Abstract
Accidental falls are preventable adverse events for older post-acute care (PAC) patients. Yet, due to the functional and medical care needs of this population, there is little guidance to inform multidisciplinary prevention efforts. This scoping review aims to characterize the evidence for multifactorial PAC fall prevention interventions. Of the 33 included studies, common PAC intervention domains included implementing facility-based strategies (e.g., staff education), evaluating patient-specific fall risk factors (e.g., function), and developing an individualized risk profile and treatment plan that targets the patient’s constellation of fall risk factors. However, there was variability across studies in how and to what extent the domains were addressed. While further research is warranted, health system efforts to prevent accidental falls in PAC should consider a patient-centered multifactorial approach that fosters a culture of safety, addresses individuals’ fall risk, and champions a multidisciplinary team.
What this paper adds
• Synthesizes current evidence for multifactorial PAC fall prevention interventions
Applications of study findings
• Addresses PAC fall prevention, an adverse event prioritized by health systems, providers, patients, and caregivers • Reinforces the importance of a multidisciplinary team approach to preventing accidental falls in PAC
An estimated 21% of older adults experience a fall after admission to United States (US) post-acute care (PAC) facility (Leland et al., 2012). Accidental falls can be devastating for PAC patients who are unable to return home immediately after acute hospital discharge. These adverse events are associated with an increased risk of hospital readmissions, long-term institutionalization, and subsequent falls and fractures (Berry et al., 2008; Mahoney et al., 2000).
Although preventable, falls can be caused by intrinsic and extrinsic risk factors, such as comorbid conditions, polypharmacy, functional limitations, and environment hazards (Tricco et al., 2017). Complicating PAC fall prevention, fall risk changes throughout the brief episode of care (Williams et al., 2007). The variation throughout the PAC stay can be attributed to new medications and functional limitations post-hospitalization, a PAC plan of care which aims to foster recovery, and being in an unfamiliar environment (Leland et al., 2012).
The objective of US PAC is to foster patients’ functional and medical recovery after a brief acute hospitalization so they can safely return to the community (Leland et al., 2018). This philosophy is reinforced by US value-based healthcare initiatives that hold health systems accountable for care quality. For example, major falls (i.e., falls resulting in an injury) and community discharge (i.e., getting back home and staying home for 30-days) are two PAC quality measures that incentivize coordinated comprehensive multidisciplinary patient-centered care. Further, these initiatives encourage the integration of fall prevention into the PAC plan of care to enhance patient safety, prevent facility adverse events, and decrease the risk of readmissions.
The multifactorial approach to fall prevention targets the individual’s multiple intrinsic and extrinsic risk factors via modification (e.g., strengthening exercises) or compensatory strategies. Recent systematic reviews support the benefits of multifactorial fall prevention approaches in hospital, clinic, and community-based settings (Cheng et al., 2018; Tinetti & Kumar, 2010; Tricco et al., 2017). Yet, there is a paucity of PAC multifactorial fall prevention literature. Thus, we investigate the evidence for multifactorial fall prevention interventions within a PAC healthcare context.
Methods
This scoping review used the methods developed by (Arksey & O’Malley, 2005) and the PRISMA extension for scoping reviews (PRISMA-ScR) reporting guidelines, to address the research question, review evidence, describe the breadth of evidence, and assess the quality (Tricco et al., 2018). The PRISMA-SCR is a modification of the original reporting guidelines for systematic reviews the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) Statement (Sarkis-Onofre et al., 2021). Existing quality appraisal tools were used to assess the risk of bias for controlled and non-controlled studies (Higgins & Green, 2011; National Heart, Lung, and Blood Institute, 2014). A predetermined protocol was developed by the research team.
Data Sources and Search Strategy
Relevant studies were retrieved across four databases: MEDLINE (PubMed), Ageline, Cumulative Index to Nursing and Allied Health Literature Complete (CINAHL Complete), and Cochrane Library. Keywords and controlled vocabulary were developed in consultation with a research librarian for use in PubMed (Online Appendix A) and modified as needed for the remaining databases.
Inclusion and Exclusion Criteria
Studies were included if the intervention: (a) included a sample of patients with an average age of 65 years or older; (b) measured a fall-related primary outcome; (c) was delivered in a post-acute care setting (e.g., skilled nursing facility, inpatient rehabilitation facility); and (d) investigated multifactorial fall prevention interventions. International healthcare settings that are similar to a US PAC context were also included (i.e., service delivery, length of stay, multidisciplinary care team). Included studies were written in English and published between January 2006–March 2021. This timeframe was informed by preliminary searches that found studies prior to 2006 did not examine PAC specifically or it was integrated with other care settings. Studies were excluded if they were (a) not an intervention; (b) conducted outside a PAC setting (e.g., community-based setting); (c) targeted one fall risk factor; or (d) were conference abstracts or dissertations. We excluded systematic reviews as they did not capture the details of each intervention study, but hand searched their reference lists.
Study Screening and Selection
Two members of the research team systematically reviewed each article for eligibility by title, abstract, and then full text. Articles that did not meet the inclusion criteria were excluded. Disagreements about eligibility were resolved through review and discussion by the entire team.
Data Collection and Analysis
Data extraction of each article was conducted by two members of the research team included the following: author(s), publication year, study design, country in which the study was conducted, sample characteristics (e.g., age, gender), care setting, intervention elements, primary outcomes, and results. Each article was categorized according to Kocak et al.’s (2011) operationalization of study levels of evidence (e.g., level I= randomized control trials, level II= non-randomized controlled trials, level III= case reports, clinical trials without a control). A consensus procedure was used to address any disagreements between authors with respect to data extraction.
After data extraction, an iterative process of reviewing fall prevention intervention elements was completed, which could then be grouped together to identify intervention components, and finally intervention domains. This consensus-driven process first included reviewing intervention elements individually (e.g., dynamic standing balance activity) and collectively. The team discussed themes that could group similar elements into intervention components (e.g., balance). Next, the fall prevention components were clustered into overarching intervention domains (e.g., facility-based interventions). Within each intervention domain, study components were categorized as a screen, evaluation, or intervention, which enabled the differentiation between processes for identifying risk (i.e., screening or evaluating) and intervening on a given domain via modification or compensation. Given the multiple domains reflected within a single multifactorial intervention, manuscripts can be reflected in multiple domains. However, not all intervention domains had to be addressed to be included in the sample. Thus, our sample is not described in mutually exclusive categories.
Results
The review of the literature resulted in an initial sample of 113,444 articles. After removing duplicates and initial screening at the title and abstract level, 3652 full articles were reviewed resulting in a final sample of 33 publications (Figure 1). Sample flow diagram.
Distribution of Domains by Study (n = 33).
Notes. E = Evaluation, I = Intervention, (+) = significant, (−) = insignificant, (+/−) = mixed, NR = significance not reported.
Overview of Care Practices
Six domains of fall prevention interventions (Table 1) emerged across the sample. The first, facility-based interventions, targeted system-wide efforts to promote safety. The remaining five reflect patient-specific fall risk domains (i.e., function, cognitive/psychosocial, home environment/adaptive equipment, medical status, individualized fall risk profile). For each of the five patient-level domains the (a) evaluation and (b) intervention components were described.
Twenty-three studies utilized facility prevention protocols. This process began by screening to identify those at risk, conveying that risk to the team, and implementing systematic preventative strategies. After conducting an admission fall risk screen all 23 studies used the results to stratify patients based by fall risk. Of those, five studies had a facility protocol to stratify patients by low, medium, or high fall risk, which informed staff of individual patient care needs. The remaining 18 studies stratified patient fall risk dichotomously (yes/no). Each facility protocol outlined a series of pre-specified fall prevention strategies that guided implementation of pre-established, system-wide approaches for all at risk patients. For example, 10 studies notified staff of a patient’s risk through a wristband worn by the patient or signage posted above the patient’s door.
Thirteen studies systematically altered the health care environment (e.g., utilizing low beds) to meet the patient-specific needs, or ensured that assistive devices and call bells were within reach of the patient within their environment. Six studies altered the patient’s environment by enhancing supervision of patients with cognitive impairment by moving the patient closer to the nurses’ station (e.g., room location) and/or altering staff workflow in the environment to increase the frequency of staff contact with the individual throughout the day.
Sixteen studies incorporated staff fall prevention education to ensure a (a) common level of knowledge and (b) systematic delivery of the intervention. Of those, six studies used didactic training and clinical reminder tools together, such as falls prevention binders that included a list of necessary tasks and falls report indexes. Across the 16 studies that incorporated staff education, seven also implemented multidisciplinary fall prevention discussions facilitated by research educators, fall prevention teams, or designated staff members. This approach intended to foster problem-solving and facilitate knowledge translation among staff. For example, Barker et al. (2016) used a multi-component implementation strategy to educate and empower staff to consistently address fall prevention. Their intervention included PAC clinical leaders and unit champions to reinforce intervention delivery, execute staff audits, provide clinical reminders, and offer non-punitive staff feedback. To foster buy-in of the study, Bonner et al. (2007) engaged in an iterative collaborative process with the facility stakeholders to (a) design the education program and staff resources and (b) foster engagement of frontline staff leaders that would champion the program.
Post-fall protocols were integrated into five studies, which included a root cause analyses along with standardized event documentation (e.g., location, cause). The goal was to comprehensively describe the incident, establish a plan, and proactively prevent subsequent falls for that resident. The information collected also helped inform facility efforts to enhance prevention efforts.
Five studies used quality improvement initiatives in their facilities to drive facility-based fall prevention approaches. These initiatives used a stepwise approach in which a facility falls champion or administrative liaison introduced one prevention strategy at a time, ensuring that staff mastered a single process before adding the next approach to the program. This quality improvement method helped facilities coordinate care among members of the care team given the diverse multifactorial, multidisciplinary fall prevention components included in the initiative.
Evaluation: Twenty-three studies evaluated a patient’s multiple fall risk factors in the context of balance, mobility, gait, and/or self-care using a variety of assessment tools (Online Appendix C). Balance was the most frequently assessed risk factor in the context of function. Balance and/or mobility were assessed using standardized assessments, such as the Timed Up and Go (n = 4) and the Performance-Oriented Mobility Assessment (n = 2). Conversely, one study used their own locally developed tool (n = 1), and two studies did not report the tools that were used. Routine acts of self-care (n = 19) were also assessed with standardized tools, such as the Barthel Index Score (n = 5). The results from these targeted assessments informed the patient’s individualized fall prevention intervention.
Intervention: Twenty-seven studies described interventions for improving function to prevent falls. Six studies implemented exercise to enhance function, which included rote exercises (n = 4), or a combination of functionally oriented and rote exercises (n = 2). One study implemented Tai Chi into the exercise routine. Six studies encouraged mobility through targeted early and frequent ambulation in addition to an exercise routine. To address balance limitations, seven studies focused on balance training and five implemented strengthening routines. Frequent and scheduled toileting was implemented in two studies to limit falls related to incontinence (e.g., rushing). ADL training was utilized in five studies to reinforce the use of fall prevention strategies in regularly occurring tasks and to manage associated fall risks (e.g., managing base of support while standing at the sink during grooming activities).
Evaluation: Twenty-two studies included screening of cognitive and psychosocial factors as a strategy to prevent falls (Online Appendix C). Twelve studies used standardized assessments such as the Mini Mental Status Exam to identify dementia or delirium. Four studies screened patients for fear of falling (e.g., Falls Efficacy Scale International). Depression was measured using the Geriatric Depression Scale (n = 7).
Intervention: Ten studies delivered interventions for patients identified as having impaired cognition. These interventions included orienting patients to the facility environment and establishing individualized schedules. Interventions were meant to stimulate thought and help patients establish daily routines such as knowing when to expect meals or visits from nursing staff. Four studies engaged patients in diversional activities conducted by facility volunteers to increase supervision and minimize fall risks, including preventing unsafe wandering (n = 1) and mitigating aggressive behaviors (n = 1). Two studies implemented agitation reduction techniques among patients with dementia, delirium, or acute confusion, through ensuring one-on-one attention from nursing staff.
Cognitive behavioral therapy (n = 1) was implemented to address fear of falling. This approach sought to address fear so that patient confidence or efficacy in maintaining balance accurately represented their level of fall risk. Patients were provided with strategies, such as asking for help when needed and developing the ability to express their feelings about falls openly. Patients were educated on techniques to safely carry out daily activities relative to their fall risk, instead of avoiding activity or engaging in risky behavior.
Evaluation: Six studies included an evaluation of the patients’ home environments in an effort to equip them with safe home practices, as well as empower patients (and caregivers) with strategies to safely use assistive devices. These services were offered in the form of an environmental audit. Only one study used standardized assessments (Online Appendix C). Information from these assessments was used to guide efforts to equip the patient (and caregiver) with the knowledge and skills to optimize safety and minimize fall risks through environmental modifications in the home.
Intervention: Education and training on the safe use of adaptive equipment (e.g., cane) was included in 16 studies and two recommended the use of hip protectors to prevent an injurious fall (i.e., fracture) among high fall risk individuals. Some of these studies educated (n = 5) patients about assistive devices, such as walkers, to ensure proper usage.
Evaluation: Eight studies examined patient fall risk in the context of medical status, including overall health, nutrition, and/or sensory capabilities. A comprehensive medication history was completed in four studies to identify prescriptions that may contribute to an individual’s fall risk (i.e., polypharmacy). Three studies included a standard medical history. Sensory systems were assessed in the context of fall risk, including hearing and vision (n = 2) or vision alone (n = 3). One study assessed malnourishment using the Mini Nutritional Assessment (Online Appendix C).
Intervention: Nine studies went beyond identification of polypharmacy to implement a medication review and intervention plan to decrease fall and fracture risk. In seven of these studies, the medical team, family, and individual worked together to explore alternatives to mitigate fall risk related to medications. Three studies targeted nutrition (e.g., adequate hydration). Patient diet was further supplemented with vitamin D (n = 4) to strengthen bones to mitigate fracture risk in case fall occurred. In addition, eight studies provided additional medical management within fall prevention programs (e.g., sleep interventions).
Evaluation: Twenty-six studies completed patient-specific fall risk evaluations for those identified as a fall risk. This process included identifying the patient’s cumulative risk factors and ascertaining their falls history. The characterization of a patient-specific fall risk profile guided the individually tailored intervention by determining an individual’s particular vulnerabilities, skills needed, and appropriate educational goals. This was done with standardized assessment tools (Online Appendix C), such as the Mores Fall Risk Scale (n = 3), an informal, locally developed tool (n = 6), or a combination of both types of tools (n = 3). In addition to the identification of these risk factors, a patient’s falls history (n = 16) was taken to better inform providers of situations in which falls may occur for each patient. Of the 26 studies, only 19 studies then implemented an individualized fall prevention plan based on this patient-specific fall risk evaluation.
Intervention: Based on the individualized fall risk profile, patient education was targeted in 11 studies, family education in two studies, and both patient and family education were implemented in five prevention programs. Patient falls prevention education was delivered through one-on-one instruction from a staff member (n = 9), packets of printed information (n = 4), or a combination of both (n = 5). To supplement social supports and better enable care at home, family and caregiver education was provided in two studies, one of which integrated goal setting to develop the individualized fall prevention plan.
Quality of Evidence
We present the methodological quality of the controlled studies (n = 26) and non-controlled studies (n = 7) in Online Appendices D and E, respectively. Of the 26 controlled studies, performance bias was the most frequent category identified as high risk (n = 17), which is expected given the difficulty in blinding participants and personnel in rehabilitation research. Selection bias was identified in eight of the 26 controlled studies where random sequence generation was not conducted and 12 studies where allocation concealment was not conducted, which is a byproduct of the designs used (e.g., observational studies, case reports). Fifteen studies had a low risk for bias in the blinding of fall outcomes. All the controlled studies had a low risk of bias with respect to short-term outcomes, but long-term data reporting was either at high risk for bias (n = 3), low risk of bias (n = 10), or not included (n = 13). Across all seven non-controlled research studies, participants represented real world patients and eligibility criteria was clearly described. Six of the seven non-controlled studies clearly defined the intervention, and one study used an assessor that was blinded to participant’s intervention exposure.
Discussion
Health systems are incentivized to optimize PAC patient safety through preventing accidental falls while simultaneously promoting recovery after an acute hospitalization. Current evidence uses both an organizational and individual patient-centered approach to PAC fall prevention. Given the care needs of PAC patients, the implementation and delivery of PAC fall prevention programs aims to enhance patient safety, foster functional and medical recovery, and facilitate a safe and effective community care transition. Across the six fall prevention domains, studies addressed: (a) screening patients for fall risk, (b) identifying the multifactorial nature of the patient’s fall risk, and (c) developing and delivering an individualized fall prevention care plan to mitigate risk. While these three overarching approaches are consistent with fall prevention in other settings (Cheng et al., 2018; Tricco et al., 2017), below we highlight some differences in the nuances of the PAC evidence and the role the organizational context plays in the intervention.
Screening for fall risk and evaluating the multifactorial nature of PAC patients’ risk such as environmental hazards, fear of falling, balance deficits, inappropriate or ill-fitting footwear, and safety is consistent with existing community-based research (Chang & Ganz, 2007; Cheng et al., 2018; Maden et al., 2021; Tricco et al., 2017). However, in these prior systematic reviews, an older adult’s fall risk was assessed at a single point in time and intervention delivered based on that assessment, absent of a multidisciplinary team. Alternatively, PAC patients are actively recovering from recent acute hospitalization. Thus, their medical and functional abilities may fluctuate throughout the PAC stay, altering their admission fall risk profile and prevention needs (Burke et al., 2015). Therefore, the PAC team may want to consider how the patients’ medical and functional changes influence the admission fall risk profile and corresponding fall prevention interventions. For example, Aizen et al. (2015) reassessed patients weekly and Barker and colleagues (2015) completed risk assessments each shift to determine current status, modify risk profile, and adjust intervention strategies.
Consistent with efforts to address fall risk factors in the community and other healthcare settings (Cheng et al., 2018; Maden et al., 2021; Tricco et al., 2017), interventions targeted a variety of modifiable and non-modifiable risk factors such as polypharmacy, footwear, vision limitations, functional ability (e.g., mobility, self-care), and the environment. However, the studies in our sample that engaged the PAC staff to deliver the intervention leveraged the expertise of the various members of the PAC care team, including geriatricians, occupational, physical, and speech therapists, social work, and nursing (e.g., Di Monaco et al., 2008; Forrest et al., 2012). Further, given the growing emphasis on fall prevention both during the PAC stay and after discharge, the environmental risk assessment may consider both contexts. While community-based fall prevention interventions often consider the home environment, both the PAC facility and the home environment where the patient will be discharged to were examined in this sample (Goljar et al., 2016; Hopewell et al., 2018; Johnston et al., 2010; Lannin et al., 2007). However, because they were not consistently addressed together in the same intervention, future research may consider both environmental contexts when developing and testing fall prevention programs.
As health systems and researchers advance the efficacy and effectiveness of PAC multifactorial fall prevention, we found intervention design foci expanded beyond the patient and considered the system in which the PAC patient received the multifactorial fall prevention intervention. This approach aligns with the broader implementation research that is driving system change. This methodology emphasizes the importance of taking into account the clinical context (e.g., existing policies, resources), the interdisciplinary team (e.g., staff training, works flow), and communication practices (Titler et al., 2016). More specifically, consistent with Nancarrow et al. (2013) work on effective interdisciplinary teams, our sample described strategies for fostering staff behavior change to implement the intervention, interdisciplinary care coordination, and communication aimed to promote an organizational culture of safety and staff buy-in. For example, establishing clear communication processes among team members to (a) convey a patient’s evolving fall risk and (b) delineate roles and responsibilities related to screening, (re)assessment, and intervention to foster collaboration across the PAC team (e.g., Stenvall et al., 2007). Future PAC fall prevention research may benefit from detailed implementation plans that engage stakeholders and address a multi-pronged approach to provider behavior change and care coordination.
Limitations
As with any study, there are limitations. The search was limited to four databases, and a gray literature search was not conducted. There is the potential for publication bias. While articles of diverse strength were part of this study, included manuscripts had to be published in a peer-reviewed journal, thereby potentially excluding articles that were not published in such journals because of negative results. Finally, while we put processes in place to restrict our sample to studies conducted in the US or international studies that were consistent with a US PAC context, the approaches and findings from other countries may not translate to a US context.
Conclusion
Although falls are preventable, the societal and economic burden of falls is far-reaching (Florence et al., 2018). As health systems strive to prevent falls in their facilities and during the community transition, such initiatives should consider a comprehensive approach that utilizes facility-wide universal strategies and individually tailored approaches that target the patient’s specific fall risk profile.
Supplemental Material
sj-pdf-1-jag-10.1177_07334648221104375 – Supplemental Material for Optimizing Post-Acute Care Patient Safety: A Scoping Review of Multifactorial Fall Prevention Interventions for Older Adults
Supplemental Material sj-pdf-1-jag-10.1177_07334648221104375 for Optimizing Post-Acute Care Patient Safety: A Scoping Review of Multifactorial Fall Prevention Interventions for Older Adults by Natalie E. Leland, Cara Lekovitch, Jenny Martínez, Stephanie Rouc, Patrick Harding and Carin Wong in Journal of Applied Gerontology
Footnotes
Acknowledgments
Thank you to Brenda Covarrubias, Brenda Fagan, and the USC Chan Division Rehabilitation Health Services Research Lab students for their assistance in pulling articles for this review. Thank you to Karin Saric for her guidance in developing the search strategy.
Author Contributions
Natalie Leland contributed to study conception and design, acquisition of data, analysis and interpretation of data, and drafting and finalizing the article. Cara Lekovitch, Jenny Martínez, Carin Wong contributed to study conception and design, data collection, analysis and interpretation of data, and article revision and final approval. Stephanie Rouch and Patrick Harding contributed to the analysis and interpretation of data, drafting and finalizing the article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Agency for Healthcare Research and Quality (K01 HS 022907).
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
