our observations of the management of care settings, their layout and movements within them. In all eight of the studied units, the seats were arranged side-by-side in rows and were mainly against the walls facing the centre of the room. This arrangement clearly inhibited interactions among the residents. Where the seats faced one another, the wide space between them allowed the free passage of wheelchairs but impeded interaction. This layout also enabled the carers to survey the residents and take stock of their welfare without direct engagement.

Regulated in their use and occupation of space and their movement around the home, the residents were restricted by both physical barriers and the care workers' interventions. Those residents found walking the corridors were routinely led to the seating areas. One of the most frequently heard carer's instruction was that a resident should ‘sit down’, even though most residents spent a large portion of the day seated. Care workers were observed deflecting offers of help from residents, who were barred from entering kitchens, laundry rooms and offices. On some units during the day, the doors to bedrooms and those leading in and out of day rooms were locked or in some way secured.

The data suggest that the maintenance of order is a driving force behind dementia-care practice. The day-to-day provision of care is organised according to an array of signifiers that function to reassure care workers, relatives and the wider organisation of the quality of care. The staff and relatives were far less questioning of the levels of contact and engagement. We found little evidence to suggest that these priorities were consistent with the expressed wishes and preferences of residents themselves. Indeed, when considered from the standpoint of communication, the orderliness of care homes presented a host of artificial barriers that adversely affect communication and meaningful social interaction.

Michael's self-expression

Michael had been a resident for over three years.Footnote ⁶ We were informed by care workers that he was ‘unable to communicate’. This assumption was reflected in the staff's comments and recorded in his file and care-plan at each stage on the ‘pathway’ of assessment, placement in the home and care-planning. Michael appeared unable to communicate through speech, his facial expression altered very little, he was unable to turn his head, and his body was stiff. He required a wheelchair to move about the home. There was observable movement in his arms and legs but he did not seem able either to call out or to wave his arms. In common with many of the residents with severe dementia that we observed, there appeared to be no agreed signal by which Michael could attract the attention of staff, so contact was mainly initiated by the care workers and he had little apparent control of when this occurred.

Michael would sometimes slide down in his chair. Each time this happened, carers helped him sit up straight again, and habitually paused to say a few words or hold his hand. This sequence occurred repeatedly and was the only way Michael seemed able to influence the contact he had with staff. We began to suspect that initiating this chain of events was an intentional act of communication on his part. As Michael sat watching the day's events, he would unfold his arms and his legs would jerk and twitch. When certain pieces of music were played, he tapped his hand or foot in perfect rhythm to the music, demonstrating control over this movement. Each time he first caught sight of his wife when she visited, his movements quickened. When she left, his movements slowed and sometimes completely ceased. During the week's filming, Michael's movements had a distinct pattern. They were associated with these key events and we identified rudimentary connections between his movements and the events. We used these patterns as a way of beginning to interpret certain messages. For instance, more intense movements seemed to indicate excitement, while the slowing of movement seemed to signal sadness or low-mood, especially following the departure of his wife.

As with many of the people who took part in the project, very few efforts were made by the care staff to interpret these patterns of self-expression. Michael's movements were viewed as involuntary and attributed to his dementia; the possibility they might hold meaning and be a key to understanding his thoughts and feelings was generally discounted. There was no evidence in his file to suggest that any structured form of observation, interpretation or recording of these patterns had been attempted. As a result, Michael's efforts to engage with others were frequently overlooked. For Michael and others in similar situations, periods of low mood or worse failed to attract the reassurance of a comforting response from those around him.

The staff's closer engagements with residents

Our observations and the video film did reveal instances of carers who had established a close relationship with a resident (often as their key-worker) and who responded to the patterns of expression that we interpreted as the means of maintaining communication with the individual concerned. These relationships were isolated, however, and it appeared that the knowledge associated with them was seldom shared. This impression was supported by our interviews with care workers, many of whom seemed to struggle to put this aspect of dementia-care practice into words. One care worker's account shifted from an individual to the generality of the residents:

Such comments suggest the need for a more widely accepted vocabulary by which to describe the detail of communicative encounters in dementia care and for a support system that records and shares these skills. In short, care workers need time to reflect upon and discuss communication with the residents. The research found that there was very little formal support for developing communication skills. In fact, many of the practice documents used by care homes routinely label people with dementia as unable to communicate or express themselves. This presumption was by no means confined to care workers, as a phrase from a social worker's assessment showed: ‘[The client is a] charming man; he discourses in a long and involved way and is talking total rubbish’. It was clear that from their arrival in the homes, many of the residents were deemed incapable of making a contribution to crucial decisions affecting their lives. Another written comment, from a resident's placement review, characterises the attitude: ‘Mrs X is sadly unable to express her views due to her mental state’. Nor were they considered capable of participating in the inspection process. As one of the inspection reports noted:

The slight attention given to the residents' self-expression effectively excludes people with dementia from any form of participation in or influence over the support they receive. This amounts to social exclusion in an absolute form. Such routine exclusion represents what might be described as ‘cognitive disablism’, and is an example of the discrimination faced by people with dementia in their everyday lives. There is a pressing need to move beyond the expediency of labelling the person with dementia as unable to communicate and to seek out the means by which to recognise, support and facilitate their input.