‘Falls not a priority’: insights on discharging older people, admitted to hospital for a fall, back to the community

Additional keywords: community health, falls prevention, hospital discharge, mixed methods, pilot study, quality of life.

Falls among community-dwelling older people are a major public health concern. Hospitalisation resulting from a fall is a frequent occurrence, affecting both older people individually and the health system more broadly. In Australia, falls account for 70% of hospitalisations due to injury among people aged >65 years and 87% among people aged > 85 years (Australian Institute of Health and Welfare 2014). Many advances in understanding causes of falls and effective management strategies have occurred, yet injury-related hospitalisation rates due to falls continue to rise, evidenced by an age-standardised increase of 2% between 2000 and 2010 (Australian Institute of Health and Welfare 2014). Falls often result in serious injury (5–10%) and 92% of hip fractures are attributed to falls (McKay and Anderson 2010). Approximately half of community-dwelling older adults report a fear of falling, which for 40% results in a downward cycle of inactivity, decreased strength and balance (Zijlstra et al. 2007). Falls are often multi-factorial, whereas health systems tend to focus on episodic care of individual diseases (Tinetti et al, 2006); this leads to a system potentially ill-equipped for multiple risk factors and the required range of health professionals.

One in three older people fall within a given year (Tinetti et al. 1988), with a three-fold risk of falling again in subsequent months (Rubenstein and Josephson 2006). The post-hospital discharge period (not necessarily related to a fall) poses an increased risk of falls. Falls rates are higher in the post-discharge period compared with the general older person population, with up to 40% falling within 6 months, and over 50% resulting in injury (Hill et al. 2011a). Thirty-four per cent of older people fall within 3 months of discharge from aged care inpatient rehabilitation (Sherrington et al. 2011). Higher risk of hospital re-admission may place stress on an already over-burdened acute care health system.

Primary healthcare services are well placed to potentially ease the burden on acute care services, by implementing falls prevention guidelines. One study showed a large proportion of general practitioners (GPs), physiotherapists and occupational therapists believed that falls prevention guidelines are feasible to implement in practice (Milisen et al. 2006), but competing priorities and inadequate training potentially limited change in clinical practice (Goodwin et al. 2011). Peer-led programs, thought to be a cost-effective mechanism for falls prevention programs in the community, are currently limited by their lack of training in exercise provision, a key component of falls prevention (Goodwin et al. 2011).

Strong evidence for certain single intervention falls prevention strategies exist, such as provision of exercise (Sherrington et al. 2011) and withdrawal of psychotropic medications (Hill and Wee 2012). Effect of falls prevention education is inconclusive (Gillespie et al. 2012); delivery of an individualised education program reduced falls rates in hospital, but had no sustained effect post-discharge (Hill et al. 2011a). Multifactorial programs based on individual falls risk assessment also reduce fall rates (Gillespie et al. 2012). Participation in, and engagement with, falls prevention strategies by the older person is a critical component. Older people are largely unaware of effective falls prevention strategies they can pursue on their return home (Hill et al. 2011b). Falls risk may not be identified, or they may feel that their current activities are enough to prevent falls (Haines et al. 2014), which may be potentially an issue in the current climate of patient-initiated contact with the primary care sector.

A gap exists in the research related to post-hospital discharge falls prevention strategies, together with an understanding of how the primary healthcare sector can be of benefit. This paper explores relevant falls prevention discharge strategies from the perspective of older people who were hospitalised for a fall and the liaison nurses assisting people to return home.

This paper reports the initial component of a pilot randomised controlled trial (RCT). Fig. 1 outlines the sequential explanatory mixed-method design, with qualitative data used to assess and explain trends of quantitative baseline characteristics (Creswell and Plano Clark 2011). Briefly, the RCT intervention included an exercise program designed and modified by an exercise physiologist, medication review by a pharmacist and provision of falls prevention education, delivered in-home post-discharge. The control arm was usual post-discharge home nursing care provided by a not-for-profit community nursing organisation in metropolitan Melbourne.

Fig. 1.  Flowchart of complete study design (pre-randomisation interviews only reported in this paper).

Semi-structured interviews with older people occurred within 1–2 weeks of return home after hospitalisation for a fall, before randomisation. Questions focused on their perceptions of falls and provision of falls prevention information, and any resultant reduction in confidence and independence (Box 1).

Semi-structured interviews with liaison nurses (nurses based in hospitals facilitating transition from hospital to home with community nursing services) were conducted at their workplace following the recruitment phase. Categories of discussion included recruitment of older people, engaging with the research team and provision of information at time of discharge (Box 1).

Bolton Clarke Human Research Ethics Committee approval was granted (HREC number 150008).

People hospitalised following a fall were identified as potential participants. Inclusion criteria were:

• Aged ≥ 65 years;

• Discharged home from an acute or sub-acute hospital;

• Deemed medically fit; and

• Requiring nursing care or support at home post-discharge.

Exclusion criteria were weight-bearing restrictions or terminal illness.

Potential participants were approached by liaison nurses and informed verbally of the study. If participants provided verbal consent, contact details were forwarded to the study team. Interested persons were telephoned within 1–2 weeks of discharge, with initial appointments subsequently made. Participant Information and Consent Forms were sent to participants before the initial visit. People with cognitive impairment may have coincidentally been part of the recruitment process. Informed consent for those with suspected cognitive impairment was sought by a member of the research team (CM, an experienced allied health clinician and researcher) judging competence according to recommendations from the Dementia Collaborative Research Centre (http://www.dementia.unsw.edu.au/, accessed May 2017).

Baseline interviews were conducted by one of the researchers (CM). All interviews were digitally audio-taped.

• Demographic characteristics

• 1. Age, gender, education level, health conditions, falls history, length of hospitalisation.

• Health-related quality of life

• 1. Assessment of Quality of Life (AQoL) – 8D (Richardson et al. 2012), measured between 0 and 1 (high defined as ≥0.84, low defined as <0.84).

• Fear of falling

• 1. Falls Efficacy Scale – International (Yardley et al. 2005), measured out of 64 (higher scores indicating greater fear of falling).

• Falls risk

• 1. Falls Risk for Older Person – Community (FROP-Com) (Russell et al. 2008) rated out of 60 (higher scores indicating higher risk).

• Functional ability

• 1. Katz Index of Independence in Activities of Daily Living Scale (Wallace and Shelkey 2008).

• Frailty

• 1. Fried Frailty Criteria (weight loss, grip strength, exhaustion, timed walk, activity level) (Fried et al. 2001).

Audio tapes were transcribed verbatim with transcripts reviewed (CM). Two independent reviewers (CM and ER) used open coding initially, guided by the interview questions, to identify recurring themes within the text (Strauss and Corbin 2008). A third reviewer (DG) analysed discrepancies as needed.

Descriptive or frequency analysis was undertaken for all baseline data using SPSS Statistics (ver. 21, IBM Corp., Armonk, NY, USA).

A role-ordered matrix approach was used to provide a synthesis of qualitative themes and baseline quantitative data. This approach allows for sorting data according to a certain set of criteria, reflecting the views of individuals within those criteria (Miles et al. 2014). The chosen criteria were acute v. rehabilitative hospital stay. The next stage was to consider bio-sociodemographic characteristics of support arrangements, injury level and quality of life.

Approximately 60 people were approached by 1 of 10 liaison nurses across nine acute and rehabilitative hospitals. Thirty-seven people agreed to be telephoned, of whom 13 agreed to an initial appointment. Reasons for non-participation were: unable to make contact (n = 6); not interested (n = 5); carer declined (n = 4); too busy (n = 2); worsening condition (n = 2); ineligible (n = 2); and death (n = 2).

Thirteen people participated in baseline interviews (Table 1). Seven (53.8%) were female and six (46.2%) were male, and the average age was 82.2 ± 9.7 years. Approximately 30% had completed junior high school (up to Year 10), identified as Australian ethnicity (84.6%) and were widowed (53.8%). All but one older person lived alone, and the average number of reported health conditions was 5.6 ± 2.7. Quality of life was rated average (0.5 ± 0.1), falls risk was high (22.2 ± 3.5) and fear of falling was average (32.5 ± 6.6). Frailty criteria indicated nine participants were frail and another four were intermediate. The Abbreviated Mental Test Score, as part of the FROP-Com, indicated four (30.8%) participants had cognitive impairment.

Table 1.  Baseline demographic data

Three themes emerged in relation to exploring falls-related discharge planning from the perspective of older people and liaison nurses; they were ‘falls are not a priority’, ‘information not given, or given and not received’ and ‘reduction in confidence and independence’.

Admission to hospital precipitated multiple investigations and interventions related to the possible underlying medical reason for the fall. However, it appeared, from the older persons’ perspective, that other potential falls risk factors were not always given due consideration, with the fall overlooked.

They haven’t tested that leg any further … the rheumatology team came around … but they were more concerned about the heart attack [P7].

The liaison nurses stated that falls are not always given priority during the hospital admission, with reason for admission usually noted to be the medical condition preceding the fall, thus reducing likelihood of further investigation.

…it’s not until you read right through their notes and they’ve got fall as a secondary or a third entry into why it was a part of their admission … they’ll usually put functional decline or [not coping] or something like that. It’s not always identified that they’ve had a fall [L4].

Many participants could not recall being spoken to about falls prevention, despite being admitted to hospital because of a fall.

I don’t know, I can’t remember. They might have but I don’t remember [P8].

The liaison nurses’ comments suggested that the fall and hospital experience may have been overwhelming for the older person, potentially limiting the person’s ability to retain falls prevention information.

It is all just too overwhelming … it’s hard in the hospital situation when they have so many people coming at them from all directions … it is hard to process [L1].

Of importance to the older person was the inability to get up from a fall and how that made them feel. No information had been provided on getting up off the floor following a fall.

I was frantically trying to lift my arm to signal to her through the window, but she couldn’t see me. I lay on the floor for half an hour [P2].

Being hospitalised for a fall provided impetus to contemplate feelings related to the fall and consequences of the fall, specifically a desire to remain confident in their abilities and independence.

At the moment I’m very tentative but I have gone for a walk each day … once you’ve started feeling that you’re weak and you can’t do this or that, well, it just – it affects how you face life and what you do every day [P1].

A lack of confidence was also felt to be linked to other factors, rather than only falls.

It’s probably more because I have the virus [HIV] that I’ve lost confidence, not because of the fall [P11].

The role-ordered matrix approach allowed the views of individual participants to be analysed according to categories of acute v. rehabilitative hospital stay, cognisant of salient characteristics of living arrangements, injury level and quality of life (Miles et al. 2014). This method enabled important distinctions to be examined, relevant to falls prevention discharge strategies. The data synthesis can be seen in Table 2, and reflects the salient characteristics and supporting quotations. A summary of four examples is provided below.

Table 2.  Role-ordered matrix

Lived alone, with limited family support. He sustained a fractured forearm from a fall, which was undiagnosed for 2 weeks. On admission to an acute care hospital, he reflected that he would have sought assistance for the pain earlier if he had known the importance of falls reporting. Quality of life was low, as evidenced by the objective test.

Sustained a fractured wrist from a fall, with a lumbar spine fracture diagnosed several days after discharge from acute care. She lived alone but reported having a very supportive daughter, with quality of life rated as average. Again, there was no recollection of falls prevention information given, with falls seemingly not prioritised by the physiotherapist, just the rehabilitation of the wrist post fracture.

Sustained a fractured humerus and several spinal vertebrae from her fall, resulting in a rehabilitative hospital stay. The participant reported that little falls prevention information was provided despite the longer hospital stay, severe injuries and decreased confidence. Her quality of life was rated as being average. Referral was made to the osteoporosis clinic, but post-discharge physiotherapy focused on exercises for shoulder range of movement rather than balance impairment.

Lived alone, with no family support. Her fall resulted in a fractured ankle and rehabilitative hospital stay, she expressed that exercises were given during rehabilitation, whereas the admission to acute hospital focused only on her silent myocardial infarction. Her confidence deteriorated at home away from the supports of a rehabilitative hospital unit, with quality of life rated as low.

This mixed-methods study has explored the under-researched area of falls prevention discharge strategies for older people hospitalised for a fall. This study was strengthened by triangulation of data, revealing insights from the perspective of both the older person and liaison nurses assisting people to return home. Importantly, this study revealed that both groups felt falls prevention was not prioritised during hospital admission, despite being admitted following a fall. Caution should be applied, however, as treating health professionals were not interviewed for this study. This study group of older people were rated as having either an intermediate or high level of frailty, a high risk of falls and all, but one, were living alone, yet insufficient falls prevention information appeared to be given. On closer examination of the data through a role-ordered matrix synthesis, insights have shown some differences between acute and rehabilitative hospital sectors.

Falls prevention did not appear to be a priority for older people who had sustained a fall, or for health professionals in contact with older people in hospital post-fall. Falls prevention literature supports this finding, with older people stating falls prevention is for others, not themselves (Haines et al. 2014). Priority, particularly in the acute sector, seemed to be given to the acute medical condition on admission. No doubt these medical conditions warrant attention, but falls assessment, according to the American and British Geriatrics Society Falls Prevention Guidelines (Panel on Prevention of Falls in Older Persons et al. 2011), should be considered once pressing medical needs are met. A falls assessment is critical for any older person who has fallen to ensure appropriate strategies are implemented to mitigate further falls (Panel on Prevention of Falls in Older Persons et al. 2011). Sub-analysis of acute v. sub-acute data shows little emphasis on falls assessment in the acute sector, regardless of severity of injury. The rehabilitative sector, however, showed evidence of balance and mobility assessment, and referral to further services. Time and resources potentially differ between sectors, with rehabilitative sectors specifically designed to improve function. There is an imperative from both sectors to ensure a fall triggers prioritised assessment, with simultaneous provision of falls prevention information. However, given the time and resource constraints in hospitals, the primary healthcare sector may be better placed to undertake post-discharge falls prevention. Victoria’s Primary Care Partnerships, a collaboration of hospitals and community healthcare services among others, are well placed to support the co-ordination and integration of services required to assess and manage falls (Day et al. 2014). It is important to note, however, that primary healthcare services currently operate through an older person initiative model, rather than the healthcare provider being the initiator; this being problematic when older people are reluctant to acknowledge falls (Haines et al. 2014).

These data also suggest that falls prevention information was not given to the older person despite them being admitted for a fall. However, it is possible that information was given but not retained. For the few participants with cognitive impairment, retention of the falls prevention message may have been affected due to memory loss, with the intervention component of the study specifically supported through written and verbal reminders. This study did not seek to exclude people with dementia, as many previous falls prevention studies have done (Shaw 2007), given that people with dementia seek to be included in research and have their voice heard (Beuscher and Grando 2009). Liaison nurse data revealed a perception that older people in general may be overwhelmed with information during an acute stay, potentially affecting recall. Other studies have shown education to increase knowledge of falls prevention strategies, but carry-over of information from hospital to community is limited (Hill et al. 2011a). Adoption of recommended strategies does not always occur (Russell et al. 2010), and people may identify falls prevention to be for others, not themselves (Haines et al. 2014). Timing, nature and delivery of falls prevention messages is crucial. Liaison nurses identified education to be potentially more effective on return to the community and routine activities. Most likely, this is more effective for the person with memory concerns, given that familiarity and consistent routines are hallmarks of Australian best practice dementia care (Guideline Adaptation Committee 2016). Appropriate systems are necessary to ensure falls prevention education occurs on discharge, again potentially in the primary healthcare sector once the initial shock of hospitalisation has subsided and the person has returned home.

A reduction in confidence and independence following the fall was evident; however, it was difficult to unpack from many other concerns facing the older person. Decreased confidence was related to co-morbidities, social isolation and unsafe living arrangements, and not necessarily to the fall itself. These findings indicate that interventions to improve confidence and independence on return to the community should be individually tailored for specific concerns and circumstances.

The main limitation of this study was few participants, restricting generalisability of findings. However, as no new themes emerged in later interviews, data saturation had been reached.

Further research related to falls prevention discharge strategies should include:

• A process for ensuring comprehensive assessment of falls risk factors in older people, before returning home from hospital or, alternatively, closer liaison with primary healthcare services as mandated through the Primary Care Partnerships.

• Critical evaluation of the type of falls prevention information (verbal and written) and the timing of information delivery to older people, including delivery of information in the post-discharge period through primary healthcare services.

• Increased awareness in older people and their families of the importance of falls and potential management strategies to reduce likelihood of future falls, including implementation of falls prevention guidelines by primary healthcare practitioners, noting the challenges of an older person-initiated model.

Falls prevention discharge strategies have some crucial gaps. Falls commonly recur post-discharge, with quality of life and functional capacity also deteriorating. The ultimate aim of falls prevention in this context is to prevent recurrence of a fall or, at the very least, avert an injurious fall. Older people are often unaware of falls prevention strategies, or feel that suggested strategies are not for them. Primary healthcare services are critical players in ensuring the needs and preferences of older people related to post-discharge falls prevention are met.

The authors declare that they have no conflicts of interest.