In situating the task of developing a delirium prevention system of care, we describe how delirium and the work of delirium prevention were currently understood and accomplished by staff. We draw on NPT mechanisms to organise the findings and illustrate interpretive points from our fieldnotes and interview data. We then review the evolving model of delirium prevention developed iteratively through the empirical research and participatory process with development teams. Finally, we present the integrated delirium prevention system of care (POD system of care), including the rationale, or theory of change, underpinning it.
Knowledge and awareness of delirium
Although knowledge of delirium and interest in enhancing practice to prevent it was a key motivating factor for geriatricians’ involvement in the research, awareness of delirium was more variable among other staff. Although junior doctors might be familiar with the term ‘delirium’ and knowledge-based understanding was seen to have improved among registrars specialising in the care of older people, there was less confidence that such knowledge was routinely translated into action to prevent delirium or manage it when it occurred. For nursing and therapy staff, delirium had not featured in their professional training. Among all staff, delirium and delirium prevention were not included as part of mandatory training or in-service education programmes. This was seen to reflect the low salience attached to delirium and delirium prevention in policy and practice, such that, unlike other aspects of acute care delivery such as falls and pressure sores, there were no specific protocols relating to it in any of the sites.
Nursing, therapy and care staff generally did not use the term ‘delirium’; instead, the term ‘confusion’ or ‘acute confusion’ was more typically employed, particularly on elderly care wards:
It’s just that perhaps they don’t recognise it as delirium . . . they don’t put a label on it.
Doctor
Whatever the term used, among these staff, delirium was primarily understood in its manifestation as a problem for ward management and in the disruption it caused for other patients. Thus, awareness (unprompted) was predominantly of hyperactive delirium that resulted in difficulties for staff from problematic behaviour such as aggression, agitation, shouting and wandering. There was acknowledgement that hypoactive delirium, resulting in withdrawn, lethargic behaviour, could easily be overlooked in an acute environment. Indeed, staff awareness and understanding of the experience of delirium from the perspective of patients and caregivers was prompted through presentation of the evidence by the research team, and patients’ and caregivers’ concrete accounts of specific episodes of delirium during the workshops.
How staff perceived the nature, impact and consequences of the ‘problem’ of delirium affected how they sought to manage it. Awareness of ‘acute’ or ‘temporary’ confusion was seen to result in information-seeking from family and friends to determine whether this was of long-standing duration or of recent origin:
They might have been getting worse over a few weeks so, you really need to speak to a carer or relative; quite often we ring home care as well. We ring district nurses: ‘How are they normally? How have they been? Have you noticed any change in their condition over the last few weeks?’ . . . often the consultant . . . if we haven’t done it, will ask for us to get information from home care or whatever.
Senior nurse
Ascertaining that the change was recent and that the behaviour was atypical might precipitate a search for underlying causative factors contributory to the delirium (e.g. sepsis on elderly care wards), so as to identify solutions to address them (e.g. pain relief following surgery on orthopaedic wards). The practice consequences of identifying delirium also highlighted the process whereby delirium affects treatment and extends inpatient stay. Therapists, for example, indicated that mobilisation might need to be delayed to allow patients the chance to recover sufficiently to engage in rehabilitation:
That’s very common [delirium with sepsis], now those patients who are . . . acutely ill and we feel are in that stage, we don’t always try and do anything with them in the early days because we’re aware of the fact that, say they’d come in with a UTI [urinary tract infection] that, sometimes, just having a couple of days to recuperate means that, when we intervene, then they’ll have a much more successful outcome . . . they’re the sort of patients who we might discuss with the nursing staff and they’ll say, ‘leave it today’, you know . . .
Therapist
Some staff used the terms ‘confusion’ and ‘acute confusion’ interchangeably. This imprecision in language use denoted a lack of clarity about the distinction between acute confusion and dementia. The practice consequence was that search for a cause might not be pursued:
I think a lot of the times . . . it’s probably put down more to dementia than it is to delirium . . . when, I guess, so many people who have dementia are . . . more susceptible to having delirium . . . And then [for people with dementia], I think it’s probably more put down to: ‘they’re . . . out of their own environment, they’ve had a traumatic operation, they’re just more confused’, rather than there’s perhaps another underlying issue that’s causing it . . .
Therapist
The conflation of delirium and dementia by staff was a source of heightened anxiety and perplexity among caregivers/relatives, as the suddenness of the change and the strangeness of the behaviour of their relative was not understood by staff. Aggression and/or refusal to participate in treatment could be interpreted as ‘lack of engagement’, resulting in the patient being perceived as unsuitable for rehabilitation or berated by staff for ‘inappropriate’ behaviour. The following is an illustrative example.
Mrs Patterson’s (a pseudonym) brother-in-law was admitted to hospital ‘with a very high temperature and inflammation in his leg. I’m not quite sure what diagnosis was put on it’. He was also disabled following a stroke several years previously:
When he was admitted . . . everything went . . . during the night he just screamed for my sister . . . I went down to the hospital the following morning . . . it was obvious to me something was wrong . . . he was shouting and aggressive . . . and demanding. The nurse said to me when I went on the ward: ‘Oh, I’m glad you’re here, I want to say this in front of you [looking at the brother-in-law] that you are a very difficult man and we don’t like the way you’re speaking to us and if you continue we will refuse to nurse you’ . . . I said to her that he’s not like this usually . . . The doctor later confirmed that he had delirium.
Mrs Patterson
Generally, then, variability in how delirium was understood among different groups of staff and the lack of investment at an organisational level in respect of training and education meant that, in NPT terms, delirium identification had low coherence. Delirium diagnosis was primarily effected through use of observational cues, although how these were interpreted and acted on depended on the expertise of those making the observations. Thus, management practices following on from observations reflected the skills and interests of individual professionals, rather than a collective staff and ward response.
Delirium prevention
Given the low coherence of delirium among staff groups across sites, it is hardly surprising that delirium prevention was not perceived as meaningful. Even when senior staff had initiated action to increase awareness of delirium risk (e.g. posters displaying risk factors), this did not inform assessment and care practices: ‘it’s not in the foreground of people’s minds’ (geriatrician). Interviews and observations indicated that knowledge of delirium among individual staff did not necessarily translate into specific beliefs and behaviours (cognitive participation) and the organisation of work practices geared towards prevention (collective action).
In one elderly care ward, we observed that senior nursing staff employed the term ‘delirium’ in describing specific patients, and demonstrated awareness and knowledge of both hypoactive and hyperactive delirium, as well as sensitivity to the distress caused to patients with it. The consultant geriatrician also had a particular interest in delirium. During observation of a nursing handover meeting on this ward, it was reported that just under one-fifth of current patients were characterised as having delirium. One of the patients with delirium discussed was perceived as needing considerable assistance with eating and drinking; another was referred to as having ‘hypoactive delirium’, ‘really drowsy’, not sufficiently alert to eat and drink, incontinent and ‘on IV [intravenous] fluids’. For the former patient, it was emphasised that all staff should be alerted to ensure support at mealtimes and to encourage drinking and eating. For the latter, it was agreed that she should be moved to a bed that was more visible from the nurses’ station, although this also provoked discussion about the disorientation such a move might cause. At the same meeting, several newly admitted patients were described as having symptoms that, to the observer, might portend risk of developing delirium: an 89-year-old patient who had experienced multiple urinary tract infections and been admitted following a fall, and an 80-year-old patient with a urine infection and pneumonia who had suffered a heart attack and needed oxygen. The symptoms were presented without reference to or discussion about delirium risk or any specific preventative action to be taken. This was recognised as typical practice by the development teams. Thus, even when there was shared understanding of delirium management, this did not facilitate noting and acting on risk factors before delirium had occurred (cognitive participation in NPT).
One exception to this general gap between knowledge and practice in delirium prevention was the development and implementation of a protocol on pain management post surgery for use in hip fracture patients on the trauma orthopaedic ward. This was aimed at delirium prevention. Staff remarked on how the protocol was routinely pursued, with positive outcomes as a consequence, particularly in reducing the severity and duration of delirium episodes. The ward manager attributed success to specific features of the hip fracture patient pathway: this was direct, linear and highly protocolised, with all patients diagnosed with hip fracture fast-tracked from the A&E department to the ward to undergo surgery within 24 hours. Insertion of the protocol into the pathway was viewed as an elaboration of existing practice, rather than as a major shift in how things were done. Practice change was reinforced by the perception among nursing staff that this was a relatively simple intervention with visible positive effect in a short period of time.
By contrast, the patient journey to elderly care wards across the three hospitals was more protracted and diverse. Triage systems and initial investigation in A&E to determine a differential diagnosis and whether or not admission was warranted might be followed by further observations in a short-stay assessment facility for up to 48 hours, which could be further protracted because of a shortage of acute ward beds. The chaotic nature of A&E and short-stay assessment facility environments, compounded by the multiple potential aetiologies of delirium in these settings, was viewed as contributing to delirium risk so that the scope for preventing incident delirium on acute wards could be adversely affected by the length of the patient journey into them. Even so, delirium prevention was considered to be feasible and worthwhile in the acute ward environment, although organisational factors shaping the patient journey through the hospital also needed attention as part of a strategic approach to prevention.
Current ward routines and practices
From interviews with staff and development team discussions, the acute care ward was reported as ‘busy’, often ‘chaotic’ and challenging: a picture reinforced by research observation. Explanatory, contributory factors offered by staff included patient mix and the policy and organisational imperative to achieve rapid patient throughput. Policy and service emphasis on hospital admission of those who required specialist medical and nursing expertise that could not be provided in alternative settings meant that patients were very acutely ill. Similarly, it was expected that patients would move on from acute care once medical and functional needs were met to secure ‘safe’ discharge. Patient moves within wards across all sites were also common, reflecting various organisational contingencies.
The hectic nature of ward life had the consequence that routine practice was described by staff as being primarily directed at responding to what was immediately presented, with priority given to diagnostic, observational and interdisciplinary assessment and care-planning. This picture was reinforced from observation. Thus, particularly for nursing and auxiliary staff, ward life was organised on the basis of a structured rhythm of time-sequenced care (washing, toileting), observations, diagnostic processes and treatment, punctuated by meals and visiting times, a pattern that was prone to disruption as a result of crisis events. Alongside this daily rhythm was the management of patient flow (admissions and discharges) and associated activities (negotiating with bed managers, discharge co-ordinators, social workers, relatives and community agencies), including record-keeping.
The variability and general understanding of delirium prevention among staff meant that it was not a significant driver of ward care practice. However, although delirium preventative interventions relate primarily to features of care quality, it is pertinent to consider how relevant routine care practices were accomplished, including the barriers and contextual factors that affected them.
Nutrition, fluids and sensory aids
Although nutrition and fluid intake were viewed as components of ‘basic’ care to be undertaken by ward staff, they were primarily delivered by health-care assistants (HCAs). In each site, around one-third of patients required some direct help at mealtimes. Others might need encouragement to eat, although this provision depended on staff availability and assumed lower priority. Similarly, tasks of washing and dressing, including ensuring that patients had spectacles, dentures and hearing aids, as appropriate, were mainly undertaken by HCAs. Even so, the importance that senior nursing staff attached to care tasks affected both the value attributed to them by junior nursing staff and the extent to which they pitched in to provide assistance.
Mobilisation
Mobilisation by physiotherapy staff of patients with particular needs was limited: it appeared to occur, at most, once daily, and was intended to be augmented with support and encouragement from ward staff. Patients who merely lacked confidence in getting up and walking on their own were reliant on nurses and HCAs to provide this. Similarly, local policies on prioritising therapy for those with the potential to resume independent living meant that, for example, in one site, patients admitted from nursing homes did not receive therapy. The engagement of nurses and HCAs routinely in mobilisation work in either an enhancing or a supportive role was viewed by staff as essential to sustaining mobility among patients, most of whom were of advanced age, frail and unsteady on their feet. How consistently this was done depended on factors such as the ward physical environment and other pressures.
In one site, the confined and cluttered space of the bays was a constraint on the ability of patients to move safely, and, as the distance between bed and toilet was not more than a few steps, routine mobilisation by nurses and HCAs was limited. Only therapists walked patients for longer distances along the corridor, where the wider space enabled freer and more confident movement. In another site, by contrast, the distance between the bed and toilet was some 10–20 m. Part of the ward routine included nurses and HCAs providing direct assistance to patients and/or keeping an eye out for them as they walked from bed to toilet. It was noted over an observation period how one patient progressed from being assisted with walking to managing independently with a nurse walking behind her, and then to walking on her own. Although here the physical environment was conducive to staff encouraging mobility, this practice was facilitated and reinforced by a care ethos that placed high value on all staff, including nurses, participating in such work. This is exemplified in the following episode observed in this site, but not in others. One of the nurses was with a patient as she encouraged her to stand up from being seated. As the patient made several attempts to propel herself from a sitting to a standing position, the nurse stood by continually encouraging her by showing her how to use her arms to push and move to the edge of the chair, and praising each effort until she stood up.
Orientation and communication
Features of the ward physical environment may act as constraints to ‘good’ care practice, for example inappropriately placed clocks, lack of space or infection control policies precluding personal possessions.
There was variation between individual staff and professional groups within and between sites in the extent to which they conversed with patients in the course of their work. Therapists, for example, typically introduced themselves to patients they were working with, engaging them in general, social and orienting conversation. The pattern was more diverse among nursing and care staff. In one site, there was a buzz of chatter in the bays as nurses and HCAs conversed socially with patients as they went about their daily routines of washing, dressing, medication rounds and mealtimes. The progress of individuals was remarked on and patients were complimented on efforts at walking or dressing. In another site, interaction between staff and patients seemed primarily directed on the task in hand: ‘here are your tablets’, ‘do you have any pain?’.
Cognitive stimulation and therapeutic activities
The hustle and bustle of ward life, particularly from early morning to mid-afternoon as described by staff, was in marked contrast to the silence and inactivity of patients once care needs and clinical observations were completed. We could discern two parallel but distinct ward rhythms: a staff rhythm marked by frenetic movement and continuous noise – buzzers, telephones and the clatter of trolleys – and a patient rhythm distinguished by a paucity of conversation and little movement. Sustained or prolonged engagement of patients by staff was absent in all sites. Development teams remarked that this was neither feasible nor valued in the context of the priority attached to moving patients quickly through the system.
Overall, some practices pertinent to delirium prevention (assistance with meals for those who needed help with eating but not for those who required encouragement) were carried out more or less consistently for some patients across all sites. Other practices (enabling support to encourage and enhance mobility among patients lacking in confidence, and personally meaningful, as opposed to task-based, communication) were accomplished more consistently in some sites than others, depending on local policies and priorities, the physical environment in which care was delivered and the existence of a care ethos that placed high value on social engagement and care. Yet other practices (spending time with patients in one-to-one conversation or engaging patients in cognitively stimulating activities) were not routinely engaged in by staff across sites, which was seen to reflect the current acute care environment. Collective action by ward staff in practices that are preventative of delirium were contingent on local policies and priorities on patient need, staffing levels, division of labour and the care culture operating. In no site were any of these explicitly linked with delirium prevention or engaged in consistently for all patients who might exhibit delirium risk factors.
