In total 1,491 citations and abstracts and 467 full texts were screened, resulting in the inclusion of 28 manuscripts for Facilitation (15–42) and 24 manuscripts for Patient Participation (23, 25, 43–64), see Fig. 2 for PRISMA diagram. The studies were conducted in a range of countries and encompassed a range of methods, many were quality improvement projects describing examples from practice, see Appendix iii for summary of study details.
Findings are presented below by CMOc and the questions addressed in the analysis.
CMOc 1: Facilitation via tools
How are MFRA tools designed to facilitate implementation of falls prevention practices?
Ten studies described use of published tools such as the Morse Fall Scale (16, 18, 23–25, 32), the Memorial Emergency Department Fall Risk Assessment Tool (MEDFRAT) (33), the Fall Risk for Older People (FROP) (37) and KINDER 1 (19, 41). Seven studies described locally developed tools e.g., developed through review of the falls literature and/or identifying common risk factors on a particular unit (17, 21, 26, 28, 34–36). In nine studies, it was unclear whether the assessment tool was publicly available or locally developed (15, 20, 22, 27, 29, 31, 37, 40, 42).
MFRA tools offered a structure of items to guide identification of individual falls risk factors, but the number and type varied. See Table 2 for comparison against items recommended in the NICE guidance.
Table 2
Example of falls risk items included in different tools
| NICE (2013) | MORSE Fall Scale | KINDER 1 | Site specific (17) |
1 | Cognitive Impairment | Mental status | Altered mental state | Disorientated |
2 | Continence | N/A | N/A | Requires assistance toileting |
3 | Falls history | History of falling | Presented to ED due to fall | Two falls in the last 12 months |
4 | Footwear | N/A | N/A | N/A |
5 | Health problems | Secondary diagnosis (more than two medical diagnoses) | N/A | N/A |
6 | Medication | N/A | N/A | High risk medication Patient taking more than four medications |
7 | Postural instability, Mobility/balance | Gait | Impaired mobility | Unsteady gait |
8 | Syncope syndrome | N/A | N/A | N/A |
9 | Visual impairment | N/A | N/A | N/A |
Other | N/A | Ambulatory aid IV / No IV | Age 70 or older Nurse judgement | At risk behaviours |
Table 2 indicates how falls risk factor items differ depending on the tool. Cognitive impairment, mobility and history of falls are commonly included items. However, items that appear similar may not prompt the same information e.g., falls history within the NICE guidance refers to how, where, when, and why falls occur which might identify syncope or other treatable causes, whilst similar items in the tools listed require a ‘yes/no’ response. Some studies described how tools were designed to support decision-making in choice of intervention in response to risk factors identified e.g., providing guidance about interventions to implement in response to falls risk factors (17, 26, 28, 29, 35); some hybrid tools also recommended standard intervention bundles for patients assessed as high risk. One study focused on intervention delivery, providing quick reference guides, organised by risk factor area, to inform choice of intervention (40). Several studies used visual tools e.g., posters, to remind staff, patients, and carers of interventions in place for individual falls risk factors (23–25, 27, 37, 38).
Nine studies described that the MFRA and care plans were integrated into the Electronic Healthcare Record (EHR), digitising documentation of falls prevention practices (15, 19, 22, 32, 33, 35, 36, 41, 42). Manuscripts that focused on the role of Health Information Technology (HIT) included assessment of the impact of digitising MFRA documentation (22, 42), automating parts of the assessment and/or care planning process e.g., automatically generating a care plan with interventions linked to the falls risk factors identified during the assessment (18, 20, 23, 24, 37, 38, 65) and an evaluation of EHR alerts that notified staff to incomplete documentation (31).
To what extent were falls risk assessments and interventions delivered?
Twelve studies assessed delivery of a MFRA as documented in clinical records with improvement post-intervention reported in 11 studies (15, 18, 20–22, 28, 29, 31, 32, 35, 42), encompassing paper-based and HIT tools. One study found MFRA delivery was consistent pre- and post-intervention (19). Documentation of a care plan in clinical records was reported in seven studies (15, 18, 31, 32, 35, 40, 42). Lytle et al (31) reported that documentation of risk assessments improved significantly, in response to electronic alerts, whilst care plans did not. Wu et al (42) showed that digitisation improved documentation of practice but care plans were not documented for all patients assessed as high risk of falls. Three studies reported use of targeted interventions, two of which demonstrated an improvement post-intervention (32, 35) and one a decline in two out three wards studied (15). Titler et al (40) reported significant improvement (p < 0.001) for use of specific interventions including for mobility, toileting, and cognition, but not for medications. In Carroll et al, (18) documentation of a MFRA and care plan improved, whilst documentation of intervention delivery did not. Three studies measured adherence displaying a bed side poster generated from Fall TIPS (23–25), a HIT intervention that aimed to involve patients, with their families and carers, in the assessment and care planning process to overcome patient non-adherence to falls prevention strategies. In summary, whilst documentation of a MFRA improved quite consistently across studies (where reported), there was variation in impact regarding documentation of care plans and interventions delivered.
Why and in what circumstances do tools facilitate falls prevention practices?
There was a paucity of data detailing staff experiences using MFRA tools, although some studies provided an explanation as to why particular tools were chosen, e.g., to reduce variation in the assessment content by providing a standardised structure (32), and to improve risk identification by introducing items tailored to the patient population (17, 19, 33). Some authors suggested tools may work simply by drawing staff attention to required practices, acting as a prompt (21, 26, 29). To work in this way, evidence indicated that tools, paper-based and HIT-based, need to be clearly visible to staff in their work processes (15, 31, 33). Automation of practices via HIT removed task loads from clinical staff – automatically linking falls risk factors to interventions and generating a care plan - but introduced novel manual work such as displaying and updating bedside posters, that brought new challenges e.g., remembering to move posters when patients swapped beds (38). One study suggested new manual tasks may be seen as a competing priority for which staff do not have time (27).
Educational strategies, such as training and feedback, were highlighted as supports for tool use because they raised staff awareness of the tool, increased their knowledge of falls prevention practices, and evidenced the importance of following tool guidance (15, 17, 29, 30). However, it was not possible to distinguish the impact of individual interventions as studies often incorporated multiple strategies to improve practice. Furthermore, HIT was found to introduce additional training needs e.g., in one study staff were motivated to use HIT but required more training than had been provided, to use the technology itself (38).
There was some data to suggest that staff responded well to tools that provided space to document clinical judgement, particularly where stratification (a practice no longer recommended by NICE) was used. For example, a hybrid tool recommended remote video monitoring to patients stratified as high risk of falls (19). The authors reported that staff felt empowered by a clinical judgement item to allocate this intervention to patients most in need and according to resources available. Other studies provided further insight into to the problems of stratification, e.g., in one study staff were confused over the definition of high risk patients because they did not always judge a patient to be at risk when indicators on the tool suggested that they were (31). One study suggested that discrepancies between tool stratification of patients as high risk and nurses’ clinical judgement may help explain why care plans were not documented consistently for patients (42).
Alongside clinical judgement, the studies pointed to a number of factors that influenced the extent to which tools acted as practice facilitators e.g., changes in patient condition and transition between wards were highlighted as circumstances that may disrupt tool use and documentation of care plans (32, 38). Lack of communication of the falls prevention plan between different professional groups and availability of physical resources, e.g., non-skid socks, may constrain delivery of interventions suggested by tools (21, 33, 34, 40). Furthermore, hospital IT infrastructure dictated what HIT was available to staff at the ward level e.g., whether automation was available or not (23).
Programme Theory Refinement
The Facilitation analysis was used to refine the CMOcs and overarching IPT (see Table 3).
Table 3
Facilitation: Programme Theory Refinement
| Mechanism | |
Context | Intervention | Staff Response | Outcome |
Where staff understand (through experience, or education or feedback) how and why falls prevention practices reduce falls risk factors. | MFRA tools are located visibly and intuitively in the EHR or ward practice and offer a structure to guide identification of fall risk factors. However, assessment tools vary in type and number of assessment items. | Reminder: Tool draws staff attention to the tasks required e.g., completing an assessment of individual falls risk factors and prompts action. | More consistent documentation and delivery of falls risk assessments but content of assessment may differ depending on tool used by service. |
Ward conditions are complex – patients’ condition may change, they may swap beds or move wards, and they may require multiple interventions. | MFRA tools are located visibly and intuitively in the EHR or ward practice and offer a structure to guide identification of fall risk factors. However, tools vary in type and number of assessment items. | Prioritisation: Staff attention is focused on care delivery rather than documenting care processes. | Documentation of care process may be less consistent, particularly after the initial falls risk factor assessment. |
Staff who are educated and experienced in identifying and managing falls risk factors. | MFRA tools are visible to staff in their work routines and provide guidance for assessing risk and linking risk factors with interventions. | Clinical Judgement: Tool guidance does not align with clinical judgement or resources available - staff apply care according to their own judgement. | Care may not be in line with tool recommendation, but action taken to manage risks using ward resources. |
IT systems support HIT function and staff are trained and experienced with use of HIT. Where staff understand (through experience, or education or feedback) how and why falls prevention practices reduce falls risk factors. | MFRA tools are located visibly and intuitively in the EHR or ward practice environment. Care plans, poster and information leaflet automatically generated from software. | Automation: Interventions to address falls risks automatically selected and documented in care plan and patient poster. Staff display poster at patient bedside and action care plan. | Reduced variation in development and documentation of care plan that links falls risks with appropriate interventions. Task load of clinical staff reduced. Falls prevention strategy more visible in poster at patient bedside. |
Manual work: Staff see manual work as competing priority with other responsibilities. | Display of poster may be disrupted by patient flow e.g., between beds and wards. |
CMOc area 2: Patient Participation
What are the characteristics of interventions designed to encourage patient participation?
Nine studies (25, 43, 46, 49, 50, 53, 54, 57, 63) examined interventions that sought to engage patients in the assessment and/or care planning process to encourage their participation in falls prevention interventions. Radecki et al (54) and Sitzer (57) introduced tools that enabled patients to self-assess their falls risk, recognising a discrepancy between patients’ and professionals’ perception. Martin et al (53) evaluated the Safe Recovery Programme in which ward staff and volunteers worked with patients on one or more occasions to develop personalised goals to prevent falls. Haines et al (49) compared two approaches, one in which patients were provided with educational materials, and a ‘complete programme’ where materials were supplemented by one or more follow-ups with a physiotherapist for goal-setting and review. Three studies examined Fall Safety Agreements (43, 50, 64) e.g., Bargmann et al (43) introduced an agreement that patients signed to confirm that they had been educated on fall risk prevention strategies, acknowledged falling could cause serious injuries and therefore agreed to ask for help to prevent falls. Five studies examined Fall TIPS (23, 25, 46, 63), an intervention in which staff, patients and their carers worked in partnership throughout the assessment and care planning processes to prevent falls. However, how patients and carers interacted with staff during these processes was not explained.
Three studies examined interventions where patient participation was encouraged during comfort rounds, also known as intentional, purposeful, or hourly rounding (44, 48, 60). During intentional rounding, staff asked about patients’ immediate and personal needs (44, 48, 60). Cann and Gardner (44) described their aim as moving from a ‘patient allocation’ to a ‘practice partnership’ model of care, within which intentional rounds were intended to support patients to participate more fully in their own care. Goldsack et al (48) examined hourly rounding with an intention of decreasing call bell usage, by engaging patients as active partners and Zadvinkis (60) conducted a survey, part of which was about intentional rounding, but no information was provided about what form rounding took.
To what extent do patients participate in falls prevention practices?
There was limited data evidencing the extent to which patients participated in falls prevention interventions. One study reported a significant reduction in patients’ use of call bells from 1277 uses per 100,000 patient hours to 523 uses (P = < 0.001) after comfort rounds (44). Two studies described the types of goal that were set during patient and professional interactions (49, 53). Common goals in both studies included working more effectively with healthcare staff, identifying environmental hazards, and using appropriate aids and equipment.
A more commonly measured impact was patient knowledge. Seven studies measured patient understanding of their falls risks and care plan with varied results (23, 43, 53, 54, 61, 63, 65) e.g., Radecki et al (54) conducted a knowledge-in-action survey which showed statistically significant improvements between baseline and intervention groups (P = .0007) in patient involvement in care planning. However, there was no significant difference in other questions, including whether the prevention plan was always followed. One study (examining Fall TIPS) reported that patient activation, a term that encompassed knowledge, skill, and confidence to participate in falls prevention, improved preintervention to postintervention at three sites, with the mean score improving from 63.82 (standard deviation [SD] ± 17.35) to 80.88 (SD ± 17.48), p < 0.0001. Bargmann et al (43) used staff incentives to increase adherence to programme implementation, which was thought to have supported an increase from 30% (5 out of 17 patients) to 95% of patients correctly stating their falls risk.
Why and in what circumstances do patients participate in falls prevention strategies?
The studies suggested patient attitudes, beliefs and understanding about falls risks may constrain their participation in falls prevention interventions e.g., patients may be reluctant to use the call bell for fear of disturbing busy nurses (45, 53), they may not accept they are at risk of falling (53, 55, 56, 58), and patients that have had a recent fall may be more likely to engage in falls prevention than patients who have not (51). Studies also reported intention to act e.g., asking for help using the toilet, may not be followed through if the help requested is not forthcoming and patients feel confident to act alone (59) or are unable to wait due to urgency (58). Additionally, patients may not be physically able to participate e.g., depending on where the call bell is placed (45).
In the Safe Recovery Programme, introduced previously, Martin et al (53) explained that individualising messages to address patient circumstances, such as those described above, may trigger participation mechanisms such as gaining permission to ask for help, empowerment to act, and increased awareness of risk. The quality of interaction between nurse and patient was highlighted as key to successful messaging. Volunteers in Martin’s study were said to have skills such as listening, teaching, and reflecting that created engaging, personalised, safe interactional spaces. Similarly, one study suggested nurses with more experience (defined as two or more years) moved the risk assessment process from ‘task mode’ to a vehicle to enhance communication and partnership that authors linked to falls reduction (54). Based on previous experience, effectively communicating the care plan to patients was emphasised as a key component of Fall TIPS, with studies evidencing a reduction in fall rates and improvements in patient activation (23, 46, 63).
Few studies included patients with cognitive impairment (49, 51, 53, 66). Martin et al (53) included patients with mild cognitive impairment and explained that posters and environmental cues e.g., call bell in place, may work as reminders for these patients to avoid risk-taking behaviours. A key finding came from Haines et all [95], where results indicated that participants with impaired cognitive function in the complete programme (that included goal-setting and review) incurred a significantly higher rate of injurious falls per 1000 patient-days than those in the control group (7.49 vs 2.89, p = 0.02). The authors stated that cognitive impairment may have constrained patients’ ability to adhere to safety plans, as well as making them more willing to report injuries from falls
In a study of nurses’ experience of falls prevention, participants described using a combination of formal assessment, monitoring and communication as part of an ongoing strategy of ‘knowing that the patient is safe’ (56). These strategies enabled nurses to be responsive to patients’ requests for help and ensure safety even if patients are unable to participate in interventions fully in response to messaging e.g., due to cognitive impairment. However, constraints were described for each strategy. For example, low staffing levels reduced nurse vigilance when making patient rounds and constrained the direct patient contact needed to know patients were safe.
Programme Theory Refinement
The Patient Participation analysis was used to refine the CMOc s and overarching IPT (see Table 4).
Table 4
Patient Participation Programme Theory Refinement
Context | Mechanism | Outcome |
Resource | Response |
Patients with capacity have different perspectives and circumstances that may influence if/how they participate in falls prevention strategies in hospital. Staff have the time and skills to create an interactional rather than task focused space for assessment and care planning. | Staff individualise falls prevention messages for patients, i.e., that account for their circumstances and perspectives. | Patient empowerment: patients are empowered (increased confidence to ask for help, knowledge about their falls prevention strategy, acceptance of their falls risks) to participate in appropriate strategies. | Patients participate in interventions such as using the call bell and avoid risk taking behaviour. |
If staff are not responsive to patients’ requests for help mobilising or performing functional tasks e.g., due to task load / awareness. | Staff individualise goal setting and falls prevention messages to patient, i.e., that account for their circumstances and perspectives. | Taking risks: Patient confident they can, or feel urgency to, mobilise by themselves e.g., to get to the toilet. | Patient at risk of falling, particularly if hurrying. |
Patients with cognitive impairment have falls risk factors like other patients but additionally may have problems with memory, attention and confusion. | Staff individualise falls prevention messages to patient i.e., that address their emotional barriers to participation. | Taking risks: Patients may struggle to understand or retain information and are unable to communicate needs unambiguously to staff, despite messaging. | Patients engage in risk taking behaviour. |
Patients with cognitive impairment have falls risk factors like other patients but additionally may have problems with memory, attention and confusion. | Staff undertake ongoing assessment of risk and monitoring of patient. | Knowing the patient is safe: Staff collate the information necessary to understand if the patient is safe from harm. | Staff intervene in behaviour that may lead to a fall. |