Practices of falls risk assessment and prevention in acute hospital settings: a realist investigation

Inpatient falls

Falls are generally defined as ‘an unexpected event in which the participant comes to rest on the ground, floor, or lower level’. 2 They are the most common type of safety incident reported by acute hospitals. 3 More than 240,000 falls are reported in acute hospitals and mental health trusts in England and Wales each year,4 although under-reporting may mean the true incidence of falls is higher. 5,6 Falls are most common in patients aged 65 years or older, representing 77% of inpatient falls. 3 The majority of falls result from multiple interacting causes, most commonly age-related physiological changes, medical causes, medications and environmental hazards. 7

Overall, 28% of inpatient falls result in some level of harm and patients aged 65 years or older are more likely to be harmed. 3 The proportion of falls resulting in any fracture ranges from 1% to 3%, with reports of hip fracture ranging from 1.1% to 2.0%. 6 In 2015–6, inpatient falls in England resulted in 2500 hip fractures. 8 Outcomes for patients who acquire hip fractures in hospital are far worse than for those in the community who acquire hip fractures, with significant differences in mortality [relative risk (RR) = 3.00; 95% confidence intervals (CIs) 1.05 to 8.57], discharge to long-term high-level nursing care facilities (RR = 2.80; 95% CIs 1.10 to 7.09), and return to preadmission activity of daily living status (RR = 0.17; 95% CIs 0.06 to 0.44). 9

Even where no physical harm occurs, falls can lead to fear of falling and associated loss of confidence. 5,8 They can result in slower recovery,8 even when physical harm is minimal, and can have longer-term consequences for the patient’s health, as fear of falling may lead to restriction of activity and associated loss of muscle and balance function, increasing risk of falling. 5 Falls can also be a cause of significant distress for families and staff. 6,8 Falls in hospital are a common cause of complaints10 and can be a source of litigation. 11 They are also associated with increased length of stay and greater amounts of health resource use. 6 NHS Improvement (now part of NHS England) estimated inpatient falls cost the NHS and social care an estimated £630 million annually. 3 It is therefore a priority to reduce the number of patients who fall, and their risk of injury, in acute hospital settings.

Falls risk assessment

The traditional approach to managing falls in acute hospitals is to complete a falls risk prediction tool, sometimes referred to as falls risk screening tools or falls risk scores (such as STRATIFY12). Such tools typically provide a list of falls risk factors, assign a numerical value to the presence or absence of the risk factor, and then sum the numerical values together to represent the individual’s risk of falling (high, medium, low). 13 Interventions are then used to target individuals at high risk. 14 There are issues with the predictive validity of such tools; a systematic review of falls risk prediction tools found only moderate accuracy, comparable to the accuracy of nursing staff clinical judgement. 13 Consequently, such tools may either provide false reassurance about patients identified as low risk or result in most patients on a ward being identified as high risk. 14 Such tools are often completed only once, typically on admission, while a patient’s risk of falling can vary over time. There is also concern that their use gives false reassurance something is being done, even if no action to address falls risks has been taken. Additionally, with a tool of this kind, actions tend to be linked to the score and can lead to a ‘one size fits all’ approach even though the issues and needs of individual high-risk patients can be very different. 14 A stepped-wedge cluster randomised controlled trial (RCT) showed removing the risk score component from falls risk prediction tools does not negatively impact falls outcomes and can reduce time spent completing paperwork. 15

In light of the limitations of falls risk prediction tools, the NICE guideline on falls in older people states they should not be used and instead a MFRA should be undertaken. 16 The recently published World Falls Guidelines, for falls prevention and management for older adults, also recommend patients in hospital should receive a MFRA and advise against using falls risk prediction tools. 17 A multifactorial approach to falls risk assessment identifies individual risk factors for each patient, which may make them at risk of falling and that can be treated, improved, or managed during their stay (what tend to be referred to in the falls research literature as ‘modifiable’ risk factors). MFRAs, unlike risk prediction tools, do not include unmodifiable risk factors (i.e. cannot be treated, improved, or managed) such as age and sex. The NICE guideline includes the following modifiable risk factors: cognitive impairment; continence problems; falls history, including causes and consequences (e.g. injury and fear of falling); unsuitable or missing footwear; health problems that may increase a patient’s risk of falling; medications that increase the risk of falls; postural instability, mobility problems and/or balance problems; syncope syndrome; and visual impairment. The NICE guideline states that a MFRA should be undertaken for all inpatients 65 years or older and inpatients aged 50–64 years judged to be at higher risk of falling due to an underlying condition. Based on this assessment, a multifactorial intervention should be provided for the patient, tailored to their individual risk factors. For example, if visual impairment is identified, it might be decided that an optician visit should be arranged if the patient has lost their glasses or, if there is no known reason for poor eyesight, an ophthalmology referral is made. 18 In this way different patients, who have different risk profiles, will receive different interventions to reduce their risk of falls.

Preventing inpatient falls completely would only be possible with unacceptable restrictions to patients’ independence, dignity and privacy, such that some falls may be considered an inevitable consequence of promoting rehabilitation and autonomy. 6,10 Thus, there is a need to balance the risk of harm from falls and the risk of deconditioning. Nonetheless, it is estimated introduction of MFRA and tailored interventions, as recommended by the NICE guideline, could reduce the incidence of inpatient falls by 25–30% and the annual cost of falls by up to 25%. 3 Despite the NICE guideline being updated to include these recommendations in 2013, the 2022 National Audit of Inpatient Falls (NAIF) report noted that 34% of Trusts are still using falls risk prediction tools and, while there has been improvement in the proportion of patients receiving documented assessment for components of the MFRA included in the NICE guideline, there has been a reduction in the proportion of patients assessed for delirium. 19 Documented vision assessment (52%) and lying and standing blood pressure (LSBP, 39%) remain concerningly low. In interventions, a mobility care plan was in place for 90% of patients who required one, a continence care plan for 78% of patients who required one, and a delirium care plan for 61% of patients who required one. This suggests variation in the extent to which the NICE guideline is being followed and opportunities are being missed to reduce the likelihood of inpatient falls.

Given these findings, it is necessary to understand the contextual factors that support and constrain use of MFRA and tailored falls prevention interventions in acute hospitals, to improve practice.

Phase 1: theory construction

Phase 2: theory testing

The CMOcs prioritised for testing encompassed similar mechanisms. To facilitate searches, we grouped mechanisms where appropriate, identifying six key concepts. The searches were conducted in two stages, based on the six concepts.

Phase 3: theory refinement

For theory refinement, NVivo was used to categorise sections of the manuscripts to support theory testing. For example, for facilitation, an overarching theme ‘Type of facilitation’ in NVivo had subthemes relating to alerts and reminders, and decision support. A sample of manuscripts was reviewed to identify additional themes, for example, in relation to study details (including intervention descriptions and study rationales) and influences on staff practices, with the development of subthemes, such as individual staff beliefs and attitudes. The themes were added to NVivo and tested for usability using a sample of manuscripts (n = 5) that varied in aims and methods and included qualitative and quantitative data. Researchers coded the manuscripts and met to discuss their experiences, suggest refinements, and develop a shared understanding of how to apply the themes to the texts. After refining the coding framework, manuscripts from the four concept searches were imported into NVivo and coded using the framework by NA and LM.

All manuscripts were coded using an overarching framework. Researchers began analysis of two CMOcs: one focused on implementation – facilitation – and one focused on falls risk reduction – patient participation. Details about the studies reported in these texts, including methods used, settings, samples, intervention description and comparator (if appropriate) were extracted into Excel.

Data from the concept-specific searches were collated. Analysis began by examining the interventions described by study authors, to understand the extent to which they reflected the resource component of the initial CMOc. The type and range of interventions were then described in narratives, prepared by the researchers. Following intervention description, manuscripts were reviewed to understand what the data suggested about the outcome of interest in the CMOcs. For example, for facilitation, the primary outcome of interest was whether recommended practices (e.g. MFRA) were implemented as intended, and for patient participation, the primary outcome of interest was the extent to which falls rates were reduced. Outcome data were recorded in tables to understand variation in intervention impact across studies.

After tabulating outcomes, researchers examined coded data and the original manuscripts to collate evidence to help explain variations in outcomes, and to assess the extent to which the mechanisms and context expressed in the original CMOc were evidenced by the literature. In doing so, they looked for data that supported and diverged from the CMOc logic. Narrative summaries were then written and were used to refine the initial CMOcs.

Following the steps of analysis depicted in Figure 1, the aim was to refine the first two CMOcs and then continue with the remaining two if time allowed. Although coded, time limitations meant we were not able to undertake theory refinement for leadership and shared responsibility, so the decision was made that data for testing the leadership CMOc would consist of the review of falls prevention policies (see Review of acute NHS Trust falls prevention policies) and data collected in the multisite case study, while data for testing the shared responsibility CMOc would consist of data gathered in the multisite case study only.

FIGURE 1.

Realist review analysis flowchart.

Sampling of case sites

Data were collected across three NHS acute Trusts in England. This number was chosen to provide a balance between breadth and depth of investigation,45 enabling identification of organisational-level factors that impact on falls risk assessment and prevention while providing confidence in the generalisability of findings that are consistent across sites. Trusts were selected to ensure variation in key NAIF indicators at the time of writing the proposal,8 the HIT in place, and to include both teaching and district general hospitals (Table 4).

We chose to sample across clinical areas in each site to enable us to be able to distinguish differences due to clinical area and Trust/unit-level factors. We decided that, in each Trust, we would undertake data collection in one older person/complex care ward and one orthopaedic ward. These areas were selected because they would provide different patient populations with different lengths of stay and different staff experience, with longer length of stay and staff having experience in managing older people at risk of falling on older person/complex care wards. Specific wards were identified through discussion with local collaborators at each site. An introduction to each ward is provided in Table 5. The two wards for Site 1 were located in different hospitals within the Trust; for Sites 2 and 3, the two wards were located within the same hospital.

Data collection

Analysis

Qualitative data analysis followed the steps of framework analysis. 48 The researchers began by familiarising themselves with the data by reading a selection of the observation and interview transcripts – a process facilitated by ongoing reading of transcripts throughout data collection. Researchers then met to discuss construction of a thematic framework to facilitate CMOc testing. Based on previous experience of indexing data for CMOcs, the decision was made to minimise the number of themes by keeping themes abstract to encompass explanation of mechanisms and contextual influences for example, for facilitation, ‘use of physical artefacts’ encompassed description of tools used, how they were used by staff and factors that appeared to support or constrain tool use. The thematic framework was then used to index the data. A series of matrix displays, based on the case dynamics matrices described by Miles and Huberman,49 were used as a next stage in analysis to facilitate cross-case analysis and obtain an overview of the data. One matrix display was produced for each site and for each CMOc being tested (12 in total). The matrix column headings summarised the content of the starting CMOc – clarifying the hypothesised contextual influences, resources offered, participant responses and impacts. In each matrix row, data were summarised, with reference to the indexed data, by Trust and ward, with rows representing organisational site interviews, orthopaedic ward staff interviews, orthopaedic ward observations, older person/complex care ward staff interviews, older person/complex care ward observations. The frameworks informed the site presentations, where findings were presented and discussed with participants, and then formed the basis of narrative summaries that described how the data supported or suggested a refinement or addition to the CMOc.

In addition, the Lay Research Group undertook an analysis of the qualitative data. After defining the task and what qualitative data analysis involved, the Lay Research Group members had about a month before the data analysis session to consider individually two sets of observation notes (one from an orthopaedic ward and one from an older person/complex care ward) and two interview transcripts (one from a patient and one from a carer). Following a general discussion about the materials and the effect of reading them on members, the Lay Research Group shared their thoughts on the individual observation notes and interviews. A number of patterns or themes were identified in the text, including themes about acknowledging patients as people and the impact staff attitudes had on falls prevention. Their analysis resulted in the recommendation of some approaches in falls care planning, such as using imagery (e.g. picture cards) in interactions with patients with communication difficulties such as people with dementia, or speak different languages. Their analysis was reported to the project management group and fed into the wider data analysis process described above. Because of this, their analysis is not reported separately in the following results chapters, but is, rather, incorporated within them (e.g. the importance of knowing patients as people is reflected in the analysis of case study data relating to patient participation in Chapter 8). Reflections on the results or outcomes of PPI itself on the study are reported in Chapter 9.

Intervention characteristics

Intervention characteristics in the CFIR cover how intervention features, such as complexity, strength of evidence, and adaptability, influence implementation. The included papers discussed implementation of national guidelines for falls prevention and the development of specific tools to assess and document falls risks and prevention strategies. They also considered interventions designed to support or improve implementation of guidelines and associated tools, such as the use of champions and falls education and training. In this section we focus on characteristics of national guidelines and the assessment tools themselves and discuss other types of intervention within the remaining domains.

Barker70 compared the 2004 and 2013 versions of NICE guideline on falls in older people, pointing out that a lack of integration between the two guidelines (the first focused on community care while the second was updated to incorporate inpatient falls) made the 2013 version difficult for clinicians to interpret, owing to visible differences in writing style and layout of the two elements. These differences were said to affect the quality of the guideline and make interpretation more difficult for clinicians, especially those who worked across care settings. Additionally, they noted that, to support clinicians, NICE had developed a falls in older people pathway, designed to be used interactively on the NICE website, although this might not be feasible for clinicians during their working day. There is a paper alternative, but it was 12 pages long, difficult to follow and not considered user friendly. This suggests ease of access to recommendations and their presentation may influence their use in practice. Furthermore, Glasper71 referenced a number of guidelines and resources, developed by NICE, the Royal College of Physicians, and the Care Quality Commission, available to support falls prevention in hospital but noted that ‘no matter how much advice bodies such as NICE produce it is often nurses who have to reconcile the reality of care delivery and the quest to reduce falls’ (p. 807), pointing to the practical day-to-day challenges of delivering recommendations, which are likely to vary between organisations depending on resources available.

Staff characteristics

The CFIR domain ‘characteristics of individuals’ refers to people’s knowledge and beliefs about the intervention. These factors may support the implementation of falls risk assessment and prevention practices, or they may constrain it, even if the intervention has features deemed supportive. For example, Miake-Lye et al. 56 discussed nurses’ ‘buy-in’ or commitment to falls prevention as an implementation support, but also noted the need to change the attitude that ‘nothing can be done’. Changing individual and group attitudes or motivation towards falls prevention programmes was discussed in many of the manuscripts reviewed,56,75,79 often with reference to the use of educational and training interventions to increase knowledge of falls risks and, consequently, attitude towards intervention delivery. For example, Glogovsky80 discussed how, when nurses understand falls prevention interventions, they are more engaged in preventing falls. A nurse supported this assertion; Johnson81 reported that reading a continuing professional development (CPD) article improved their knowledge of falls risk assessment tools, including the importance of completing an assessment for older inpatients both on admission and regularly during their hospital stay. They explained that this knowledge increased their confidence in undertaking the assessments and that they were more aware of the importance of taking a holistic approach, incorporating risk factors that they had not considered previously, such as oxygen tubing, which can present a trip hazard. The CPD article outlined the physical and psychological factors nurses should consider when undertaking assessments, and the effects that older people may experience because of a fall. The importance of engaging practitioners undertaking the assessment in this way was highlighted by Lindus,82 who described a nurse’s realisation of ‘going through the motions’ when undertaking a falls risk assessment. The nurse reflected that they were shown how to complete the falls prevention paperwork but had not made the association between the risk and the specific patient assessed.

Process of implementation

‘Process of implementation’ in CFIR refers to strategies or tactics that might influence implementation. These might include engaging appropriate individuals in implementation and reflecting on and evaluating progress. Some of the previous discussion has highlighted examples of such strategies, for example, the role of falls champions in the process of implementation, and the use of audit and feedback to evaluate the process of delivery and associated outcomes. QI frameworks were also used to support implementation of recommended falls practices, through a process of incremental improvement in the care processes under study. Approaches cited include the Define, Measure, Analyse, Improve, Control process80 and Plan, Do, Study, Act. 60 These types of improvement methods were also reported in posters from falls collaboratives in NHS Trusts and NHS Improvement falls collaborative case studies, retrieved from the Google searches, in which numerous interventions were used to improve process and outcome, often including use of champions and education and training.

Inner setting

‘Inner setting’ in the CFIR refers to features of organisations, such as implementation climate and leadership. Miake-Lye et al. 56 identified leadership support from hospital directors and unit directors or clinical champions on wards as a theme of effective implementation in their systematic review of falls prevention programmes. They commented that while leadership may be considered an implementation support for programmes in general, more specific to falls prevention is multidisciplinary development and guidance within falls prevention programmes and overcoming the idea that falls are inevitable. Supporting this idea, Sutton et al. 74 suggested that the positive impact of leadership included inspiring others and ensuring falls prevention was a priority on wards, while Spoelstra et al. 75 discussed that organisation-wide training to educate staff about falls prevention would help develop a culture of safety that may engage staff in the practices necessary to ensure a safe environment for patients. West et al. 90 assessed nursing interventions to reduce patient falls and suggested that implementation of existing falls policies might be constrained by a high rate of turnover in leadership roles on clinical units and at the middle-management level, which meant leaders lacked experience to address implementation challenges. They noted that auditing compliance with local falls policies provided an opportunity to ‘educate novice leaders’ about the importance of evaluating the clinical practice of their staff but acknowledged that the daily challenges of operations and staffing issues may get in the way.

The staffing levels of wards and units was also discussed as a potential support or constraint on implementation of MFRA and prevention strategies. Christy68 argued that adequate staffing not only gives nurses time to provide hourly rounds to meet patient needs, reducing the likelihood of patients mobilising unassisted, it also gives them sufficient time to think critically and intervene to prevent falls. In their systematic review, Choi et al. 69 reported that most included studies implemented a considerable number of care process-related interventions that may demand time and effort from nurses. Environment-related interventions, the authors noted, could help to reduce falls by supporting the work of nurses, but despite evidence about the importance of the physical environment only a few hospitals were identified as introducing such interventions as part of their multifaceted falls intervention strategies.

Outer setting

‘Outer setting’ in the CFIR encompasses external policies and incentives, as well as whether patient needs and resources are prioritised by organisations. The influence of external policies was discussed in several texts. For example, Barker70 considered the impact of the NICE guidance that instigated our study, indicating potential constraints such as lack of funding. They note that the costing statement for the guideline stated that implementation is unlikely to have a significant cost impact for the NHS, based on the possibly incorrect assumption that most parts of MFRA and prevention are already standard practice.

In the USA, Spoelstra et al. 75 speculated that a motivation for healthcare managers and staff to reduce falls rates was the decision by the Centre for Medicare and Medicaid Services in 2008 not to continue funding costs associated with inpatient falls. Other policy drivers highlighted in this paper included hospitals’ desire to attain Magnet status and the requirement by the Joint Commission for accredited hospitals to conduct falls risk assessments for their patients. Mashta76 reflected on how falls prevention became a top priority for many UK hospitals as one of the eight ‘high impact actions’ developed by nurses at the request of England’s Chief Nurse in 2010.

Delivery of an MFRA and care plan

These If–Then statements relate to implementation, that is, what works to ensure that consistent, comprehensive, and ongoing MFRA and tailored interventions are delivered to eligible patients.

1. If there are nurses who are trained and dedicated to delivering falls prevention (e.g. falls leaders or champions), then they will lead implementation of the multifactorial falls prevention strategy on their ward, ensuring a consistent and co-ordinated effort to deliver the strategy across professional groups.

2. If assessment tools (including HIT) and processes are relatively quick and easy to use, and/or facilitate existing routines, then nurses will choose to integrate them into their work processes, helping to ensure that eligible patients receive a MFRA and appropriate interventions.

3. If ward staff receive feedback about their successes and challenges in delivering their multifactorial falls prevention strategy, which aims to promote learning rather than apportion blame, then they will feel supported and empowered to overcome implementation challenges.

4. If a community is in place across hospitals for falls leaders/champions to share experiences of delivering multifactorial falls prevention strategies, then they will learn from other hospitals and implement successful approaches to support delivery of such strategies.

5. If there is training and education for falls prevention available to ward staff, then they will understand the importance and benefits of multifactorial strategies for patients, carers, and the service, and feel empowered to, and/or accountable for, delivering ‘best practice’ in the form of patient-centred assessments and interventions to all eligible patients.

Delivery of interventions tailored to individual falls risk factors

These If–Then statements relate to falls prevention, that is, what works to prevent/reduce patient falls in an acute setting.

1. If a multifactorial, patient-centred strategy is delivered, then patients (and/or their carers) will understand their risk of falling and why particular preventative interventions are being used, and feel empowered to take part in those interventions, thereby reducing the risk of falling. For example, patients could use the call-bell/ask ward staff for help in getting up and moving about.

2. If multifactorial strategies include sharing information between ward staff (e.g. through posters/safety huddles/handovers), then they will develop a shared responsibility for falls prevention and become more vigilant of patients at risk from falls.

3. If falls training, education and tools are available to ward staff, then they will develop expertise in falls that they will apply in practice, for example, by asking patients more in-depth questions during multifactorial assessments to ascertain their risk of falling, leading to identification of risks relevant to each patient.

4. If falls incidences are recorded, analysed and discussed by ward staff, for example, in safety huddles, then they will learn about how, where and when falls are most likely to occur, enabling them to make changes to work patterns and routines to support multifactorial falls prevention strategies, for example, by locating high risk patients’ beds in visible locations.

5. If multidisciplinary team members are available at the right times, then they will participate in the falls risk assessment and care planning process, leading to team engagement and co-operation in delivering the multifactorial strategy.

6. If wards are adequately staffed, then nurses will have the time to think critically about the falls prevention strategy and work to ensure patients’ needs are met, for example, by providing enough for patients to drink, managing pain effectively, and helping patients to get up to visit the toilet; therefore patients who have difficulty walking safely on their own will be less likely to do so without assistance.

Organisational-level leadership for falls prevention

In each site, the ultimate responsibility for falls prevention fell to Boards of Directors and Chief Executives, and this was delegated to various individuals and bodies, including Chief Nurses or Directors of Nursing at organisational-levels and matrons at clinical service unit or departmental-levels. Each Trust also had committees or groups to which Boards of Directors delegated responsibility for falls prevention, and some also linked to multiprovider Falls Collaboratives. At Site 2, for example, the Collaborative was a city-wide multidisciplinary group reporting to several committees including the Trust Professional Board, which aimed to provide strategic direction for falls prevention and to feed learning and good practice back to key stakeholders, including ward managers. Membership comprised clinical and academic experts in falls management, including matrons, falls, and improving quality leads and local commissioners. Sites 1 and 3 had similar Collaboratives. At Site 3, for example, a multidisciplinary Collaborative reported to several committees, including the Trust’s Clinical Outcomes Group, and its membership included matrons, falls champions or link practitioners, and representatives from therapy, data and moving and handling teams.

Each Trust also had organisational teams who supported implementation of falls policies. These teams typically had both organisation-facing, assurance elements to their work as well as ward-facing, supportive elements. In Site 1, for instance, an organisational team, working alongside several other quality and safety-focused teams, provided assurance on key nursing performance indicators and patient safety markers, such as completion of MFRAs and falls care plans. They reported outcomes to governance committees, the Board of Directors and the chief nurse. The team also offered falls prevention induction and refresher training to staff. At Site 2, a team of falls specialist nurses provided corporate reporting and support for wards, for example, training falls link practitioners (see Falls champions). At Site 3, a smaller Trust than Sites 1 and 2, the Dementia Lead was responsible for implementing interventions and strategies to reduce the risk of patients falling and providing staff education and guidance on falls prevention. They also oversaw an enhanced care team, which offered one-to-one support for people at increased risk of falling and maintained strong links with matrons. The role of matrons in falls leadership was also highlighted in interviews across sites.

Falls champions

The role of falls champions was identified in all three sites, although postholders were referred to as falls link practitioners. This role was allocated to nurses and HCAs, undertaken alongside their normal roles. In practice, however, there were active link practitioners in Sites 2 and 3 only. During observations, we observed them working on their wards, as this fieldnote extract from the Site 2 older person/complex care ward describes:

The falls link nurse is chatting with the patient in bed 5. They are trying to convince them to have an ECG but the patient is saying that they don’t want one. The link nurse is explaining it’s because of their fall this morning – they’ve done her blood pressure, her blood, and an X-ray of her hip.

Duties of falls link practitioners were outlined in organisational falls prevention policies in Sites 2 and 3, although staff acknowledged that pressures of work on wards, aggravated by COVID-19, meant it was not always possible to fulfil the duties described in the policies. Trusts aimed to appoint HCAs as well as nurses as falls link practitioners, because the former made a fundamental contribution to falls prevention and they believed staff were more likely to be inspired by someone in a similar role. Reflecting this, in Site 2, a role descriptor specified that link practitioners were expected, among other duties, to act as role models for falls reduction in their clinical areas, and to provide advice and education for the assessment, intervention, and management of patients who had fallen or were at risk of falling. They were also expected to facilitate regular audits of falls management practice on wards and raise patient safety concerns.

Falls link practitioners at Site 2 received training in quarterly 2- to 3-hour sessions where specialist nurses provided up-to-date information about falls prevention practices and procedures. However, it could be difficult for link practitioners to attend the training they were entitled to, owing to the pressure of work on the wards. On one of the wards we observed, the falls link practitioner showed us a falls file they maintained, which was kept in the staff room and accessible to all staff. Contents included a role descriptor for falls link practitioners and PowerPoint slides about enhanced observations of care (EOC), with information about calculating risk scores. These scores were used to identify patients in need of one-to-one care or being placed on a cohort bay (i.e. a bay where patients assessed as being at risk of falling were co-located). The file also included information about adult falls risk assessment and care planning, including how to create a falls care plan on the EHR; a section on postfalls actions; and copies of audits about falls process compliance (the last audit was dated around 8 months before).

In the orthopaedic ward in Site 2, a senior nurse explained that falls link practitioners updated and maintained a falls board. This board listed, under a heading of ‘Falls Prevention’, the following strategies: call-bell within reach, low rise bed if available, cohort bay if available, correct footwear, and risk assessment completed and updated. The board also displayed a note that EOC scores should be updated daily. The nurse explained that a recent increase in falls was thought to be due to staff scoring patients too low on the EOC scores, therefore work was being undertaken to ensure scores were completed accurately. There was also a link nurses’ board, which gave the names of postholders.

At Site 3, the falls link practitioner role was being reviewed and relaunched, given that it had not been possible to prioritise link practitioner work during COVID-19 and not all wards had them. Falls link practitioners at Site 3 were expected to support good practice in falls prevention and management in their wards by attending training and relevant meetings, including the Falls Collaborative, and by giving information and advice to colleagues, patients and carers. Visual displays about falls prevention, like those described in Site 2, were observed in Site 3. Whether updating this information was the role of the falls link practitioner was unclear, but the nurse allocated this role in the older person/complex care ward described reinforcing messages about falls prevention to colleagues. However, like Site 2, the Site 3 link practitioner described finding the role challenging because they were busy as a nurse. Efforts were made to address this issue, for example, by booking meeting and training dates well in advance, so that ward managers could take link practitioners’ absences into account in rotas, and by offering shorter training events, for which link practitioners received CPD points and certificates, to encourage attendance. Such training was intended to empower falls link practitioners to educate colleagues about falls prevention, and we saw some evidence of that in the falls boards on wards, which disseminated information about prevention practices.

We were not able to identify falls link practitioners on either ward at Site 1: on one ward, the post was vacant, and on the other ward, a member of staff explained that a more fluid approach was being taken, whereby a nurse or HCA was designated each day to lead on falls: ‘In the Safety Huddle on a morning the nursing staff would take it in turns. So they’ll identify someone each day to be a champion for falls risks, nutrition, skin care, toileting, etc.’ (Physiotherapist, Orthopaedic Ward, Site 1). However, in observing the safety huddle on this ward, we did not see this happening. The participant noted that, while the role of falls link practitioner might be useful, it would be difficult for one person to find time to do it. They felt the current model of ‘collective’ working, with multiple staff members sharing responsibility for falls prevention on the ward, was ‘fine as it is’. As in Site 2 and Site 3, falls prevention messaging directed to staff was displayed on wards.

Leadership of falls prevention practices on wards

Falls prevention practices of interest in this study focused on the implementation of MFRAs and interventions tailored to address individual risk factors. At Site 1 the organisational falls prevention policy stated that ward managers (rather than falls champions) should ensure all patients were assessed for their risk of falling according to the Trust’s guidelines. The Site 1 policy also stated that ward managers should make sure staff were aware of their falls assessment and prevention responsibilities, including undertaking falls refresher training every 3 years; investigate causes of all incidents and ensure actions and recommendations were acted on to prevent recurrence; and ensure falls prevention information was displayed in patient areas and provided to patients at risk of falling. A ward manager in the Site 1 older person/complex care ward explained how, as a manager, they identified areas for improvement via monthly review of metrics and, through shared learning with other wards, implemented changes that had improved compliance in completing risk assessments on their ward. The ward manager had introduced a handover board, which helped staff focus on improving practice in specific areas, such as carrying out falls risk assessments. Ultimately this messaging was directed at nursing staff because, across all sites and wards, nurses were responsible for completing falls risk assessments and care plan for patients, and sisters and nurses-in-charge of shifts supported delivery of certain falls prevention interventions. In other words, nurses with different levels of seniority on the ward had day-to-day responsibility for delivering key areas of falls prevention practice, including leading discussion about patients’ falls risks in safety huddles and multidisciplinary team meetings.

Staff told us that falls link practitioners’ authority was less well defined than ward managers or nurses-in-charge of shifts and reflected that falls link practitioners might feel ‘too scared’ to make decisions that involved significant changes such as bed moves or waiving an aspect of policy in the interests of patient safely ‘because that would be whoever’s in charge’ (Nurse-in-charge, Site 2), namely, the ward manager or nurse-in-charge. When observing events such as safety huddles and handovers, we did not see link practitioners taking a leading role. In addition, on one ward in which there had been an increase in falls, the falls link practitioner was not involved in a review to address the problem, and instead more senior nurses were involved.

Provision of training and support

Training and support for ward staff to implement falls prevention practices was provided in different ways. Most falls prevention training for new staff or refresher training for existing staff was not provided by link practitioners, but by organisational-level teams. For example, at Site 1, falls prevention training was provided as part of new staff induction by a central team and covered enhanced care, falls prevention, and medication. A senior nurse explained how recommendations from NAIF informed the development of their Trust’s falls training programme, which was mandatory for all new starters.

In comparison to Site 1, a ward manager at Site 3 pointed out that many experienced staff were assessors and supervisors for student nurses, providing on-the-job training themselves, and were therefore ‘very good at questioning why we do things’ (Ward manager, Site 3). At Site 2, compulsory training for new starters was also described, but the face-to-face version had been replaced by online training, as part of COVID-19 social distancing measures. Staff felt online training was less effective, especially for inexperienced staff or international staff new to UK health care. One ward had an influx of new staff who had not received face-to-face training and the ward manager believed this could have impacted delivery of falls prevention practices. Alongside formal training, a Site 2 falls specialist nurse described how informal discussion was used to support implementation of falls prevention practices; a lot of their work focused on EOC scores, ensuring they were accurate and prompting appropriate action, that is, that the right level of patient supervision was in place.

In terms of practical implementation of support at ward level, a key strategy discussed was messaging, for example, providing visual displays and reminders about expectations and policy on wards. A senior nurse at Site 1 explained that LSBP was included in their mandatory training programme in response to recommendations from NAIF, but added that ongoing communication and reminders to ward staff were needed to embed the practice in routine processes (and, in fact, we observed LSBP being taken at Site 1). However, a staff member at this site suggested an enthusiastic leader could, in fact, demotivate tired colleagues by encouraging them to do tasks they did not have time to do and what they needed, rather, was practical help. A ward manager emphasised this point: ‘I think, if you have a team that are frazzled, and the morale is low, the attitude isn’t there to want to engage and collaborate. People get a little bit, “Oh, this is pointless”’ (Ward manager, older person/complex care ward, Site 1). An organisational leader at Site 3 thought the best response was sensitivity to the demands staff faced, which meant taking incremental approaches to change. The need for sensitivity was echoed by a ward manager at Site 2, who prioritised staff well-being and offered practical support. In addition, a ward manager in Site 3 talked about the importance of ‘leading by example [ … ] So I do think my role is to hopefully teach a little bit, pass on my knowledge to the members of staff’ (Ward manager, orthopaedic ward, Site 3). However, a Site 2 matron discussed that, in the current NHS workforce, staff tended to move into leadership positions earlier than in the past and, therefore, did not have the same ‘exposure’ that supports decision-making and the ability to lead by example.

Consistent delivery of falls prevention practices

Record review indicated that falls risk assessments and care plans were documented quite consistently across sites (see Figures 4 and 5).

FIGURE 4.

Falls risk assessment documented on admission.

FIGURE 5.

Falls prevention care plan documented after first falls risk assessment.

There was variation in the number of falls per month and falls with injury (see Appendix 4, Figures 9–11). Although this was likely to be a natural variation rather than statistically significant, discussion with senior nurses at Site 1 provided some insight into why they believed falls and injury numbers varied from month to month. Falls incidents were attributed to not having staff to consistently provide one-to-one care to keep patients safe from falls. A ward manager in the Site 1 older person/complex care ward explained that their patient population fluctuated in terms of dependency, and patients could also be confused, agitated, delirious, or ‘wandersome’, which put them at risk of falls, so that ‘the biggest risk that we have when we’re delivering care is identifying those patients who require more intensive supervision’ (Ward Manager, older person/complex care ward, Site 1). Staff in other wards and sites described similar situations in terms of patient population and had attributed rises in falls incidents to similar issues.

Participants emphasised that staffing levels should consider not only patient numbers, but patient acuity. COVID-19 meant staff shortages were frequently observed during fieldwork, due to both staff sickness and staff being ‘pinched’ to work elsewhere in the hospital. While we did not capture data on staffing levels for all shifts observed, the data we captured suggested that the number of nurses and HCAs per shift averaged 80% of the numbers required for safe staffing, ranging from 50% to 100% for nurses and 40% to 100% for HCAs.

Some staff believed provision of resources was not solely a matter of prioritisation, but also a practical and logistical challenge. Pointing to the staffing crisis in the NHS, they explained that even where funds were made available by senior management to employ staff, it was often difficult to fill posts, especially for nurses and HCAs caring for older people, which were seen as particularly difficult jobs. The impact of experienced staff leaving the NHS was believed to add to this problem. However, a member of staff on the Site 3 older person/complex care ward talked about how, just as patients might have both modifiable and non-modifiable falls risk factors, so organisations had to work with factors that were modifiable and non-modifiable; while they might not be able to make more staff available, they could change the way staff were managed, and this could make a difference.

Theory refinement

Based on the data presented above, we were able to make refinements and additions to the leadership theory (Table 10), highlighting the importance of nursing leadership, the value of sensitive leadership, and the impact of resource constraints on falls prevention practices, in terms of staffing and training.

Informal communication between professional groups

Information was also communicated informally between professional groups. We most frequently observed communication between nurses and HCAs, but communication involving physiotherapists, doctors, and pharmacists was also observed. Examples of informal communication observed included staff checking with each other if jobs such as skin rounds and taking blood pressure had been completed, staff requesting help from each other to help move/mobilise or monitor patients, and providing updates to colleagues about patients’ condition, for example, if a patient was in pain and required medication. A Site 3 nurse explained that updates from physiotherapists and OTs were provided verbally. Nurses shared these updates with HCAs (who often had primary responsibility for aiding patient mobilisation) and made a note of the information to later update the EHR. Additionally, non-nursing staff would proactively ask for information. For example, a Site 3 nutritional assistant, who had worked for the organisation for over 10 years (as a HCA and then nutritional assistant), was familiar with the roles of different professional groups, so knew who to ask for the required information. Furthermore, the nutritional assistant supported patients in mobilising even though it was not their role specifically. To work in cohesive ways, such as this, it was important to know the roles and responsibilities of different professional groups, to understand who to ask for advice and refer issues to.

Availability of professional groups on wards facilitated informal communication about patient care needs. For example, a Site 3 pharmacist described being able to talk to nursing staff about patients because the nurses were always there. Wards were staffed in different ways, which influenced how and the extent to which informal communication took place. For example, the pharmacist described how, on the orthopaedic ward, fractured neck of femur patients had orthogeriatric input on a daily basis, providing an opportunity for the pharmacist to discuss with them medications that could increase their risk of falling. As another example, the Site 2 orthopaedic ward, with 28 beds, had a dedicated physiotherapist team, whose office was next to the nurses’ station. A physiotherapist in this team explained they were constantly discussing patients’ mobility needs with other ward staff. The Site 2 older person/complex care ward was not resourced in the same way as orthopaedics. During an observation of this ward, a physiotherapist discussed that they worked across three wards, and on the day of observation there was one physiotherapist and a physiotherapy assistant for 60 patients. A senior support worker on this ward emphasised there could be conflicting priorities between physiotherapists and the nursing team, for example, physiotherapists sometimes transferred several patients from bed to chair. While the support worker acknowledged this might contribute to patient recovery, it could also create challenges for nursing teams (especially when short staffed) who later needed to put the patients back into bed with assistance.

Consistency of staffing was also discussed as an influence on communication. For example, a consultant at Site 3 described bank nurses, who worked on wards on a temporary basis, for example, to cover staff shortages, but did not know the patients as well as permanent staff, which could impact on information communicated to other professional groups. A matron in Site 1 also highlighted the importance of consistent staffing for cohesive team working.

Visual communication tools

Visual tools were also used to communicate information about patients to ward staff. Patient bed boards, placed on the walls above patient beds, were used across wards and sites. In Site 2, a red falling leaf symbol was displayed on bed boards to identify patients at risk of falling, as well as being displayed on electronic whiteboards. In Site 3, magnets that visualised information such as mobility needs, skin needs, and falls risk were attached to bed boards. In terms of falls risk, the magnet displayed items such as ‘high risk’ and ‘low risk’, with boxes that staff could tick to indicate the patient’s condition. Bed boards were typically used to communicate patient transfer (from bed to chair) and mobilisation needs. Table 12 provides examples of notes taken about bed boards within one ward.

Table 12 demonstrates that information to support mobilisation differed by patient. During observations, across sites, information displayed on bed boards was often noted as variable, incomplete, or not up to date. Mobilisation information needed to be available at a glance for staff members to have the necessary information to support them to mobilise, especially those who did not know patients, such as bank staff. On the Site 1 older person/complex care ward, we observed a patient walking with the assistance of two student nurses to the toilet when the patient said they could not see because they did not have their glasses on; that the patient needed glasses when mobilising was not information that was typically shared, either verbally or visually.

Number and range of included studies

Figure 6 presents the results of the theory testing searches for all four prioritised CMOcs. In total, 28 manuscripts were used to test the facilitation CMOc. The studies were conducted in a range of countries including the USA,34–38,106–116 Canada,116,117 Spain,118 Australia,119–121 Singapore,122,123 Brazil,124 Japan,125 Taiwan,126 Belgium127 and the UK18 (see Appendix 6, Table 24 for study details). Studies tended to focus on nursing practice, although several studies considered the multidisciplinary team in delivery of interventions. 18,106,113

FIGURE 6.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses diagram for theory testing searches.

The studies encompassed a range of aims and objectives. Twelve studies were single-site QI projects that aimed to reduce falls incidence using interventions to promote adherence to existing falls prevention programmes or by developing, implementing and evaluating a novel falls prevention strategy. 106,107,109–112,114,117–119,124,125 Other studies shared similar aims to the QI projects, describing the implementation and impact of falls prevention programmes through a RCT, uncontrolled before and after survey, comparison of matched hospitals and survey. 18,113,122,127 The remaining studies examined HIT, exploring supports and constraints on implementation of HIT,36 the impact of the digitising the nursing process,37,108,115,126 auditing compliance with intervention protocol34,38,128 and clinicians’ attitudes towards using HIT for falls prevention. 120,121

The studies took place in a variety of settings: four in Emergency Departments (EDs),107,111,112,114 while others took place in acute medical and surgical units, older person evaluation units and acute wards, and intensive care units (ICUs). 18,34,35,109,110,113,117–124,127

Falls prevention tools

To understand the role of tools in facilitating implementation of falls prevention practices, we reviewed manuscripts to identify what tools were offered to staff; we grouped these as follows: (1) falls risk assessment tools; (2) care planning and intervention delivery tools; and (3) HIT tools.

Outcomes

Implementation of recommended practices was the key outcome of interest in the CMOc. Typically, implementation was assessed using quantitative measures, such as number of patients receiving falls risk assessments on admission pre- and postintervention, or measures of fidelity, for example, displaying a visual tool as required by an intervention protocol. In Appendix 7, Table 25, we summarise outcome data, where provided, on the following: (1) delivery of a falls risk assessment; (2) documentation of a falls prevention care plan; (3) use of interventions targeted to individual falls risks, and (4) falls incidents and rates.

How, why and in what contexts do MFRA and care planning tools work?

Although most studies described MFRA tools and pointed to their role in the facilitation process, there was a paucity of data about how and why staff responded to these tools. A further challenge in the analysis was that, typically, MFRA tools were not used in isolation. Often, multiple interventions were used to address implementation barriers as part of QI strategies, for example, staff incentives, education, training, and feedback were introduced alongside the assessment and care planning tools of interest. Therefore, we focused on tools as the resource of interest and examined what shaped clinician use and impact as key contextual influences of interest (including additional interventions used).

Analysis suggests the tools can facilitate delivery of recommended practices by (1) prompting staff to complete recommended tasks and actions and (2) automating parts of the falls prevention process. Contextual influences that help explain variation in tool impact include the following: (1) staff knowledge, experience, and attitude (sometimes targeted by educational and training interventions used as part of the study); (2) the complexity of ward work, including patient condition and movement between wards; (3) application of clinical judgment; and (4) physical resources needed to deliver interventions and multidisciplinary teamwork.

Theory refinement

Based on the literature presented above, we were able to make refinements and additions to the facilitation theory, moving from a single CMOc to four CMOcs concerned with facilitation (Table 15). The evidence supporting the claims of the CMOcs and GRADE-CERQual assessment of the evidence are presented in Appendix 8, Table 26.

Falls risk assessment and care planning tools

We explored whether the nine items recommended by NICE for a comprehensive MFRA were included explicitly in each site’s falls risk assessment, falls care plan or whether it could be identified in other nursing assessments (Table 16).

Table 16 indicates that, typically, the nine items recommended by NICE were included as a prompt in at least one space within the nursing documentation. However, they were not always explicitly incorporated in the falls risk assessment tool and there was variation between sites in phrasing of questions/prompts and tool content. Site 2, for example, included seven items recommended by NICE within its falls risk assessment tool and additional items, such as ‘Patient anxious about falling’.

At Site 3, completing the three-item assessment automatically calculated a score that categorised patients as high, medium or low risk. There was awareness that categorising patients in this way did not reflect the principles of NICE or the spirit of the Trust’s own falls prevention policy, which stated that people at risk of falling would be offered a MFRA and, where assessed at increased risk of falling, an individualised multifactorial intervention. However, this site was in the process of updating its EHR and the assessment documentation was being adapted to be more inclusive of individual falls risk factors and provide a better guide for staff in delivery of falls prevention practices.

While Site 3 was the only organisation using a score to stratify patients by falls risk, on all wards studied, staff explained that they attempted to identify patients considered at high risk of falls. Site 2 used the EOC scores to identify high-risk patients. High-risk patients were typically understood as needing some level of supervision to maintain safety, for example, one-to-one care, being placed on a cohort bay, intentional rounding.

Based on falls risk assessment and care planning documentation alone, it was not always clear the extent to which individual risk factors were comprehensively addressed for each patient at different sites. For example, in Site 1, the first care plan documented was called the ‘Intervention Plan’ and prompted staff to document actions against 15 items (Table 17).

A ‘Daily Care Plan’, an abbreviated version of the intervention plan, was used for updates after the Intervention Plan. Similarly, the care planning documentation in Site 3 prompted staff to enter responses for individual risks but was not as comprehensive as that provided in Site 1.

In Site 2, while the assessment included more NICE items than in Sites 1 and 3, the care plan did not prompt staff to enter responses for individual risks. A Falls Specialist Nurse explained that the documentation allowed nurses to link interventions to risks, but questioned the extent to which this happened in practice. Observation revealed that the ward manager on the orthopaedic ward had worked with the IT department to code falls interventions into the EHR to facilitate documentation. This meant that if they typed in ‘falls0’ or ‘falls1’, standard text relating to actions was inserted automatically; ‘falls0’ generated text about appropriate footwear, whereas ‘falls1’ generates text about call-bell use.

Examining the falls prevention documentation alone suggests why there is variation in how, and the extent to which, individual falls risk factors are identified and modified for patients across sites. However, nursing documentation in its entirety includes assessments that contribute to falls prevention even if they are not explicitly labelled on the EHR as such. For example, in Site 1, the falls risk assessment was part of the Nursing Specialist Assessment (NSA), completed for all patients on admission to a ward. Alongside the falls risk assessment, documentation included sections titled Cognition and Mental Capacity, Mobility, Elimination – Bladder, and Elimination – Bowel.

Use of tools in falls prevention

An aim of data collection was to observe falls risk assessments taking place in real time, to better understand how tools were used by nurses to identify and modify patient falls risk factors. During fieldwork researchers explicitly requested to shadow nurses while completing falls risk assessments. However, few assessments were observed in real time. Care plan updates were observed but often as a demonstration for the researchers rather than as part of normal work. Ward staff explained that patients are assessed when they are admitted to the ward and admissions can happen at any time throughout the day. Furthermore, care plans are typically updated during quiet times, often on the night shift, as data from the clinical record review indicated.

While laptops on wheels were observed on all wards, these were not used for completing falls documentation at the bedside. Assessing patients’ care needs was described as a dynamic process, not necessarily reliant on direct use of tools. For example, a Falls Specialist Nurse in Site 2 explained how a nurse might assess a patient’s condition and potential needs based on direct observation and conversation, identifying potential falls risks, for example, use of glasses and walking aids, for incorporation into the assessment. Similarly, a nurse in Site 3 described this type of dynamic assessment of patients, explaining that they ‘know’ quickly if a patient will take risks, presumably by direct observation and early interactions, and that the assessment is an ongoing process because patients’ conditions change. Observations supported interviewees’ accounts, showing how information was collated in a dynamic way and was not reliant on use of EHR tools, with the relevant sections of the EHR being completed at a later point, typically at the nurses’ station. Observations also revealed challenges in collating information about risks, for example, patients may not be able to communicate effectively if they are confused or in pain, an issue we return to in Chapter 8.

Theory refinement

Through the multisite case study, we were able to make a series of refinements and additions to the facilitation CMOcs (Table 18). Specifically, CMOc F1 becomes two CMOcs, reflecting what happens when tools are completed retrospectively; CMOc F2 is revised to reflect the dynamic nature of falls risk assessment and the potential for completion of documentation to become a tick-box exercise; and CMOc F5 is added to reflect the consequences of supervision being a key falls prevention strategy. We were unable to test automation of practice via HIT, as described in CMOc F4, because this technology was not available at sites, although our findings regarding information on bed boards being variable or incomplete reported in Chapter 6 does support the theory that updating such displays is a competing priority with other responsibilities.

Number and range of included studies

Twenty-four manuscripts were included in CMOc testing34,38,63,129–149 (see Appendix 9, Table 27 for study details). Ten studies34,38,63,129,130,134,135,138,139,142 examined interventions that sought to engage patients in the assessment and/or care planning process to encourage their participation in falls prevention interventions. Five studies examined Fall TIPS. 34,38,63,129,146 Studies were conducted in the USA,34,38,63,129,130,132,133,135,136,139–142,144–146,149 Australia,131,134,147,148 the UK,143 Sweden137 and New Zealand. 138 As with testing the facilitation CMO, it was found that most empirical literature related to patient participation focused on nursing practice. Several studies describe interventions to involve carers, as well as patients, in falls prevention. 38,130,134

Eight studies were QI projects,38,130,131,133,135,139,142 which aimed to reduce falls incidences in single sites with interventions that promoted adherence to existing or novel falls prevention programmes. Nine studies used quantitative methods,34,63,129,134,136,137,144–146 three of which evaluated falls prevention practices, using randomised and non-randomised control trial designs. 63,129,134 Seven studies used qualitative methods to explore patient and professional experiences of falls prevention practices132,140,141,143,147,148 and Martin et al. 138 evaluated an intervention that promoted patient participation in falls prevention using realist (mixed) methods.

All studies except for Martin et al. 138 took place in acute hospitals, in a range of clinical areas such as medical-surgical units130,136 and oncology units,38,133,137,139 among others. The study by Martin et al. 138 took place within an older person’s health rehabilitation hospital; we included this study, despite its non-acute setting, because of its relevance to this CMOc, given that many of its findings about tailoring communication for patients with specific characteristics are transferable to acute settings (and pertinent to the implementation of MFRA and tailored interventions). Few studies explicitly discussed working with patients with cognitive differences, such as dementia and delirium, or of patients whose ability to communicate with staff was constrained by language differences. 130,131,133,135,145 Only Haines et al. 134 explicitly included both groups. The Fall TIPS intervention explicitly included patients with different languages, and implicitly included patients with cognitive impairment, given that the intervention was offered to all adult inpatients in participating units. 34,38,63,129,146 Patients with cognitive impairment or who did not speak the language of the country in which the intervention took place were specifically excluded from two intervention studies. 139,142 Martin et al. 138 excluded patients with moderate to severe cognitive impairment [defined as Montreal Cognitive Assessment (MoCA)16 > 23/30] or delirium (screened using 4AT17 and then MoCA) but included mildly cognitively impaired patients and those who were hard of hearing, and specifically addressed their capacity to engage in falls assessment and prevention, thereby providing valuable evidence. Studies reporting patient perspectives excluded patients with severe and moderate cognitive impairment or who could not communicate in the language of the country in which the research took place,132,136,137,140,143,144 but two136,143 included patients with mild cognitive impairment.

Interventions to promote patient participation in falls prevention

In the CMOc, we think of resources as what healthcare professionals offered to patients through their interactions, specifically that they (1) provided an opportunity for patients’ needs and preferences to be incorporated in decision-making about their care and (2) communicated their individual care plan effectively to patients. The interventions varied in content and form; they can be conceived, broadly, on a continuum from staff-directed initiatives, with patients taking a somewhat passive role, to practices in which patients take a more active role as partners in the falls prevention team, illustrated in Figure 8. Drawing on the patient-centred framework described by Kullberg et al.,137 the continuum moves from one-way information transfer to more or less passive patients, to two-way information exchange, in which patients and healthcare professionals contribute in partnership. In the studies reviewed, we identified interventions that incorporated approaches along the continuum and for different purposes. Below we describe these interventions, beginning with those that most closely reflect the CMOc.

FIGURE 8.

Patient involvement continuum.

Outcomes

How, why and in what contexts does patient participation work?

From interview and focus group data collected as part of Martin et al. ’s138 realist evaluation, they elucidated four mechanisms underpinning why patients may participate in interventions:

• Gaining permission: where ward staff explained ward systems to patients and made it clear that they could call for help and assistance when needed.

• Increased awareness of risk: if staff made patients aware of the risk of falling via personalised conversations, patients were more likely to draw on relevant strategies to keep themselves safe.

• Empowerment to act: where goals were aligned with patient values, they were more likely to act in their own interests to keep themselves safe.

• Reminders: environmental cues prompted patients to ask for help and plan activities.

The findings from this study provide insight into how messaging works, for example, in discussion of increased awareness of risk, Martin et al. noted the importance of creating an ‘interactional space in which patients could integrate new knowledge into their existing beliefs’ (p. 266). The educators and volunteers involved in intervention delivery emphasised that creating such a space required time to establish rapport, to engage with and motivate patients, and personalise interactions. ‘Taking time’ was said to promote learning and changes in beliefs around falls risk and the authors reported that the goals developed in these conversations appeared to be most effective at creating behavioural change. A fifth mechanism was also reported by Martin et al.: the continuous prioritisation of falls prevention messaging by ward staff. This was where feedback from the educators and visual reminders, such as posters, reinforced the amount and consistency of messaging provided to patients by staff.

Martin et al. 138 reported that the volunteers in their study, as well as being retired nurses, were a similar age to some patients and, using their skills of listening, teaching and reflecting, created ‘engaging, personalised and safe interactional spaces’. Through this, they were able to ‘draw out’ patients to talk about some issues in more depth. Therefore, to provide the type of interaction that engages patients in conversations that lead to participation in falls prevention may be dependent on the skills of the assessor and their ability to create interactional rather than task-focused spaces. Radecki et al. 139 pointed to similar qualities. They explained that units that demonstrated improvement in falls rates had a higher percentage of experienced nurses (defined as more than 2 years’ experience) and discussed that this may have influenced the outcome (although there was a ceiling effect in one unit under study that had a baseline measure of no falls incidents). Radecki et al. 139 elaborated that nurses with less experience may be in ‘task mode’ meaning that they view the intervention as a task rather than a process through which communication and partnership with the patient is enhanced.

Bargmann and Brundrett130 reported a reduced falls rate following implementation of their falls prevention bundle and patient safety agreement. A key component of the intervention was education for professionals to support implementation of the intervention as intended. Compliance in using the intervention was measured and, after the initial audit, staff incentives (contributions to purchase snacks) were introduced to increase compliance towards a target of 90%. Following the incentive programme, adherence with bed alarms and the fall safety agreement rose to 89% after 3 months. The programme was said to have improved the culture of safety on the unit, which in turn supported an increase to 95% of patients correctly stating their fall risk.

Theory refinement

The mechanism in the original CMOc, taking patients’ needs and preferences into account in care plan development and communicating the prevention plan to the patient triggered their participation in the plan and reduced their risk of falling, in contexts where patients had capacity to engage in assessment and care planning. The studies provided greater detail about how and why this type of intervention may work and the contexts that influence impact.

Thinking about outcomes, in the first instance, the studies helped us build explanations in two areas: (1) how and why patients participate in falls prevention activities as an end-point; and (2) how patient participation may contribute to reducing incidents of falls as the overall aim of falls prevention strategies. For example, patients may participate in falls prevention by using call-bells (an impact in itself) in response to interventions, but if staff do not respond to their requests for help, they may attempt to mobilise on their own, particularly where there is a sense of urgency (e.g. for the toilet), creating a falls risk and potentially leading to falls.

In terms of mechanisms of patient participation, we focused on patient responses and what influences these responses. In summary, patients can experience emotional and physical barriers to participation. For example, they may be reluctant to use call-bells for fear of disturbing busy nurses or may not accept they are at risk of falling if they are not obviously limited physically. For these reasons, rather than staff simply considering needs and preferences in general, as stated in the original CMOc, the studies suggest that messages in the assessment, care planning, and intervention process need to be individualised to the emotional barriers experienced by patients to trigger mechanisms, such as gaining permission and acceptance of falls risk. We grouped these as empowering patients, for example, generating knowledge or confidence to participate in strategies to keep them safe. The idea that these types of interventions are more likely to be successful with patients without severe cognitive impairment was supported to a certain extent, with the question remaining of how can falls prevention strategies be delivered to patients with differing levels of cognitive impairment in a patient-centred way? A further influence on patient participation suggested in the studies was the type or quality of interaction between professional and patient, with risk assessment and care planning being seen not as a ‘task’ but as a vehicle to enhance communication, an attitude that may come with experience as staff become familiar with the process.

Finally, partnerships between professionals and patients/carers have the potential to impact on falls risks. Given sufficient resources, staff can be responsive to patients’ requests for help and can keep patients informed about any changes to their care plans or falls risks, ensuring that patients are not forced to engage in risky behaviour, thereby reducing their likelihood of falling.

Based on the data, Table 19 presents refinements to the initial CMOc, with the addition of new CMOcs. The evidence supporting the claims of the CMOcs and GRADE-CERQual assessment of the evidence are presented in Appendix 8, Table 26.

Interventions in which patients were encouraged to participate

Within the case sites, the interventions in which we observed patients participating included using the call-bell to ask for help when mobilising; wearing non-slip socks; wearing the correct glasses or hearing aids (if needed) and using mobility devices, such as walking sticks or frames; engaging in physiotherapy and rehabilitation; and taking time to get up/walking carefully (especially important for patients with postural hypotension). Other interventions, such as bed and chair alarms and sensors that alerted when a patient got up, did not require active patient participation.

Use of interventions differed across wards, depending on patient need and condition. For example, bed-bound patients recovering from surgery on orthopaedic wards might at first be unable to undertake physiotherapy and might rely heavily on using call-bells to obtain assistance from staff.

Interactions to encourage patient participation

Context mechanism outcome configuration P1 hypothesised that patient participation would be realised where ward staff created ‘interactional spaces’ that enabled falls prevention messages to be tailored to patient circumstances, addressing constraints to their participation. Although, as noted in Chapter 7, nurses’ capacity to create such spaces to share falls prevention messages was limited by the dynamic and implicit nature of falls risk assessments and care planning and by competing priorities on their time, staff emphasised the importance of ‘getting to know’ patients so that they could communicate and frame information in ways more likely to be accepted by and, perhaps, acted on by patients.

On the wards studied, large proportions of patients experienced some level of cognitive impairment, and some of these patients could struggle to understand or retain the meaning of direct or explicit messaging, such as instructions to wait for assistance before mobilising (the situation was complex, as cognitive impairments are experienced differently by different people at different times). However, staff told us that even acutely impaired patients might be able to understand the implicit dimensions of messages, such as whether a staff member really wanted to help them, and might respond to such messaging in ways that supported their safe care. In other words, the quality of the communication could constitute a form of messaging that triggered a degree of participation. For example, a cognitively impaired patient who could not understand they needed to wait for assistance before walking might be soothed when being supported to walk by a staff member they trusted, and potentially less likely to fall owing to agitated behaviour. We observed these kinds of interaction many times on the wards between patients, nurses, HCAs, and other staff, even though nurses were not always resourced to spend time with patients in the way the CMOc suggests.

Notwithstanding the importance of implicit messaging, many cognitively impaired patients were perceived by staff to be unable to participate fully in keeping themselves safe and thus required supervision to prevent them from falling. To support them in this, staff explained it was important to understand their behaviour. Wandering, for example, where patients became restless and walked the corridors, was also referred to as ‘walking with purpose’ because it tended to have underlying meaning for the person. If staff understood such behaviour, it could inform their care, for example, if they knew that a patient tended to be particularly active in the morning, they would provide supervision at that time, but perhaps not at other times. In other words, getting to know patients, understanding their perspectives and behaviours was recognised as a mechanism to inform messaging (where appropriate) and care delivery.

In comparison to messaging about patient-specific, tailored interventions, all patients in all sites were supposed to have access to call-bells to request assistance if needed. A nurse in Site 3 explained that patients were able to discern whether staff were sincere when they encouraged them to use call-bells and alert to conflicting messages, which might be conveyed implicitly through facial expression or tone of voice, reflecting the point made above about the impact of implicit messaging, this time for all patients and not only those with a cognitive impairment. Therefore, to generate patient assurance (or permission), this nurse emphasised to colleagues the need to convey an attitude of wanting to help to all patients on a shift-by-shift basis.

Nurses often explained they had limited time to get to know patients. For this reason, they sometimes relied on other staff to pass on information that could inform messaging and care delivery, such as HCAs. Some wards, namely the orthopaedic and older person/complex care wards at Site 3 and the orthopaedic ward at Site 2, employed staff with a specific remit for patient engagement, known as engagement support staff (Site 3) or activity co-ordinators (Site 2), who spent time with patients, chatting, reading to them, playing games, or encouraging them to engage in crafts. A nurse at Site 2 explained that the activity co-ordinator could encourage patients to participate in keeping themselves safe and could also pass information to the nursing team. Such information was shared either verbally or by completing documentation about patients’ individual needs (e.g. safe mobilisation needs). All three Trusts had such documentation, which was called ‘This is Me’ at Site 1, ‘Getting to Know Me’ at Site 2 and ‘See Who I Am’ at Site 3. Similar support from engagement support workers was observed at Site 3.

Physiotherapists and OTs also had time to interact with patients during therapy sessions and emphasised the need for two-way communication: ‘Making sure that [patients] understand what we’re doing and why and try and get someone on board with that and look at different ways around it’ (Physiotherapist, orthopaedic ward, Site 1). During fieldwork we observed physiotherapists attempting to encourage patients to participate in therapy (and thereby decrease their risk of falling), using different strategies if patients were not inclined to engage, for example, offering the patient a cup of tea and biscuit after the session. A pharmacist also highlighted the importance of ‘having patients [as] part of the conversation’, exploring medication options with them to find the best solutions to their needs and concluding: ‘I think it makes for a better, longer lasting position with patients and it gets them bought into what you’re trying to do’ (Pharmacist, orthopaedic ward, Site 3).

‘Knowing the patient is safe’

For patients with a cognitive impairment, CMOc P4 hypothesised that the nursing practice of knowing the patient is safe, rather than tailored messaging, would work to modify falls risk factors. The practice of knowing the patient is safe consisted of collating information via formal assessment, monitoring, and communication. These practices were enacted for all patients but, for some patients with cognitive impairments, could be done in a way that negated the need for their active participation in response to explicit messaging, for example, in relation to using the call-bell to ask for support.

Patients with cognitive impairment were often placed in cohort bays in Sites 1 and 2 or received one-to-one care. At Site 3, the layout of both wards made it difficult to establish a cohort bay as such, but staff told us they tried to group at-risk patients in more visible beds near to nurses’ stations, and sometimes patients were seated in the doorways to their rooms, with a tray-table in front of them, so that staff could monitor them more easily. Even where bays were conducive for cohorting, staff described challenges of using this strategy to know patients were safe. For example, other patients on a cohort bay might be at risk while the staff member on the bay was busy with one patient behind the curtains, and there were insufficient resources to provide more staff to monitor patients while this work was being done. Staff also talked about how monitoring patients on a cohort bay could be both stressful and boring, leading to staff ‘tuning out’ to a degree if they were there for a long time, and therefore becoming less vigilant. Similarly, they reported being tempted, sometimes, to leave the bay briefly to help colleagues, especially when patients appeared to be settled, or if they were asked directly to help others.

Intentional rounding (sometimes called comfort rounds or care and communication rounding), where staff checked on patients at regular intervals, was also used. A care worker on the Site 1 orthopaedic ward explained that during these 2- or 4-hourly rounds, they asked patients how they were, if they needed any help going to the toilet and checked call-bells were in reach, thereby helping them to participate in call-bell and toileting-related interventions. Staff talked about how a proactive approach can prevent patients from getting up without the required assistance, although the tension between preventing patients from falling and preventing deconditioning was acknowledged, as this quote from a Site 1 organisational-level interviewee describes:

One of the things we’ve done in the past year or so is to bring about a continence promotion team because we know that good management of continence will prevent falls. This is about some of this proactive questioning as part of rounding about do you need the toilet? Is there anything else that you need that I can get for you? So we’re pre-empting patient’s care needs before they almost even know that they need it. So you lessen their demand for needing to get up and about. Conversely, we should be encouraging our patients to get up and about and mobilise because we don’t want them to decondition, we don’t want them to if they’re a bit of a falls risk. It’s a bit of a dichotomy really in terms of what we can promote and how strong we can promote it without adversely affecting other harms.

The role of families and carers

During observations, the support provided by families and carers was disrupted due to COVID-19 because visiting was restricted and families and carers experienced difficulties contacting busy staff by telephone. However, data collection revealed that families, carers and visitors could play an important role in encouraging loved ones to participate and passing information to staff that could help them understand patients’ circumstances and behaviour. This was particularly important for those patients with cognitive impairments, who were unable to give staff such information themselves. A nurse in the Site 2 orthopaedic ward provided an example of speaking with family members about a patient with cognitive impairment who told the nurse the patient drank coffee every morning and not tea, enabling staff to understand why the patient was trying to get up (to find another drink). Speaking to family members to understand patient behaviour enabled ward staff to reduce the likelihood of patients taking actions that would increase their risk of falling, without the need for patient-directed messaging.

Additionally, carers could be a valuable source of support and information for patients with capacity, when, for example, they did not feel well enough to talk or when a personal or emotional issue made them reluctant to talk. For example, a carer in Site 1, whose mother experienced a fall in a toilet on the ward, thought that she had fallen while trying to remove her incontinence pants alone. The carer explained ‘my mum would never say to anybody: “I need help getting my incontinence pants off”’ through embarrassment, which suggests that, where staff are not able to create interactional spaces to enable patients to reveal sensitive issues, this may put them at risk due to embarrassment or other factors. In the above example, the carer also explained that her mother had difficulty understanding what staff were saying to her when they spoke with an accent, but she did not want to say this for fear of appearing critical. However, the carer said that, if asked, she would have been able to correct any misconceptions that had arisen between staff and her mother.

Carers also provided additional supervision for patients, supporting ward staff, watching out not just for their own loved ones, but also others on the same bay. For example, at a weekend observation on a falls cohort bay on the orthopaedic ward at Site 2, visitors were present for two of the three patients, all of whom had cognitive impairments. The visitors interacted with their loved ones in various ways: one visitor gave her relative a shave and helped him to eat, another showed her relative some photographs and joked with him. Put simply, alongside aiding staff in supervision, they took on jobs that would otherwise fall to staff, such as personal care. The visitors also interacted with the patient with no visitors, which he seemed to appreciate. The patient sat calmly in his chair while they were there, whistling and chatting, and did not attempt to mobilise unsafely.

In terms of the impact of restricted visiting during the COVID-19 pandemic, staff discussed the demotivating effect on patients of not having visitors, as well as the loss of information to themselves and added logistical challenges of contacting relatives and carers by telephone. They told us they missed this additional help, although some also said visitors caused extra work and felt that visiting restrictions had, in fact, taken some pressure off them. Towards the end of the observation period, however, visiting restrictions were eased.

Visual cues and reminders to patients

In addition to interactions to inform patient messaging and care delivery, the wards displayed a number of visual reminders of the main falls prevention messages to patients. For example, all wards had some variant of a ‘Call, Don’t Fall’ poster, displayed on bays, corridors, and in toilets and some had reminders about appropriate footwear. All sites had leaflets or videos with falls prevention advice for patients and their families, covering matters like using call-bells and appropriate walking aids, although patient awareness of their falls risks (see Patients as participants in falls prevention practices) raises questions about to what extent these were given to patients. We saw falls prevention display boards on wards, although it was not always clear to us whether these were directed at staff or patients (some may have been directed at both).

Visual aids were also used to remind patients who might struggle to understand written labels or instructions in English (e.g. icons to identify toilets or showers). Some wards used red-coloured Zimmer frames, which we were told were easier for patients with cognitive impairments to locate.

Patients as participants in falls prevention practices

To understand the extent to which patients participated in falls prevention interventions in response to staff interactions, we asked the patients we interviewed (all of whom had capacity) what they remembered about their falls risk assessment and falls prevention messages on admission to the ward. Across the three Trusts, 72% of patients/carers interviewed said they were not aware of having had a falls risk assessment, even though the record review of Trust EHR systems showed that such assessments were indeed taking place. Patient awareness of assessment varied between Trusts and wards, as did patient understanding of their own falls risks: some were very aware (often because they had fallen before or had a pre-existing condition that made falling more likely and had learnt through experience), while others appeared less well informed, making vague statements about feeling shaky or simply that they fell because they were ‘old’.

Patients frequently remembered being encouraged to use the call-bell and talked about seeing ward posters that reinforced this message to them. Some also remembered being encouraged to mobilise with appropriate footwear and walking aids, and recalled individualised messaging from physiotherapists, such as how to use a Zimmer frame, given their physical limitations or gait.

Patients talked about circumstances and perspectives likely to shape how they responded to, and acted on, the messaging directed at them by ward staff. They acknowledged receiving much information when they were admitted to hospital, which could be overwhelming, calling for clarity and conciseness, but on the other hand, they needed to feel respected: ‘You’ve got to be able to put it so that the person listening to you doesn’t think, “Oh for God’s sake. Obviously I know that”’ (Patient, orthopaedic ward, Site 2). Some also expressed the desire not to bother busy nurses.

Finally, we also saw patients helping each other to participate (e.g. telling staff a fellow patient needed help or comforting each other) or, sometimes, inhibiting participation by upsetting or agitating each other.

Patient experiences of participation

All patients were provided with, and advised to use, call-bells to request assistance when needed. CMOc P2 hypothesised that, where staff were not responsive to patients’ use of call-bells, they may take actions that increase their risk of falling, for example, trying to mobilise without the required assistance. Patients in all three sites described this experience and the wait time was especially distressing for those who needed the toilet urgently. A patient in Site 2 explained they did not want to take actions that increased their risk of falling, but sometimes felt they had to, to avoid making ‘a mess’. Another patient from Site 2 explained they had gone to the toilet by themself twice in the night before the interview, despite being aware that two people were supposed to accompany them, partly because they believed it would be difficult to obtain assistance from two busy staff members. Our observations captured similar experiences, for example, a bed-bound patient was observed repeatedly pressing the call-bell for help to go to the toilet on the Site 1 orthopaedic ward. This patient could not get out of bed unassisted, so was unable to risk mobilising on their own, but during the same observation another more mobile patient, who had earlier been assisted to the toilet by staff, took that risk. On this occasion, there was only a few moments’ delay between pressing the call-bell and a staff member arriving to help, but the patient had already rushed to the toilet alone, which illustrates the level of responsiveness sometimes needed to avoid patients taking actions that increase their risk of falling. Delayed staff responses to patients were observed on cohort bays also, when the staff member on the bay was busy with one patient at the same time as another needed help. Therefore, patients receive messaging about using call-bells (although nursing attitude in delivery may vary) but a key constraint on acting on this messaging is staff response time. Several patients commented that they thought staff ability to respond quickly to call-bells or to support other forms of participation was related to workload and short-staffing. Certainly, we saw occasions where a single HCA was trying to respond to the requests of several different patients on a bay, resulting in some patients having to wait.

However, when staff did have the time necessary to support patients, the impact of this was observed. For example, on the Site 1 older person/complex care ward, we observed a HCA supporting a patient to walk down the corridor with a Zimmer frame. The support worker asked the patient what they needed (whether they would like a wheelchair) and responded in a way that was attuned to their needs, both physically (i.e. walking behind the patient so they could support them if they started to fall) and emotionally (reassuring them that they had time). Similar examples of staff interacting in ways that encouraged patients to accept help were also captured. For example, a patient at Site 1, who had been ashamed to experience diarrhoea while bed-bound, described the ‘kindness’ and ‘consideration’ they experienced from the nurses, which helped them to feel more acceptable, which, in turn, involved accepting help.

Patients with cognitive impairment

Context mechanism outcome configuration P4 hypothesised that, where patients had difficulty understanding or retaining information or were unable to communicate their needs to staff, staff needed to perform practices not reliant on patient participation to help prevent falls. Patients with cognitive impairment could transition between periods of calm to agitation quickly and behaviours such as ‘walking with purpose’, or trying to leave the ward, were also observed and falls incidents were attributed to this. Staff told us that their behaviour could be affected by other contextual factors, too, such as length of stay, the extent to which they were able to spend time with familiar people, and the number of times they moved wards or beds.

Even where wards were staffed to provide supervision to all patients in need, other skills were needed to help calm cognitively impaired patients in moments of distress or agitation, such as speaking calmly, responding to them at their own level of understanding, and not contradicting them. For example, on the Site 2 orthopaedic ward, we observed a HCA talking to a patient who had stood up abruptly, saying they wanted to go home. Rather than telling the patient they could not go home, the HCA negotiated a better time to go there, repeating the conversation as needed. As a result, the patient was able to walk safely when they wanted but also accepted support from the staff members present. The need to remind patients, perhaps repeatedly, without becoming bored or irritated was observed as an important trait in caring for cognitively impaired patients. These findings reflect the importance of targeted, implicit messaging outlined in Interactions to encourage patient participation.

Theory refinement

The data collected in staff, patient and carer interviews and ethnographic observations enabled further refinements to be made to the patient participation CMOcs, outlined in Table 20. Specifically, CMOcs P1 and P2 have been refined to capture the specific contextual factors we observed – multiple patients requiring assistance and staffing levels; a strategy we observed nursing staff using, namely advising patients to use the call-bell; and the consequences of this. CMOc P3 is revised to show how sensitive communication can support patients to avoid taking risks of the kind described in the literature-based version of this CMOc. CMOc P4 is revised to reflect the strategy we observed of receiving information about patients from other staff, family and carers, supplementing assessment and monitoring. It becomes two CMOcs, reflecting the fact that this was a strategy used with both patients with cognitive impairments and patients without. CMOc P5 is added to capture the strategy of normalising patients’ experiences, generating more willingness to accept help.

Using secondary data to develop a programme theory

Through our realist review, we developed an initial programme theory that explains what supports and constrains routine use of MFRA and falls prevention interventions, as presented in Chapter 4. Within the study, we were able to use secondary data to test two of the CMOcs that made up that programme theory, concerned with facilitation of MFRA and patient participation.

Analysis suggested that if tools are visible in staff workflow, they may facilitate implementation of MFRAs and tailored interventions by prompting completion of required tasks. 18,110,111,118,122 The literature also pointed to the complexity of care delivery, for example, in terms of patient acuity and limited resources, that may disrupt use of tool guidance and documentation of care processes. Automating practices via HIT appears a promising implementation support34,38,115 but may introduce additional training requirements to support technology use and additional manual tasks that may be seen as a competing priority on staff time. 109,120

The literature suggested patients’ circumstances should inform the messaging directed to them to enable them to participate in falls prevention interventions. 138,146–148 The quality of interaction between patients and professionals appeared to underpin successful messaging,138,139 but creating such interactions relied on staff experience, skills, and time, resources that may be constrained outside the context of an intervention study. 138,139 Furthermore, some patients are not able to remember or understand messaging due to cognitive impairment, with one study indicating potential harm to these patients. 134 Therefore, other strategies are needed to support falls prevention in these populations. The review also revealed a lack of research involving those with cognitive impairment or who do not speak English.

Theory development was achieved via iterative literature searches, building on practitioner ideas with evidence from empirical studies, and allowed for inclusion of different types of data. Including different methodologies can be considered a study strength, providing examples from clinical practice, but proved challenging to synthesise because there was much variation, for example, in outcomes assessed, description of methods, and data reported. Use of GRADE-CERQual to assess the strength of evidence associated with the CMOcs can also be considered a review strength; it supported team discussion about key findings, often leading to further CMOc refinements. We would recommend use of GRADE-CERQual in future realist reviews.

Refining the programme theory through a multisite case study

In our multisite case study, we used multiple methods of data collection – ethnographic observations, interviews with staff, patients and carers, and review of patient records – to test the four prioritised CMOcs from our initial programme theory. This combination of methods provides rich description and fresh insights regarding the realities of MFRA and prevention in acute hospitals in England, addressing a previous gap in the literature and strengthening the knowledge base.

Developing guidance

When we presented our findings to case sites, participants agreed with our analysis. In discussing their implications, two key themes emerged: (1) the need for lessons learnt to be disseminated to all professional groups through leaflets and training materials; and (2) the need for leaflets for patients and carers, individualised to patients, providing them with information about their falls risks and how to reduce their risk of falling. Participants said stories and vignettes from our fieldnotes would be a good way to disseminate information because they are ‘real-life’ examples and some sites asked for permission to use our materials within their own falls prevention training.

In response, we produced a draft guidance leaflet based on our findings (see Implications for practice for implications of our findings for practice), which includes one vignette. We produced a set of PowerPoint slides that provide additional vignettes and discussion prompts that can be accessed by individuals or incorporated into training. We sought feedback on our draft guidance leaflet via a NAIF webinar to falls leads (June 2023) and a symposium at the Royal College of Nursing International Research Conference (September 2023). We have revised the guidance leaflet considering the feedback.

Providing individualised leaflets for patients and carers could be high-tech or low-tech; the Fall TIPS software generates leaflets for patients,63 but an alternative is a leaflet where the relevant options can be ticked. We have produced a draft paper-based patient leaflet that allows for this. Our patient leaflet has been revised based on feedback from our Lay Research Group. We will further develop it through a workshop with a wider group of service users and carers, including people living with dementia, and intend to incorporate appropriate icons or images, as well as text. Using existing patient leaflets for falls prevention as examples, such as those developed by the Royal College of Physicians,169 key questions to explore are as follows: (1) Given patients’ request for conciseness, what additional information should be included? For example, should more information be given about actions that multidisciplinary teams can take (e.g. reviewing medications)? (2) Which icons are most easily understandable by a wide range of audiences? (3) What else is needed to ensure the leaflet is understandable, useful and acceptable?

Once these activities are complete, we will send the guidance leaflet and patient leaflet for professional typesetting and layout design before dissemination. In addition to sharing the materials with our case sites and via social media, the Royal College of Nursing Professional Lead for Older People and Dementia will work with us to seek Royal College of Nursing endorsement of the materials and disseminate them more widely. We have been invited to write a practically focused article for the practitioner journal Nursing Older People and will use this to promote our guidance materials, providing links to the materials on our website. We have also agreed with the British Geriatrics Society that, after their conference, we will write a blog for the British Geriatrics Society website, which will also provide links to the materials. We are in conversation, too, with NAIF about how they can support wider dissemination.

Leadership

Our findings emphasise the importance of experienced nursing leaders, the need for role clarity and the authority to target resources. For role clarity and to avoid duplication of effort, we suggest that managers should make sure there is a clear distinction and communication between ward manager/nurse-in-charge roles and falls link practitioners about falls prevention. If falls link practitioners are to fulfil their roles, adequate time to do so needs to be allocated and protected, including time for training.

Shared responsibility

There is a need to recognise that MFRA and prevention should not be, and cannot be, the responsibility of nursing staff alone; a comprehensive MFRA requires the input of other professional groups. Greater multidisciplinary team, and patient and carer, involvement in MFRA and selection and delivery of interventions is required. This message should be emphasised by Board members and managers. Additionally, we recommend the following actions:

• Managers should ensure the ethos of the NICE guideline is understood by ward teams, that is, it is not just about documentation of MFRAs but multidisciplinary and patient and carer input to ensure modifiable risk factors are addressed.

• Managers, front-line healthcare professionals, and support workers should work together to consider how systems and processes can be revised to better support multidisciplinary approaches to falls risk assessment and prevention.

• Staff who lead and organise safety huddles should involve colleagues from different disciplines, clinical and non-clinical, wherever possible in huddles, for example, by holding them at times they can attend (normally between 9.00 and 17.00).

• Managers should ensure that there are processes in place to update bed boards regularly and accurately, to ensure that all members of the multidisciplinary team have access to accurate information about the assistance a patient needs to mobilise.

• Managers and IT teams should consider how EHRs can better support multidisciplinary working. This could include, for example, ensuring the falls risk assessment is fully integrated with and easily accessible by clinicians, such as doctors and allied health professionals and automatically triggering a request for a medication review when this is identified as a risk factor for a patient.

• Managers, front-line healthcare professionals, and support workers should reflect together on how decisions about which patients require supervision are made and consider what alternatives there may be to supervision.

Facilitation

Documentation burden, combined with constraints on nursing time, potentially reduce the quality of falls documentation, while also requiring nurses to make a choice between ‘must do’ work and ‘being with’ patients. 119 While nurse staffing issues contribute to this, the current NHS workforce crisis combined with resource constraints mean we cannot address the problem by simply recruiting more nurses. Therefore, managers and IT teams should work with nursing staff to identify ways to reduce documentation burden. This could include removing duplication of information and autopopulating items where the information is already available in EHRs. Our findings related to facilitation also suggest the following:

• Managers and IT teams should work together to ensure that items included in falls risk assessment tools align with NICE guidance. Thought should be given to ways in which items are phrased and how information is entered to avoid tick-box responses.

• Board members and managers should emphasise to healthcare professionals that falls risk assessment tools and falls care plans are tools to support practice. This means moving away from a culture where documentation is seen as something to be audited and blame assigned when not completed to a culture of learning and continuous development.

Patient participation

Trusts should consider how they can ensure patients receive individualised information about their falls risks and how they can reduce their risk of falling. We have described how patient leaflets could be individualised but, whatever the solution, for successful implementation, it must not create an additional burden for nursing staff.

It is also important to recognise that providing patients with a leaflet is unlikely to be enough, with our findings emphasising the importance of interactions between staff and patients; it is interesting to note changes to the patient participation CMOcs following our observations, emphasising that, in the absence of formal interventions, patience and kindness may reduce behaviours that can lead to falls, a point that we have made clear in our guidance leaflet for staff. Trusts should consider providing staff with guidance on brief but sensitive ways to talk with patients about risks and preventing falls. Our research highlighted the role of engagement support workers in supporting people with dementia, to reduce their risk of falling; while there is some evidence of benefits of such roles,172 further research is required to evaluate their potential to support falls prevention.

Leadership

CINAHL (EBSCOhost)

Search date: 6 May 2021

Records found: 122

EMBASE Classic + EMBASE (Ovid) 1947 to 5 May 2021

Search date: 6 May 2021

Records found: 187

----------- ------------ -------------- ----------- ----------- -------- -------------

1. hospitalization/ (420,992)

2. hospital department/or exp ward/ (439,715)

3. exp hospital/ (1,302,934)

4. medical staff/ (40,310)

5. nursing staff/ (74,427)

6. rehabilitation center/ (17,387)

7. subacute care/ (1151)

8. exp clinical handover/ or collaborative care team/ or exp rapid response team/ (11,414)

9. exp hospital patient/ (199,241)

10. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kw. (51,102)

11. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kw. (12,324)

12. inpatient?.tw,kw. (199,814)

13. (acute adj5 patient?).tw,kw. (369,927)

14. hospital*.tw,kw. (2,200,719)

15. or/1-14 [hospital] (3,282,770)

16. *falling/ (12,578)

17. falling/pc (2954)

18. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kw. (25,940)

19. or/16-18 [Falls prevention or risk assessment] (32,853)

20. implement*.tw,kw. (724,768)

21. (sustain* adj4 (program* or practice* or intervention?)).tw,kw. (14,660)

22. Adopt*.tw,kw./freq=2 (49,567)

23. Adopt*.ti. (20,665)

24. deliver*.tw,kw./freq=2 (374,349)

25. deliver*.ti. (176,933)

26. (adher* or comply or complian*).tw,kw. (526,988)

27. fidelity.tw,kw. (35,433)

28. implementation science/ (1982)

29. exp protocol compliance/ (15,723)

30. engage*.ti. (20,887)

31. engage*.tw,kw./freq=2 (52,149)

32. (intervention adj2 deliver*).tw,kw. (7304)

33. (Cognitive adj2 participat*).tw,kw. (729)

34. “buy in”.tw,kw. (2788)

35. or/20-34 [Implementation or Adherence to Guidelines and strategies] (1,730,025)

36. leadership/ (76,777)

37. total quality management/ (72,191)

38. leader*.tw,kw. (102,125)

39. champion*.tw,kw. (10,691)

40. (organi#ation* adj3 (support* or strateg*)).tw,kw. (12,500)

41. exp *“organization and management”/ (577,268)

42. facilitator*.ti. (6305)

43. facilitat*.tw,kw./freq=2 (94,889)

44. ((quality or safety or nurse* or matron*) adj3 lead*).tw,kw. (13,991)

45. or/36-44 [Leaderships] (841,009)

46. risk assessment/ (618,392)

47. (risk adj2 (assess* or evaluat*)).tw,kw. (196,115)

48. exp practice guideline/ (601,575)

49. guideline*.tw,kw. (613,857)

50. exp clinical protocol/ (106,751)

51. protocol?.tw,kw./freq=2 (185,559)

52. protocol?.ti. (86,947)

53. or/46-52 [Multifactorial risk assessment] (1,767,880)

54. 15 and 19 and 35 and 45 and 53 [Leadership CMO1 – final] (187)

Ovid MEDLINE(R) ALL 1946 to 5 May 2021

Search date: 6 May 2021

Records found: 145

1. Hospitalization/ (115,592)

2. Subacute Care/ (1169)

3. Hospital Units/ (10,319)

4. exp Hospitals/ (284,007)

5. medical staff, hospital/ or nursing staff, hospital/ (66,506)

6. exp Patient Care Team/ (70,042)

7. Rehabilitation Centers/ (8356)

8. Inpatients/ (23,637)

9. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kf. (34,662)

10. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kf. (6803)

11. (acute adj5 patient?).tw,kf. (227,174)

12. inpatient?.tw,kf. (116,838)

13. hospital*.tw,kf. (1,383,898)

14. or/1-13 [hospital] (1,835,443)

15. Accidental Falls/pc [Prevention and Control] (9549)

16. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kf. (17,142)

17. 15 or 16 [Falls] (21,139)

18. implement*.tw,kf. (545,160)

19. (sustain* adj4 (program* or practice* or intervention?)).tw,kf. (11,220)

20. Adopt*.tw,kf./freq=2 (37,248)

21. Adopt*.ti. (16,394)

22. deliver*.tw,kf./freq=2 (258,583)

23. deliver*.ti. (137,239)

24. (adher* or comply or complian*).tw,kf. (343,916)

25. fidelity.tw,kf. (30,445)

26. Health Plan Implementation/ (6379)

27. implementation science/ or technology transfer/ (2784)

28. Guideline Adherence/ (33,374)

29. engage*.tw,kf./freq=2 (39,474)

30. engage*.ti. (17,196)

31. (intervention adj2 deliver*).tw,kf. (5693)

32. (Cognitive adj2 participat*).tw,kf. (547)

33. “buy in”.tw,kf. (1792)

34. or/18-33 [Implementation or Adherence to Guidelines and strategies] (1,263,268)

35. Leadership/ (42,613)

36. Total Quality Management/ (12,588)

37. exp *“organization and administration”/ (729,180)

38. leader*.tw,kf. (83,493)

39. champion*.tw,kf. (7722)

40. (organi#ation* adj3 (support* or strateg*)).tw,kf. (10,007)

41. facilitator*.ti. (5290)

42. facilitat*.tw,kf./freq=2 (72,113)

43. ((quality or safety or nurse* or matron*) adj3 lead*).tw,kf. (10,657)

44. 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 [Leadership] (899,951)

45. exp Risk Assessment/ (284,216)

46. (risk adj2 (assess* or evaluat*)).tw,kf. (137,873)

47. guideline/ or practice guideline/ (35,660)

48. guideline*.tw,kf. (386,469)

49. protocol?.tw,kf./freq=2 (126,031)

50. protocol?.ti. (69,596)

51. exp Clinical Protocols/ (174,448)

52. or/45-51 [Multifactorial risk assessment] (1,055,597)

53. 14 and 17 and 34 and 44 and 52 [Leadership CMO1 final] (145)

NICE Evidence

Search date: 7 May 2021

Records found: 53

(“falls prevention” or “falls assessment” or “falls reduction”) and (hospital or inpatient or “acute care” or “acute ward” or “acute patient” or “rehabilitation unit” or “rehabilitation ward”) and (leadership or champions) and (“risk assessment” or “risk evaluation” or “evaluation of risk”)

Sorted by Relevance.| Sort by Date

Filters applied:

Evidence type: Evidence Summaries

Evidence type: Primary Research

Evidence type: Systematic Reviews

Evidence type: Audit and Inspection Reports

Evidence type: Health Technology Assessments

Web of Science Core Collection Databases available at the University of Leeds (searched simultaneously):

Arts and Humanities Citation Index (Web of Science) 1975–present

Conference Proceedings Citation Index- Science (Web of Science) 1990–present

Conference Proceedings Citation Index- Social Science and Humanities (Web of Science) 1990–present

Science Citation Index-Expanded (Web of Science) 1900–present

Social Sciences Citation Index (Web of Science) 1900–present

Emerging Sources Citation Index (Web of Science) 2015–present

Search date: 6 May 2021

Records found: 38

Indexes=SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI Timespan=1900–2021

Facilitation

CINAHL (EBSCOhost)

Search date: 6 May 2021

Records found: 32

EMBASE Classic + EMBASE (Ovid) 1947 to 5 May 2021

Search date: 6 May 2021

Records found: 45

EMBASE Classic + EMBASE <1947 to 5 May 2021>

Search strategy:

---------- ------------ ------------ ---------- -------------- ----------- -----------

1. hospitalization/ (420,992)

2. hospital department/ or exp ward/ (439,715)

3. exp hospital/ (1,302,934)

4. medical staff/ (40,310)

5. nursing staff/ (74,427)

6. rehabilitation center/ (17,387)

7. subacute care/ (1151)

8. exp clinical handover/ or collaborative care team/ or exp rapid response team/ (11,414)

9. exp hospital patient/ (199,241)

10. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kw. (51,102)

11. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kw. (12,324)

12. inpatient?.tw,kw. (199,814)

13. (acute adj5 patient?).tw,kw. (369,927)

14. hospital*.tw,kw. (2,200,719)

15. or/1-14 [hospital] (3,282,770)

16. *falling/ (12,578)

17. falling/pc (2954)

18. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kw. (25,940)

19. or/16-18 [Falls prevention or risk assessment] (32,853)

20. implement*.tw,kw. (724,768)

21. (sustain* adj4 (program* or practice* or intervention?)).tw,kw. (14,660)

22. Adopt*.tw,kw./freq=2 (49,567)

23. Adopt*.ti. (20,665)

24. deliver*.tw,kw./freq=2 (374,349)

25. (adher* or comply or complian*).tw,kw. (526,988)

26. fidelity.tw,kw. (35,433)

27. implementation science/ (1982)

28. exp protocol compliance/ (15,723)

29. engagement.tw,kw. (97,441)

30. engage*.tw,kw./freq=2 (52,149)

31. (intervention adj2 deliver*).tw,kw. (7304)

32. (Cognitive adj2 participat*).tw,kw. (729)

33. “buy in”.tw,kw. (2788)

34. or/20-33 [Engagement or Implementation] (1,725,436)

35. clinical assessment tool/ (25,003)

36. (assess* adj4 tool?).tw,kw. (94,602)

37. (electronic adj2 record?).tw,kw. (83,628)

38. *fall risk assessment/ (605)

39. medical informatics/or nursing informatics/ (23,031)

40. exp hospital information system/ (25,135)

41. exp information technology device/ (201,010)

42. (information adj3 (technolog* or system or computeri* or electronic)).tw,kw. (70,350)

43. digital*.tw,kw. (211,640)

44. (“risk assess*” adj4 (bundle* or instrument* or care plan*)).tw,kw. (604)

45. ((multi-factorial or multifactorial) adj4 risk assessment*).tw,kw. (130)

46. (risk* adj3 screen* adj3 tool*).tw,kw. (1181)

47. clinical decision support system/ (3800)

48. clinic* decision* support system*.tw,kw. (3087)

49. health information management.tw,kw. (909)

50. medical information system/or bedside information system/ (22,022)

51. (electronic* adj3 nurs* adj3 document*).tw,kw. (137)

52. or/35-51 [Assessment tools including health info technology] (692,085)

53. workflow/ (26,854)

54. integration/ (6459)

55. data integration/ (544)

56. facilitation.tw,kw. (39,629)

57. (workflow* or work-flow*).tw,kw. (43,560)

58. system? fit*.tw,kw. (459)

59. Embed*.tw,kw. (176,919)

60. integrat*.tw,kw. (705,509)

61. (routine* or routini*).tw,kw. (628,385)

62. or/53-61 [Workflows concept] (1,543,566)

63. 15 and 19 and 34 and 52 and 62 [HIT tools CMO4 – final] (45)

Ovid MEDLINE(R) ALL 1946 to 5 May 2021

Search date: 6 May 2021

Records found: 30

Ovid MEDLINE(R) ALL <1946 to 5 May 2021>

Search strategy:

--------- ----------- ----------- ---------- ------------ ------------- --------------

1. Hospitalization/ (115,592)

2. Subacute Care/ (1169)

3. Hospital Units/ (10,319)

4. exp Hospitals/ (284,007)

5. medical staff, hospital/or nursing staff, hospital/ (66,506)

6. exp Patient Care Team/ (70,042)

7. Rehabilitation Centers/ (8356)

8. Inpatients/ (23,637)

9. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kf. (34,662)

10. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kf. (6803)

11. (acute adj5 patient?).tw,kf. (227,174)

12. inpatient?.tw,kf. (116,838)

13. hospital*.tw,kf. (1,383,898)

14. or/1-13 [hospital] (1,835,443)

15. Accidental Falls/pc [Prevention and Control] (9549)

16. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kf. (17,142)

17. 15 or 16 [Falls] (21,139)

18. implement*.tw,kf. (545,160)

19. (sustain* adj4 (program* or practice* or intervention?)).tw,kf. (11,220)

20. Adopt*.tw,kf./freq=2 (37,248)

21. Adopt*.ti. (16,394)

22. deliver*.tw,kf./freq=2 (258,583)

23. deliver*.ti. (137,239)

24. (adher* or comply or complian*).tw,kf. (343,916)

25. fidelity.tw,kf. (30,445)

26. Health Plan Implementation/ (6379)

27. implementation science/or technology transfer/ (2784)

28. Guideline Adherence/ (33,374)

29. engage*.tw,kf./freq=2 (39,474)

30. engage*.ti. (17,196)

31. (intervention adj2 deliver*).tw,kf. (5693)

32. (Cognitive adj2 participat*).tw,kf. (547)

33. “buy in”.tw,kf. (1792)

34. or/18-33 [Implementation or Adherence to Guidelines and strategies] (1,263,268)

35. Risk Assessment/mt [Methods] (34,919)

36. (assess* adj4 tool?).tw,kf. (64,154)

37. (electronic adj2 record?).tw,kf. (43,674)

38. medical informatics/or health information exchange/or exp medical informatics applications/or exp medical informatics computing/or nursing informatics/ (460,638)

39. exp Hospital Information Systems/ (28,271)

40. exp health information management/ (1788)

41. exp Medical Records Systems, Computerized/ (42,909)

42. (information adj3 (technolog* or system or computeri* or electronic)).tw,kf. (52,606)

43. digital*.tw,kf. (157,475)

44. (“risk assess*” adj4 (bundle* or instrument* or care plan*)).tw,kf. (487)

45. clinic* decision* support system*.tw,kf. (2415)

46. (electronic* adj3 nurs* adj3 document*).tw,kf. (118)

47. Decision Support Systems, Clinical/ (8422)

48. or/35-47 [Assessment tools including health IT] (787,014)

49. workflow/ (6346)

50. systems integration/ (9571)

51. facilitation.tw,kf. (30,997)

52. (workflow* or work-flow*).tw,kf. (29,615)

53. system? fit*.tw,kf. (333)

54. Embed*.tw,kf. (138,578)

55. integrat*.tw,kf. (568,134)

56. (routine* or routini*).tw,kf. (405,465)

57. or/49-56 [Workflows] (1,141,149)

58. 14 and 17 and 34 and 48 and 57 [Facilitation Tools CMO 4] (30)

NICE Evidence

Search date: 7 May 2021

Records found: 43

(“falls prevention” or “falls assessment” or “falls reduction”) and (hospital or inpatient or “acute care” or “acute ward” or “acute patient” or “rehabilitation unit” or “rehabilitation ward”) and (“clinical decision support system” or “assessment tool”) and (workflow or routine or “system fit”)

Sorted by Relevance.| Sort by Date

Filters applied:

Evidence type: Evidence Summaries

Evidence type: Primary Research

Evidence type: Systematic Reviews

Evidence type: Audit and Inspection Reports

Evidence type: Health Technology Assessments

Web of Science Core Collection

Databases available at the University of Leeds (searched simultaneously): Arts and Humanities Citation Index (Web of Science) 1975–present

• Conference Proceedings Citation Index – Science (Web of Science) 1990–present;

• Conference Proceedings Citation Index – Social Science and Humanities (Web of Science) 1990–present;

• Science Citation Index-Expanded (Web of Science) 1900–present;

• Social Sciences Citation Index (Web of Science) 1900–present;

• Emerging Sources Citation Index (Web of Science) 2015–present.

Search date: 6 May 2021

Records found: 25

Indexes=SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI Timespan=1900–2021

Shared responsibility

CINAHL (EBSCOhost)

Search date: 6 May 2021

Records found: 42

EMBASE Classic + EMBASE (Ovid) 1947 to 2021 May 05

Search date: 6 May 2021

Records found: 42

EMBASE Classic + EMBASE <1947 to 5 May 2021>

Search strategy:

-------- ---------- ------------- ----------- ----------- ------------ ---------------

1. hospitalization/ (420,992)

2. hospital department/ or exp ward/ (439,715)

3. exp hospital/ (1,302,934)

4. medical staff/ (40,310)

5. nursing staff/ (74,427)

6. rehabilitation center/ (17,387)

7. subacute care/ (1151)

8. exp clinical handover/ or collaborative care team/ or exp rapid response team/ (11,414)

9. exp hospital patient/ (199,241)

10. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kw. (51,102)

11. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kw. (12,324)

12. inpatient?.tw,kw. (199,814)

13. (acute adj5 patient?).tw,kw. (369,927)

14. hospital*.tw,kw. (2,200,719)

15. or/1-14 [hospital] (3,282,770)

16. *falling/ (12,578)

17. falling/pc (2954)

18. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kw. (25,940)

19. or/16-18 [Falls prevention or risk assessment] (32,853)

20. risk assessment/ (618,392)

21. (risk adj2 (assess* or evaluat*)).tw,kw. (196,115)

22. exp practice guideline/ (601,575)

23. guideline*.tw,kw. (613,857)

24. exp clinical protocol/ (106,751)

25. protocol?.tw,kw./freq=2 (185,559)

26. or/20-25 [Multifactorial risk assessment] (1,743,357)

27. multidisciplinary team/ or collaborative care team/ (12,645)

28. collaborat*.tw,kw./freq=2 (50,596)

29. collaborat*.ti. (40,601)

30. (team* adj3 (share* or sharing or communic*)).tw,kw. (7724)

31. (team* adj3 (multidisciplin* or interdisciplin* or multi-disciplin* or inter-disciplin* or interprofessional* or inter-professional*)).tw,kw. (61,882)

32. cooperation/ or teamwork/ (64,251)

33. public relations/ (62,742)

34. (team* adj5 huddle?).tw,kw. (337)

35. (team* adj5 handover?).tw,kw. (230)

36. (teamwork* or team-work*).tw,kw. (20,866)

37. or/27-36 [Teamwork] (247,513)

38. ((share* or sharing) adj3 (responsibility or responsible or accountability or accountable)).tw,kw. (3181)

39. (joint adj3 (responsibility or responsible or accountability or accountable)).tw,kw. (586)

40. ownership.tw,kw. (17,297)

41. communicat*.tw,kw. (442,253)

42. exp interpersonal communication/ (717,292)

43. *“organization and management”/ (65,602)

44. alertness/ (19,538)

45. vigilan*.tw,kw. (32,155)

46. (responsib* adj3 (staff or professional? or nurs* or doctor? or clinician?)).tw,kw. (11,898)

47. or/38-46 [Shared Responsibility] (1,148,502)

48. 15 and 19 and 26 and 37 and 47 [Falls in Hospital and Risk Assessment and Teamwork and Shared Responsibility] (42)

Ovid MEDLINE(R) ALL 1946 to 5 May 2021

Search date: 6 May 2021

Records found: 55

Search strategy:

-------- ---------- ------------ ------------ ------------ ------------ --------------

1. Hospitalization/ (115,592)

2. Subacute Care/ (1169)

3. Hospital Units/ (10,319)

4. exp Hospitals/ (284,007)

5. medical staff, hospital/or nursing staff, hospital/ (66,506)

6. exp Patient Care Team/ (70,042)

7. Rehabilitation Centers/ (8356)

8. Inpatients/ (23,637)

9. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kf. (34,662)

10. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kf. (6803)

11. inpatient?.tw,kf. (116,838)

12. (acute adj5 patient?).tw,kf. (227,174)

13. hospital*.tw,kf. (1,383,898)

14. or/1-13 [hospital] (1,835,443)

15. Accidental Falls/pc [Prevention and Control] (9549)

16. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kf. (17,142)

17. 15 or 16 [Falls] (21,139)

18. exp Risk Assessment/ (284,216)

19. (risk adj2 (assess* or evaluat*)).tw,kf. (137,873)

20. guideline/or practice guideline/ (35,660)

21. guideline*.tw,kf. (386,469)

22. protocol?.tw,kf./freq=2 (126,031)

23. protocol?.ti. (69,596)

24. exp Clinical Protocols/ (174,448)

25. or/18-24 [Multifactorial risk assessment] (1,055,597)

26. exp Patient Care Team/ (70,042)

27. Cooperative Behavior/ (44,475)

28. collaborat*.tw,kf./freq=2 (35,346)

29. collaborat*.ti. (32,563)

30. (team* adj3 (share* or sharing or communic*)).tw,kf. (4639)

31. (team* adj3 (multidisciplin* or interdisciplin* or multi-disciplin* or inter-disciplin* or interprofessional* or inter-professional*)).tw,kf. (34,335)

32. exp interprofessional relations/or physician-nurse relations/ (70,564)

33. (team* adj5 huddle?).tw,kf. (106)

34. (team* adj5 handover?).tw,kf. (115)

35. (teamwork* or team-work*).tw,kf. (14,050)

36. or/26-35 [Teamwork] (228,807)

37. ((share* or sharing) adj3 (responsibility or responsible or accountability or accountable)).tw,kf. (2440)

38. (joint* adj3 (responsibility or responsible or accountability or accountable)).tw,kf. (604)

39. ownership.tw,kf. (13,717)

40. communicat*.tw,kf. (325,944)

41. exp Interpersonal Relations/ (334,498)

42. exp *“Organization and Administration”/ (729,180)

43. vigilan*.tw,kf. (22,242)

44. (responsib* adj3 (staff or professional? or nurs* or doctor? or clinician?)).tw,kf. (8985)

45. or/37-44 [Shared Responsibility] (1,323,953)

46. 14 and 17 and 25 and 36 and 45 [Falls in Hospital and Risk Assessment and Teamwork and Shared Responsibility] (55)

NICE Evidence

Search date: 7 May 2021

Records found: 27

(“falls prevention” or “falls assessment” or “falls reduction”) and (hospital or inpatient or “acute care” or “acute ward” or “acute patient” or “rehabilitation unit” or “rehabilitation ward”) and (“team huddle” or “team handover” or “interdisciplinary team” or “team share” or teamwork or interprofessional or “care team”) and (communication or collaboration or responsibility or ownership or accountability) and (“risk assessment” or “risk evaluation” or “evaluation of risk”)

Sorted by Relevance.| Sort by Date

Filters applied:

Evidence type: Evidence Summaries

Evidence type: Primary Research

Evidence type: Systematic Reviews

Evidence type: Audit and Inspection Reports

Evidence type: Health Technology Assessments

Web of Science Core Collection

Databases available at the University of Leeds (searched simultaneously):

• Arts and Humanities Citation Index (Web of Science) 1975–present

• Conference Proceedings Citation Index- Science (Web of Science) 1990–present

• Conference Proceedings Citation Index- Social Science and Humanities (Web of Science) 1990-present

• Science Citation Index-Expanded (Web of Science) 1900–present

• Social Sciences Citation Index (Web of Science) 1900–present

• Emerging Sources Citation Index (Web of Science) 2015–present

Search date: 6 May 2021

Records found: 9

Indexes=SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI Timespan=1900–2021

Patient participation

CINAHL (EBSCOhost)

Search date: 6 May 2021

Records found: 35

EMBASE Classic + EMBASE (Ovid) 1947 to 5 May 2021

Search date: 6 May 2021

Records found: 47

Search strategy:

--------- ----------- ----------- --------- ----------- ---------- ---------- ---------

1. hospitalization/ (420,992)

2. hospital department/or exp ward/ (439,715)

3. exp hospital/ (1,302,934)

4. medical staff/ (40,310)

5. nursing staff/ (74,427)

6. rehabilitation center/ (17,387)

7. subacute care/ (1151)

8. exp clinical handover/or collaborative care team/or exp rapid response team/ (11,414)

9. exp hospital patient/ (199,241)

10. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kw. (51,102)

11. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kw. (12,324)

12. inpatient?.tw,kw. (199,814)

13. (acute adj5 patient?).tw,kw. (369,927)

14. hospital*.tw,kw. (2,200,719)

15. or/1-14 [hospital] (3,282,770)

16. *falling/ (12,578)

17. falling/pc (2954)

18. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kw. (25,940)

19. or/16-18 [Falls prevention or risk assessment] (32,853)

20. risk assessment/ (618,392)

21. (risk adj2 (assess* or evaluat*)).tw,kw. (196,115)

22. exp practice guideline/ (601,575)

23. guideline*.tw,kw. (613,857)

24. exp clinical protocol/ (106,751)

25. protocol?.tw,kw./freq=2 (185,559)

26. or/20-25 [Multifactorial risk assessment] (1,743,357)

27. patient care/ (317,371)

28. shared medical appointment/ (137)

29. (patient? adj2 (centred or centered)).tw,kw. (36,785)

30. (patient? adj5 perspective?).tw,kw. (33,299)

31. (patient? adj4 need?).tw,kw. (123,089)

32. patient participation/ (29,364)

33. (engag* adj3 (patient? or famil* or carer? or caregiver? or player*)).tw,kw. (20,011)

34. (multifacet* or multi-facet* or tailor*).tw,kw. (164,230)

35. ((person-centred or person-centered) adj2 care).tw,kw. (3453)

36. patient autonomy/ (6173)

37. (patient? adj3 autonom*).tw,kw. (12,059)

38. ((individuali* or personali*) adj2 care).tw,kw. (8927)

39. or/27-38 [New Patient Centred Care] (689,386)

40. empathy/ (28,635)

41. compassion*.tw,kw. (16,029)

42. exp interpersonal communication/ (717,292)

43. listen*.tw,kw. (42,836)

44. respect/ (1344)

45. (empower* adj3 (staff or professional? or nurs* or doctor? or clinician? or patient?)).tw,kw. (9706)

46. empowerment/ (10,858)

47. respect*.tw,kw./freq=2 (1,001,127)

48. respect*.ti. (14,093)

49. communicat*.tw,kw./freq=2 (122,590)

50. communicat*.ti. (83,765)

51. or/40-50 [Compassion Empowerment] (1,877,721)

52. 15 and 19 and 26 and 39 and 51 [Falls in Hospital and Risk Assessment and Person-centred care and Empowerment Compassion] (47)

Ovid MEDLINE(R) ALL 1946 to 5 May 2021

Search date: 6 May 2021

Records found: 20

Search strategy:

--------- ---------- ---------- ------------ ------------ -------------- -------------

1. Hospitalization/ (115,592)

2. Subacute Care/ (1169)

3. Hospital Units/ (10,319)

4. exp Hospitals/ (284,007)

5. medical staff, hospital/or nursing staff, hospital/ (66,506)

6. exp Patient Care Team/ (70,042)

7. Rehabilitation Centers/ (8356)

8. Inpatients/ (23,637)

9. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kf. (34,662)

10. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kf. (6803)

11. (acute adj5 patient?).tw,kf. (227,174)

12. inpatient?.tw,kf. (116,838)

13. hospital*.tw,kf. (1,383,898)

14. or/1-13 [hospital] (1,835,443)

15. Accidental Falls/pc [Prevention and Control] (9549)

16. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kf. (17,142)

17. 15 or 16 [Falls] (21,139)

18. exp Risk Assessment/ (284,216)

19. (risk adj2 (assess* or evaluat*)).tw,kf. (137,873)

20. guideline/or practice guideline/ (35,660)

21. guideline*.tw,kf. (386,469)

22. protocol?.tw,kf./freq=2 (126,031)

23. protocol?.ti. (69,596)

24. exp Clinical Protocols/ (174,448)

25. or/18-24 [Multifactorial risk assessment] (1,055,597)

26. exp Patient-Centered Care/ (21,417)

27. Shared Medical Appointments/ (33)

28. (patient? adj2 (centred or centered)).tw,kf. (26,112)

29. (patient? adj5 perspective?).tw,kf. (22,183)

30. (patient? adj4 need?).tw,kf. (73,644)

31. Patient Participation/ (26,982)

32. (engag* adj3 (patient? or famil* or carer? or caregiver? or player*)).tw,kf. (13,206)

33. (multifacet* or multi-facet* or tailor*).tw,kf. (128,445)

34. (person adj2 (centred or centered)).tw,kf. (6622)

35. personal autonomy/ (17,428)

36. autonom*.tw,kf. (147,232)

37. ((individuali* or personali*) adj2 care).tw,kf. (6083)

38. or/26-37 [Patient Centred Care] (446,228)

39. Empathy/ (19,888)

40. compassion*.tw,kf. (11,571)

41. listen*.tw,kf. (35,198)

42. exp Interpersonal Relations/ (334,498)

43. respect*.tw,kf./freq=2 (655,328)

44. respect*.ti. (11,346)

45. respect/ (443)

46. empower*.tw,kf. (29,195)

47. empowerment/ (409)

48. communicat*.tw,kf./freq=2 (90,369)

49. communicat*.ti. (68,439)

50. or/39-49 [Compassion Empowerment] (1,165,982)

51. 14 and 17 and 25 and 38 and 50 [Falls in Hospital and Risk Assessment and Person-centred care and Empowerment Compassion] (20)

NICE Evidence

Search date: 7 May 2021

Records found: 12

(“falls prevention” or “falls assessment” or “falls reduction”) and (hospital or inpatient or “acute care” or “acute ward” or “acute patient” or “rehabilitation unit” or “rehabilitation ward”) and (“patient centered” or “patient centered” or patient centred or patient centered multifaceted or tailored or autonomy or “patient participation”) and (“risk assessment” or “risk evaluation” or “evaluation of risk”)

Sorted by Relevance.| Sort by Date

Filters applied:

Evidence type: Evidence Summaries

Evidence type: Primary Research

Evidence type: Systematic Reviews

Evidence type: Audit and Inspection Reports

Evidence type: Health Technology Assessments

Web of Science Core Collection

Databases available at the University of Leeds (searched simultaneously):

• Arts and Humanities Citation Index (Web of Science) 1975–present;

• Conference Proceedings Citation Index- Science (Web of Science) 1990–present;

• Conference Proceedings Citation Index- Social Science and Humanities (Web of Science) 1990–present;

• Science Citation Index-Expanded (Web of Science) 1900–present;

• Social Sciences Citation Index (Web of Science) 1900–present;

• Emerging Sources Citation Index (Web of Science) 2015–present.

Search date: 6 May 2021

Records found: 29

Indexes=SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI Timespan=1900–2021

Facilitation

CINAHL (EBSCOhost)

Search date: 1 August 2022

Records found: 43

EMBASE Classic + EMBASE <1947 to 29 July 2022>

Search date: 1 August 2022

Records found: 57

Search strategy:

-------- --------- --------- ---------- --------- ---------- ----------- --------------

1. hospitalization/ (478,011)

2. hospital department/or exp ward/ (500,590)

3. exp hospital/ (1,406,920)

4. medical staff/ (42,122)

5. nursing staff/ (76,712)

6. rehabilitation center/ (18,901)

7. subacute care/ (1508)

8. exp clinical handover/or collaborative care team/or exp rapid response team/ (13,877)

9. exp hospital patient/ (222,261)

10. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kw. (54,304)

11. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kw. (13,043)

12. inpatient?.tw,kw. (220,506)

13. (acute adj5 patient?).tw,kw. (397,349)

14. hospital*.tw,kw. (2,403,565)

15. or/1-14 [hospital] (3,569,217)

16. *falling/ (13,345)

17. falling/pc (2951)

18. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kw. (28,030)

19. or/16-18 [Falls prevention or risk assessment] (35,283)

20. implement*.tw,kw. (817,689)

21. (sustain* adj4 (program* or practice* or intervention?)).tw,kw. (16,587)

22. Adopt*.tw,kw./freq=2 (53,825)

23. Adopt*.ti. (22,299)

24. deliver*.tw,kw./freq=2 (391,321)

25. (adher* or comply or complian*).tw,kw. (565,478)

26. fidelity.tw,kw. (39,034)

27. implementation science/ (2954)

28. exp protocol compliance/ (18,558)

29. engagement.tw,kw. (112,073)

30. engage*.tw,kw./freq=2 (59,254)

31. (intervention adj2 deliver*).tw,kw. (8426)

32. (Cognitive adj2 participat*).tw,kw. (805)

33. “buy in”.tw,kw. (3176)

34. or/20-33 [Engagement or Implementation] (1,883,705)

35. clinical assessment tool/ (26,509)

36. (assess* adj4 tool?).tw,kw. (106,630)

37. (electronic adj2 record?).tw,kw. (98,321)

38. *fall risk assessment/ (660)

39. medical informatics/or nursing informatics/ (23,891)

40. exp hospital information system/ (27,015)

41. exp information technology device/ (214,637)

42. (information adj3 (technolog* or system or computeri* or electronic)).tw,kw. (73,764)

43. digital*.tw,kw. (236,637)

44. (“risk assess*” adj4 (bundle* or instrument* or care plan*)).tw,kw. (663)

45. ((multi-factorial or multifactorial) adj4 risk assessment*).tw,kw. (148)

46. (risk* adj3 screen* adj3 tool*).tw,kw. (1353)

47. clinical decision support system/ (4833)

48. clinic* decision* support system*.tw,kw. (3630)

49. health information management.tw,kw. (998)

50. medical information system/or bedside information system/ (22,768)

51. (electronic* adj3 nurs* adj3 document*).tw,kw. (155)

52. or/35-51 [Assessment tools including health info technology] (763,497)

53. workflow/ (33,077)

54. integration/ (6797)

55. data integration/ (1220)

56. facilitation.tw,kw. (40,639)

57. (workflow* or work-flow*).tw,kw. (53,221)

58. system? fit*.tw,kw. (493)

59. Embed*.tw,kw. (191,639)

60. integrat*.tw,kw. (778,087)

61. (routine* or routini*).tw,kw. (673,222)

62. or/53-61 [Workflows concept] (1,679,788)

63. 15 and 19 and 34 and 52 and 62 [HIT tools CMO4 – final] (57) Medline

Ovid MEDLINE(R) ALL <1946 to 29 July 2022>

Search date: 1 August 2022

Records found: 38

Search strategy:

-------- -------- ---------- ------------ ----------- --------- ---------- ------------

1. Hospitalization/ (129,166)

2. Subacute Care/ (1351)

3. Hospital Units/ (10,446)

4. exp Hospitals/ (306,482)

5. medical staff, hospital/or nursing staff, hospital/ (68,147)

6. exp Patient Care Team/ (72,215)

7. Rehabilitation Centers/ (8576)

8. Inpatients/ (27,507)

9. ((acute or sub-acute or subacute) adj3 (care or ward?)).tw,kf. (38,212)

10. ((rehabilitation or geriatric) adj (ward? or unit? or department?)).tw,kf. (7341)

11. (acute adj5 patient?).tw,kf. (245,691)

12. inpatient?.tw,kf. (130,036)

13. hospital*.tw,kf. (1,525,967)

14. or/1-13 [hospital] (2,001,274)

15. Accidental Falls/pc [Prevention & Control] (10,342)

16. (fall? adj2 (assess* or risk? or prevent* or reduc* or occur* or frequen* or screen*)).tw,kf. (19,084)

17. 15 or 16 [Falls] (23,286)

18. implement*.tw,kf. (628,881)

19. (sustain* adj4 (program* or practice* or intervention?)).tw,kf. (13,120)

20. Adopt*.tw,kf./freq=2 (42,809)

21. Adopt*.ti. (18,039)

22. deliver*.tw,kf./freq=2 (288,133)

23. deliver*.ti. (150,863)

24. (adher* or comply or complian*).tw,kf. (377,119)

25. fidelity.tw,kf. (34,472)

26. Health Plan Implementation/ (6629)

27. implementation science/or technology transfer/ (3267)

28. Guideline Adherence/ (34,788)

29. engage*.tw,kf./freq=2 (47,833)

30. engage*.ti. (20,598)

31. (intervention adj2 deliver*).tw,kf. (6761)

32. (Cognitive adj2 participat*).tw,kf. (627)

33. “buy in”.tw,kf. (2105)

34. or/18-33 [Implementation or Adherence to Guidelines and strategies] (1,419,513)

35. Risk Assessment/mt [Methods] (38,091)

36. (assess* adj4 tool?).tw,kf. (74,557)

37. (electronic adj2 record?).tw,kf. (53,541)

38. medical informatics/or health information exchange/or exp medical informatics applications/or exp medical informatics computing/or nursing informatics/ (490,892)

39. exp Hospital Information Systems/ (30,283)

40. exp health information management/ (1943)

41. exp Medical Records Systems, Computerized/ (46,872)

42. (information adj3 (technolog* or system or computeri* or electronic)).tw,kf. (59,321)

43. digital*.tw,kf. (181,775)

44. (“risk assess*” adj4 (bundle* or instrument* or care plan*)).tw,kf. (525)

45. clinic* decision* support system*.tw,kf. (2939)

46. (electronic* adj3 nurs* adj3 document*).tw,kf. (123)

47. Decision Support Systems, Clinical/ (9161)

48. or/35-47 [Assessment tools including health IT] (868,790)

49. workflow/ (8215)

50. systems integration/ (9627)

51. facilitation.tw,kf. (32,835)

52. (workflow* or work-flow*).tw,kf. (37,281)

53. system? fit*.tw,kf. (368)

54. Embed*.tw,kf. (153,897)

55. integrat*.tw,kf. (644,447)

56. (routine* or routini*).tw,kf. (440,520)

57. or/49-56 [Workflows] (1,271,193)

58. 14 and 17 and 34 and 48 and 57 [Facilitation Tools CMO 4] (38)

Web of Science Core Collection

Databases available at the University of Leeds (searched simultaneously):

• Arts and Humanities Citation Index (Web of Science) 1975–present;

• Conference Proceedings Citation Index- Science (Web of Science) 1990–present;

• Conference Proceedings Citation Index- Social Science and Humanities (Web of Science) 1990–present;

• Science Citation Index-Expanded (Web of Science) 1900–present;

• Social Sciences Citation Index (Web of Science) 1900–present;

• Emerging Sources Citation Index (Web of Science) 2015–present.

Search date: 1 August 2022

Records found: 38

Indexes=SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI Timespan=1900–2021

1. # 24 #23 AND #22 AND #10 AND #6 AND #5

2. # 23 TOPIC: (facilitation or workflow* or work-flow* or embed* or integrat* or routine* or routini* or “system* fit*”)

3. # 22 #21 OR #20 OR #19 OR #18 OR #17 OR #16 OR #15 OR #14 OR #13 OR #12 OR #11

4. # 21 TOPIC: (electronic* near/3 nurs* near/3 document*)

5. # 20 TOPIC: (“health information management”)

6. # 19 TOPIC: (“clinic* decision* support system*”)

7. # 18 TOPIC: (risk* near/3 screen* near/3 tool*)

8. # 17 TOPIC: ((multi-factorial or multifactorial) near/4 risk assessment*)

9. # 16 TS=(“risk assess*” near/4 (“care plan*”))

10. # 15 TS=(“risk assess*” near/4 (bundle* or instrument*))

11. # 14 TOPIC: (digital*)

12. # 13 TOPIC: (electronic near/2 record$)

13. # 12 TOPIC: (information near/3 (technolog* or system or computeri* or electronic))

14. # 11 TOPIC: (assess* near/4 tool$)

15. # 10 #9 OR #8 OR #7

16. # 9 TOPIC: (sustain* near/4 (program* or practice* or intervention$))

17. # 8 TI=(adopt* or deliver* or engage*)

18. # 7 TS=(implement* or adher* or comply or complian* or fidelity or “buy in” or (Cognitive near/2 participat*))

19. # 6 TS=(fall$ near/2 (assess* or risk$ or prevent* or reduc* or occur* or frequen* or screen*))

20. # 5 #4 OR #3 OR #2 OR #1

21. # 4 TOPIC: (acute near/5 patient$)

22. # 3 TOPIC: ((rehabilitation or geriatric) near/1 (ward$ or unit$ or department$))

23. # 2 TOPIC: ((acute or sub-acute or subacute) near/3 (care or ward$))

24. # 1 TOPIC: (hospital* OR inpatient*)