Design

A qualitative descriptive approach was used [12] with study data collected using semi-structured interviews and analysed using inductive thematic analysis. Ethics approval for the conduct of this study was received from the Human Research Ethics Committee at Eastern Health (E20-023-69030). This study is reported according to the Consolidated Criteria for Reporting Qualitative Research [13].

Setting

The setting for this study was one 32 bed Geriatric Evaluation and Management ward from a health service in Melbourne, Australia. The nurse-to-patient ratios on the study ward were one nurse to five patients on a morning shift, one nurse to six patients on an afternoon shift and one nurse to ten patients overnight. There were 43 nurses on the study ward roster. Of these 16% were registered nurses on morning shifts, 14% on evening shifts and 12.5% overnight, and the remainder were enrolled nurses. Registered nurses are Bachelor prepared and enrolled nurses are Diploma prepared and work under the supervision of registered nurses. The average length of patient stay on the study ward was 21.2 days.

Falls data from 1 July 2019–30 June 2020 provided context for this study. During this time there were 222 falls in 110 patients: the median number of falls per patient was one (range 1–24 falls per patient). At the patient level, 52.7% (n = 58) were females and the median age was 86 years (interquartile range 80 to 89). The median subacute care length of stay in total was 22 days (interquartile range 14 to 40 days). Of the 110 patients in this study, 2 died during their subacute care admission. The most common location of falls was patient rooms (72.1%, n = 160), the most common activity at the time of the fall was standing or walking (25.2%, n = 56) and two-thirds of falls were unwitnessed (64%, n = 143). Various falls prevention strategies were in place at the time of the fall, including falls risk alert sign over the bed (90.1%, n = 200), direct bathroom supervision (82.9%, n = 184), falls alarm (61.7%, n = 137) and lo-lo bed (hospital bed that can be lowered to floor level) (30.6%, n = 68). Most falls (96.8%, n = 213) resulted in no or minimal physical harm.

Participants

Purposive sampling was used to recruit participants. All nursing staff permanently employed on the study ward were invited to participate in a semi-structured interview during ward meetings and in-service education sessions. Study recruitment was managed by one researcher (JC) who had no line management or operational responsibilities. Written informed consent was obtained from all participants.

Data collection

All interviews were conducted by one female researcher, who is a PhD-qualified registered nurse with experience in qualitative research and clinician interviews (JC), between 27 January and 12 March 2021. The interviewer holds a joint academic-industry professorial appointment at the same health service but has no direct-line management or clinical interaction with the participants. Eleven interviews were face-to-face, and the remaining interview was conducted by telephone: they were audio-recorded and ranged from 11 to 18 minutes in duration. The semi-structured interview guide was developed following literature review and consultation with the research team, health service Comprehensive Care Clinical Risk Governance Committee, and health service consumer representatives. The interview guide is presented in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Interview guide.

https://doi.org/10.1371/journal.pone.0287537.t001

Data were transcribed and coded following each interview. At the conclusion of every third interview, the codes were reviewed to determine whether new information was being obtained, or whether information obtained from preceding interviews was confirmed, thus indicating data saturation [14]. Data saturation was apparent after ten interviews so two more interviews were undertaken to confirm that no new information was evolving. This methodological decision was based on a study suggesting that data saturation was achieved within 12 interviews when participants were asked the same interview questions [15].

Data analysis

The interviews were transcribed verbatim using a professional transcribing service. Inductive thematic analysis was undertaken using six steps from the framework developed by Braun and Clarke [16,17]: familiarisation with the data; generating initial codes; searching for themes; reviewing themes; defining and naming themes and producing the report. A single researcher (JC) checked the transcripts for accuracy against the audio files, entered them into NVIVO software, and after close reading and re-reading, performed the initial coding. An open coding process was used, so codes were not determined a priori, but developed and modified during the coding process [16,17]. A second researcher (DB) reviewed the codes and through an iterative process and discussions, subthemes and themes were identified from the data. Differences of opinion during initial coding were resolved by re-reading the interview transcripts, listening to the interviews and discussion. The themes, subthemes, codes, and transcript content for each code were organised into tables, circulated to the research team, and discussed at fortnightly to monthly team meetings where themes and subthemes were refined to the final versions presented in this paper.

The tenants of rigor of qualitative research are credibility, transferability, dependability, and confirmability leading to trustworthiness [18]. The systematic development of the interview guides established credibility, and dependability was obtained by the sound methodological approach undertaken. Conformability was established by using examples from the interviews to ensure that nurses’ voices were represented. A reflexive approach to thematic analysis requires researchers to question their assumptions, highlights researchers’ skills as resources, and requires researchers’ reflexive engagement with the data in its interpretation [17]. The research team was diverse and comprised nursing researchers, experts in subacute care nursing, ward and executive management, a clinical risk manager, and a consumer. Each member of the team therefore brought a specific perspective to the conversations regarding data analysis and theme and subtheme generation. Given the pragmatic nature of this work, the peer-to-peer conversations and meeting notes served a similar purpose to that of a reflexive journal, and enabled tracking of methodological discussions and decisions. Member checking (return of transcripts to participants) did not occur to decrease participant burden. Example interview data and their respective codes, subthemes and themes are shown in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Examples of codes, subthemes and themes.

https://doi.org/10.1371/journal.pone.0287537.t002

Theme 1: Negative impacts of falls alarms

Nurses’ negative experiences of falls alarms were related to noise and technological limitations. Nurses’ concerns and frustrations with the noise from falls alarms were multifaceted. The noise disturbed the patients in whom the falls alarm was in place, as well as other patients on the ward “…the person has to hear the noise, and everyone else has to hear the noise (EN 11)”. Nurses described sleep disturbance as a major issue and patients perceiving the noise as ‘high pitched’ and ‘horrible’. There was also additional frustration from being unable to moderate the noise without physically having to disarm the alarm at the bedside: “…the alarm can’t be remotely monitored…” (Nurse 11—RN) and “… you can’t make it stop until someone goes there and stops it” (Nurse 12—RN).

The technological limitations of falls alarms described by were many nurses. False alarms were commonly triggered by slight patient movement such as “…shifting in their seat…” (Nurse 1—EN), ‘…wriggling in bed…” (Nurse 2—ANUM) or “…simply breathing…” (Nurse 11 –EN). The sensitivity of the falls alarms, high number of false alarms, and noise resulted in both nurses and patients turning them off: “…sometimes I know they [nurses]… turn it off for the meantime… you’re not supposed to” (Nurse 5—ANUM) and “…often the patients …stand up by themselves and turn them off. (Nurse 2—RN). Nurses also described problems with the falls alarm pagers. There were days where there were not “…enough pagers to go around” (Nurse 9—EN) so nurses left them at the nurses’ station, hoping someone would notice when one went off. There were mismatches where the falls alarm would activate but the pager did not, and known issues related to pagers not working when they were too far away from the alarm: “…the pagers are only set up for about 50 metres” (Nurse 2—ANUM).

Theme 2: Juggling the safety-risk conflict

Nurses described tension between using falls alarms to decrease the risk of a fall and falls alarms creating other risks. Falls alarms triggering agitation, particularly in patients with cognitive impairment who could not make sense of the noise, was a common experience. Nurses spoke about the noise from the falls alarm “…panicking them [patients]…” (Nurse 6 –RN), “…triggering behaviours…” (Nurse 6—RN), and “…promoting aggravation, agitation…” (Nurse 10—EN) thus the falls alarms were viewed as harmful: “…making them [patients] worse … irritating them [patients] more…” (Nurse 8 –RN). The resultant patient distress and agitation was perceived to increase the risk of falling, rather than reducing falls risk as intended:

…they’ve got delirium or dementia and are so anxious about that noise. That, to me is going to promote a fall rather than them just standing up and going for a walk (Nurse 12 –RN)

I think it [falls alarm] makes them more of a falls risk because they panic when they hear this beeping going off… they try and reach it to turn it off so it makes it a greater falls …. In those circumstances, I think they’re safer without it [falls alarm] (Nurse 9 –EN)

The goal of care for patients on the study ward was rehabilitation and promoting independence in preparation for discharge. The use of falls alarms in these patients was counterintuitive and perceived as either harmful or ineffective by nurses. Nurses described that part of preparation for discharge was to allow necessary risk taking “…we need to allow them to take a certain amount of risks … we really need to take the alarm away” (Nurse 9 –EN) as patients would not have a falls alarm at home. Further, one of the unintended consequences of falls alarms was reduced mobility as patients were “…not able to even move side to side or reposition without an alarm going off” (Nurse 10 –EN) which was in direct opposition to the goals of reconditioning and increasing independence. Many nurses acknowledged that patients with falls alarms in place, still experienced falls “…if they’re going to fall, the bed/chair alarm isn’t going to protect them … it’s not a protection, it’s just a warning (Nurse 11 –RN).

The need to respond to falls alarms frequently interrupted nursing care and nurses described the stress of balancing competing priorities and clinical risks. Several nurses described their experience of attending to intimate patient care for one patient and hearing a falls alarm activated for another patient. There was no one course of action that was not distressing for nurses. They could leave the patient they were attending to respond to the falls alarm, resulting in clinical risk and indignity for the patient:

Mentally, that’s quite challenging… knowing that I’ve left my buddy standing there holding a patient that’s nearly naked with a towel over them while I’m trying to answer another bed chair alarm. (Nurse 12 –RN)

Alternatively, nurses could stay with the patient for whom they are caring and wonder what was happening to the patient in whom the alarm was triggered:

You might be in the bathroom attending to somebody. You can’t leave them to go and answer a bed/chair alarm that’s going off… you can’t leave that person… they could fall over. (Nurse 2 –ANUM)

Several nurses perceived that falls alarms created clinical risk because whilst nurses were responding to falls alarms (many of which were false alarms), other patients missed nursing care. Nurses felt that “… the time we spend running to the alarms, we can actually spend with the patients” (Nurse 6 –RN) and “…we have more people on alarms than we have nurses that can manage them” (Nurse 9—EN). As a result, patients with falls alarms received more attention at the detriment of other patients, or the time spent responding to falls alarms meant that missed care had to be handover over, both of which invoked feelings of guilt. The other source of clinical risk reported was interruptions to safety-critical tasks like medication management: “… when you run to another patient …. you cannot remember which medications you have dispensed (Nurse 5 –ANUM).

Theme 3: Negotiating falls alarm use

Falls alarm use was highly dependent on nurses’ decision making and many influences on nurses’ decisions regarding falls alarm use, such as years of experience, staffing levels, patient characteristics, family expectations and organisational policy, were associated with alarm overuse. There was a view that inexperienced nurses had a lower threshold for using a falls alarm “… our new grads [graduate nurses] think all the confused patient has to be on a bed chair alarm” (Nurse 3—ANUM) and that experienced nurses were more willing to “…give them [patients] a chance” (Nurse 10—EN). There was also a sense that falls alarms were warranted overnight to compensate for reduced staffing: “…[we need falls alarms] especially at night … there’s only three staff on” (Nurse 2—ANUM).

Patient characteristics were an important influence on nurses’ decisions to implement falls alarms. Nurses reported they were more likely to implement falls alarms if patients were “…unsteady on their feet … not pressing the bell…” (Nurse 8 –RN), had cognitive impairment, or were unable to remember and follow instructions such as using the call bell or not getting up unattended: “…it’s a combination of cognition, understanding instructions, and being able to do that safely… (Nurse 11—RN). A converse view was held by some nurses who considered it was better to manage falls risk in other ways, such as allowing safe wandering: “… if they’re confused and want to wander in their room… you take the bed/chair alarms off …” (Nurse 2—ANUM). Family expectations often swayed nurses to use falls alarms and nurses felt that alarms were evidence to families that falls preventions strategies were in place:

There’s an expectation … I tell the family we are trying our best to keep mum in a safe environment …if something happens to her despite of all the preventive equipment that we have in place, they are more understanding (Nurse 3—ANUM)

Nurses talked about different levels of decision-making autonomy and the decisions to implement a falls alarm was seen as easier than the decision to remove the device:

I can actually implement putting one on if there’s a need. I don’t feel as in control to actually say no, I don’t want one [a falls alarm] if someone is on one already … to remove it, I don’t think there’s as much autonomy (Nurse 8—RN)

Similarly, nurses reported ‘inheriting’ alarm use in patients arriving from other areas: “…from an acute hospital or sometimes from palliative care and they’ll be on a bed chair alarm” (Nurse 9—EN). Feeding into nurses’ perceived lack of decision making autonomy was the influence of organisational policy on their decisions to use or not use falls alarms. There were perceptions of reactive alarms use in response to a fall, particularly for falls with a clear explanation, and that falls alarms were a ‘box ticking’ exercise: “…because she’d a fall, they decided that they’d tick the box” (Nurse 12—RN) used for medico-legal reasons “…probably a legal thing that nobody wants to be sued” (Nurse 9—EN).

Conclusion

Nurses’ experience of using falls alarms was predominantly negative and there was tension between falls alarms having limited impact on patient safety and risks associated with their use. Nurses described a need to support nurse decision making related to falls alarms use in practice and policy, and a desire to be empowered to manage falls risk in other ways.