Situating Preventive Action in a Moral and Clinical Context: A Qualitative Synthesis on Fall Prevention

Search

Fifteen databases were systematically searched by an expert librarian in 2016 and 2017: CINAHL, COCHRANE LIBRARY, CUIDATGE, CUIDEN PLUS, EMBASE, ENFISPO, IBECS, LILACS, MEDES, OPEN GREY-Reports, PEDro, PsycINFO, PubMED, TESEO, and Dissertations and Thesis Global. Among the key words and descriptors used to setup the strategy were: falls, fall prevention, accidental falls, nurses, home health aides, home care personnel, nursing care, aged, and senescence. There were no date restrictions. A manual search was performed by two members of the review research team in which widely disseminated and less widely disseminated journals as well as the reference lists of included studies were screened.

Inclusion Criteria

Inclusion criteria were qualitative primary studies about formal care providers (these are: registered nurses, nursing aids, auxiliary nurses, and care assistants) and fall prevention in adults aged 65 years and above in health care facilities (these are: long-term care, hospital, rehabilitation, and acute care centers).

Screening and Data Extraction

Appraising the quality of qualitative studies involves issuing a judgment about their methodological quality, as well as about the nature and extent of the data and whether these are adequate to answer the research question (Sutcliffe et al., 2017). Hence, in the present synthesis, studies were appraised in a continuous process during critical appraisal, before data extraction, and during data analysis. The included studies were screened according to the prespecified criteria of relevance, credibility of findings, methodological coherence, and ethics (Abad-Corpa et al., 2012). Studies were classified per type of analysis into interpretative, descriptive, and exploratory (Sandelowski & Barroso, 2007). All screened studies were, at this point, included to maximize the potential contribution that the papers could make to the synthesis. Later, following close examination for data extraction, studies were excluded due to being mistaken for qualitative research or not containing qualitative findings (Sandelowski & Barroso, 2007), Sift I. During data analysis further studies were excluded due to poor or irrelevant data (Lewin et al., 2015) Sift II, and due to lack of relevance for the emerging analysis, Sift III. A total of 17 studies were included in this synthesis (Supplementary file 1).

Data extraction was conducted using a customized form derived from the main study (Supplementary file 2).

The selection of qualitative studies, their quality appraisal, and data extraction were conducted by members of the research team with competence in qualitative research; they worked in pairs blinded to the appraisal of the other reviewer and consulted a third researcher for discrepancies and to achieve consensus.

Analysis

Qualitative syntheses should transform information coming from primary studies and construct a “connected whole” (Thomas et al., 2017, p. 182). They transcend the sum of the parts; they should generate ideas providing higher and better understanding about the phenomenon under study (Sandelowski & Barroso, 2007; Thorne, 2017, 2019).

We used an interpretative approach for this synthesis. Symbolic interactionism provided the analytical lens. This perspective addresses the importance of meaning to understand people’s actions and the capacity of human beings to make indications to themselves to orient action (Blumer, 1969). According to this perspective, meanings are not given but arise out of peoples’ interactions within specific contexts, and they are interpreted and revised to guide action. Thus, people notice and assess what is presented to them and map lines of action (Blumer, 1969). This theoretical perspective emphasizes people’s agency in the social world and underscores the importance of the conditions under which action–interaction takes place.

A challenge of qualitative synthesis is determining what constitutes a finding. Qualitative research team members took a broad view agreeing that findings were interpretations made by researchers in studies, as well as those studies’ primary data (Thomas et al., 2017). A pair of qualitative research team members carried out the analysis, which was regularly discussed and agreed upon in meetings with the research team. Primary studies were entered into the program NVivo Version 12, 2018 (QSR International) for their findings to be coded. With the aim of preserving the anonymity of sources and putting the analysis into context, each primary study was given a code indicating the type of care provider and the type of health care facility (Supplementary file 3).

Grounded theory procedures of constant comparison, microanalysis, coding, developing memos, and diagrams (Charmaz, 2014; Glaser, 2014; Strauss, 1987) were used to synthesize the findings of primary studies. Analysis developed as a process in four concurrent and overlapping stages. The first stage involved microanalysis, that is careful, line-by-line reading of studies for open coding; categories such as “complex decision,” “continuous assessment,” and “impact of the fall” emerged here as well as some of their properties such as “fluctuating level of risk” and “latent fall risk.” In the second stage emerging categories were organized by grounded theory theoretical codes of “conditions,” “context,” and “phenomenon.” The third stage was focus coding, where categories were developed, and bibliography consulted. Throughout the coding process a constant comparison strategy was used to promote the emergence and development of categories.

Once coding was finished, in the fourth stage, qualitative researchers re-read all the studies to validate the emerging categories and to find relevant data that might have been missed. Throughout these four stages, researchers developed memos with different levels of abstraction. At first, they were very descriptive and focused on codes; as analysis progressed memos became more conceptual describing categories, and in the last stage they established links among these categories.

The analysis ended with the integration of categories around one core category. For this, qualitative researchers compared and sorted memos into main categories. The theoretical code of conditions was used to link these categories. Four categories pertaining to conditions of action–interaction were identified. During this process researchers refined and developed memos further, until they were integrated into a conceptual structure (Glaser, 2014) depicting the conditions that formal care providers consider and assess before executing preventive actions in health care facilities.

Characteristics of Included Studies

Studies were mainly from the United States and Nordic countries. Eight were carried out during the 2010s and only three at the end of the 1990s. Study subjects provided care in hospitals and long-term health care facilities (hereafter referred to as health care facilities) for frail older adults and other people at risk of falls; there were no studies from community centers or noninstitutional care settings. In most studies study subjects were professional nurses, while two studies focused on auxiliary nurses (hereafter both referred to as formal care providers). Generic qualitative studies were the predominant methodological approaches. Two of the studies used interpretative analysis, one used interpretative/descriptive analysis, three studies used a descriptive analysis and one used descriptive/exploratory analysis; the rest were exploratory (Supplementary file 4). All selected studies included direct quotations.

Primary studies were synthesized with the emerging core category of “a complex decision” and described by the emerging conditions that make that decision complex. Each condition is described below according to their properties that were inducted from the studies reviewed. Synthesized findings follow, with a code to identify the cited study (Supplementary file 3).

Permanent Threat of a Fall

Risk factors are ever present in health care facilities and falls are the product of interacting risk factors. Under these circumstances formal care providers strive to prevent falls from happening. Latent fall risk, fluctuating level of risk and interacting risk factors are properties that describe this category.

Fall risks are latent in health care facilities, formal care providers feel that they must be aware of fall risks and always ready to act upon them (AuxH9; Prof-AuxSS5; ProfH-SS7). Older people’s conditions of fragility, chronic illness, and treatments expose them to falls (Prof-AuxSS7). In addition, the physical and social environment present many risk factors for the older person (ProfH6). To the orthopedic devices, technological resources, and architectural barriers present in this physical context, we must add those risks derived from living communally with other patients, what a nurse eloquently describes in a study as “social risk” (ProfH6). However, these risk factors are not fixed, but fluctuate. Patients’ health conditions tend to become worse as time passes. Also, their mood and behavior are not always the same; similarly, their responses to treatments are not either:

Patient status could also fluctuate throughout the day. Therefore, the risk of falling was seen as something valid here and now, but that could easily change. (AuxH9)

Indeed, as primary studies report, the level of fall risk increases in certain situations, such as during shift changes, at meal and hygiene times as well as during peak or busy times (Prof-AuxSS7). Thus, formal care providers meet a fluctuating level of risk of their patients having a fall. In addition to this, fall risks do not act independently but interact between them. They combine in such a way that they create a risk situation. A study describes falls as a product of a synergism and compounding of risk factors related to aging and cohabitation in a health care facility:

An example of synergism of multiple risk factors from our study is the residents’ desire to be independent, combined with the limited availability of staff, in an environment crowded with obstacles that was not adequately designed with older residents in mind. (Prof-AuxSS4)

Older patients’ interactions with others and with the environment create potential situations of risk that formal care providers ought to anticipate and decide upon the best course of action to prevent a fall. This leads to the second condition in formal care providers’ decision contexts.

Continous Flow of Information

The pervasiveness, variability, and instability of risk factors and their interactive nature make falls hard to predict. To cope with this situation, primary studies reveal that formal care providers engage in a continuous process of assessing their patients’ condition, their interactions, and the physical environment. Thus, they are actively immersed in a continuous and variable flow of information, a category that emerged in the present synthesis. Continuous assessment and envisagement of risk situations are the properties that describe this category.

The primary studies reviewed show care providers’ daily work in constantly gathering information:

When LPNs at either ward met patients, they noticed how they walked and acted, and whether they looked ill . . . The LPNs were never completely relaxed and wanted to be near the patients in order to watch them more closely. (AuxH9)

In fall prevention, formal care providers need to be one step ahead as a primary study reported (Prof-AuxSS6) and therefore, require up-to-date information. In the studies reviewed, formal care providers consider themselves to be real time observers (Prof-AuxSS5) on the lookout for signs of heightened risks, such as changes in the patient’s behavior (Prof-AuxSS7), urges for independence (ProfH5), the effects of medication (Prof-AuxSS6; ProfH6), and indications that they are wandering (Prof-AuxSS8). They also pay attention to what relatives say about the older person’s character (ProfFH5).

Formal care providers in their assessments not only notice fall risks but also envision risk situations. Thus, they decipher what they observe and what does not match (Prof-AuxSS5). They interpret the signals from patients and the environment and integrate specific knowledge about related risk factors:

. . . identifying patients who had the potential to fall was not based on just one indicator. One nurse described this process well: “I don’t know. It’s something about the patients you see . . . they’re trying to get off the toilet by themselves . . . you see they are not grasping the information . . . maybe it’s their sitting position or balance . . . maybe they think their abilities are a lot better than they are . . . Those things lead me to sense they are going to fall. It is just knowing.” (ProfH5)

The analysis of findings from primary studies show that fall situations are constructed beforehand by formal care providers. The assessments they make lead them to foresee the development of a risk situation, this constitutes by itself information to guide their preventive actions. As a care provider said in a study: putting “the whole picture together” to be able to do something (ProfH5). Assessment is in fact the daily interpretation of signs and situations that might lead to a fall. In other words, attaching meaning to signs and cues. The information that results feeds into the formal care providers’ decision context.

Lack of Control

Lack of control over results and resources are the properties that emerged from primary studies and describe this category.

Although formal care providers’ interventions are grounded in continuously identifying those at risk and foreseeing risk situations, they do not have certainty beforehand as to whether the chosen action to prevent a fall will yield the expected result. As one study affirms, formal care providers “find themselves continuously struggling with ‘uncertainty’ as they attempt to keep their residents safe from falls” (Prof-AuxSS8).

The studies reviewed show that formal care providers lack control over results, as they do not always have all the information required to make the best decision (Prof-AuxSS7; Prof-AuxSS8) and the information might not be accurate, as risk prediction tools are not without fault. Thus, formal care providers in the studies reviewed commented that these tools have different levels of efficacy (Prof-AuxSS6) and that scales of fall risk have a ceiling effect and are not sufficiently sensitive for some older people (ProfH4; ProfH6). Also, preventive devices such as alarm sensors, whether attached to the bed, to the patient’s wrist or to a rug, are not always able to alert a formal care provider to the risk of a potential fall (Prof-AuxSS6).

More decisively, formal care providers are not always able to anticipate their patients’ responses to a preventive measure. Thus, patients themselves constitute the “unknown factor” (Pro-AuxSS8), their reactions and interactions with the environment are difficult to predict to the extent that a safety measure might turn into a risk of falling, as a study illustrates:

Protecting the patients from fall hazards did not always have the desired results. An auxiliary nurse described a situation where removing a rug from a resident’s room did not necessarily minimise the risk of falling: We had a lady here who had a large rug in her room. . . . we had to send the rug to the cleaners. So, everything was fine for a while and we weren’t afraid that she would fall. But she went around looking for the rug the whole time and she was uneasy and restless because of that. (Prof-AuxSS5)

The fact that a preventive measure might turn out as a risk factor, highlights formal care providers’ lack of control over results and the complexity of decision-making in fall prevention.

In addition, formal care providers in the studies reviewed stated the lack of control they have over resources, as well as the limitations of the environment where they work. In the studies reviewed, practical issues over which they have no decision are repeatedly mentioned: material resources (ProfH6), the layout of the units and institutional policies (ProfH3; ProfH4), and the staff characteristics assigned to their units (Prof-AuxSS6; Prof-AuxSS7; ProfH3; ProfH4; AuxSS1). Formal care providers are aware of their limited room for maneuver:

As nurses we often feel we could have done more to prevent these falls but realize things like staffing, other unexpected occurrences and changing needs of the residents on a daily basis often do not allow extra surveillance and the prevention of falls by our residents. (Prof-AuxSS8)

Furthermore, time is a valuable resource in the care of fragile and dependent persons, over which formal care providers do not have total control. The time that formal care providers have to care for their patients is reduced and compressed, while activities, and hence responsibilities, are added as primary studies reported (ProfH4; ProfH3; AuxSS4). In their eyes, institutions—which promote and implement policies to prevent falls—often fall short in effectively supporting them:

A general sentiment of the participants was that the management did not have much in place to support staff but rather created a lot of hassle when patients fell. (ProfH3)

The present synthesis shows that the practical limitations formal care providers face, in addition to the uncertainty over the results of fall prevention measures, shape the preventive decisions they make.

Ethical Dilemmas and Moral Issues

Intertwined with clinical and practical conditions in fall prevention are the ethical dilemmas, embedded with moral issues; the fourth condition identified in the fall prevention decision context:

The result showed a complex picture grounded in the ethical dilemma personnel face when, in the course of providing care, they must choose between residents’ safety and their rights to integrity and autonomy. (Prof-AuxSS6)

Ethical dilemmas and the consequences of a fall are the properties that describe this category.

Data from reviewed studies show that two dilemmas are present when formal care providers ponder how to prevent a fall. One is whether safety justifies the loss of autonomy of the older person; the other being whether it is legitimate to damage their dignity in the interest of their own safety. The studies about the use of restraints to avoid falls illustrate these dilemmas clearly:

It appears that nurses experienced some ethical conflicts in using restraints on elderly patients. Indeed, the decision to use restraints is not an easy one and sometimes nurses find themselves in a dilemma. (ProfH8)

Safety of other patients and staff should also be considered, as one patient’s safety may turn out to be unsafe for others. Measures such as the use of a walking stick or a walking aide, may become an instrument of aggression toward others (ProfH6). Finding the balance between safety and autonomy, safety, and dignity, is not easy (AuxH9) as the decision to be made will always affect the safety, autonomy, and dignity of a patient:

This descriptive category focuses on dementia and age-related conditions, and comprises the “conceptions,” “forgetfulness,” “anxiety and confusion,” “ability to express oneself and understand,” and “bodily build and function.” The informants highlight the challenge of balancing integrity and autonomy vs. the risk of falling when these circumstances exist. (Prof-AuxSS6; emphasis added)

Formal care providers appraise what is best for each patient and situation, the options they face are often negative, and opt for the lesser of two evils (Prof-AuxSS7). They seek ethically balanced decisions.

Formal care providers feel morally obliged to care for their patients (Prof-AuxSS8), and they are expected to prevent them from falling, “when we addressed the issue of preventing falls, the primary responsibility was clearly placed on the nurse” a study says (ProfH5). When a fall happens, formal care providers feel guilt and self-blame, they feel they have failed as professionals (ProfH5; Prof-AuxSS5). A fall makes them feel angry (AuxH9; Prof-AuxSS7; ProfH5; ProfH6; ProfH4; ProfH1) and morally distressed (Prof-AuxSS8); it is a painful event for them (Prof-AuxSS8).

Through their work, they are witnesses to the “devastating” consequences (ProfH6) that falls have for patients: injuries that isolate them, make them fragile, and increase their pain (ProfH1). As formal care providers, they cannot escape or forget these consequences, since they must deal with them. The possibility of a fall happening has a loop effect for formal care providers (Prof-AuxSS8) and an impact that lasts and will condition what course of action is taken:

They [nurses] stated that it was often difficult to judge when to apply a restraint. They felt that it was the dread of the responsibility that steered them toward putting patients in restraints. (ProfH3)

Falls have consequences for formal care providers. There is a lot at stake for them in fall prevention, “nobody is ever the same after a serious fall,” is asserted in a study (Prof-AuxSS8). Thus, not only patient safety is considered, but also the sense of security of formal care providers:

It appeared that whatever reasons nurses cited for restraining elderly patients, one outcome was that nurses felt a sense of security for themselves. (ProfH8)

Indeed, the potential consequences that a fall might have for formal care providers taint the appraisal that they make; on occasions, they are inclined to act “just in case” (ProfH1) and on other occasions they express ambivalence about the best course of action:

One licensed practice nurse who frequently was assigned to one-to-one observations expressed her ambivalence, “It is hard to keep someone from falling, even if you are within arm’s length, without hurting yourself. It’s your job to guide the person to floor, but it is not always possible without hurting yourself.” (ProfH6; emphasis added)

Therefore, in addition to the ethical dilemmas, moral issues enter the decision-making context of fall prevention.

Limitations

Qualitative syntheses pose challenges (Britten et al., 2017). Among those are the deficiencies in indexing qualitative studies in databases. The manual search of different sources carried out during the present study has mitigated indexing problems in electronic databases; however, it may have left some relevant studies out. In this synthesis we have analyzed the interpretations that other researchers had made of the studies’ data; therefore, it has the limitation of the quality of such interpretations (Britten et al., 2017). Although the detail and the narrative richness of the studies reviewed are considerable, the emerging categories have not achieved saturation, but they have been described to the maximum possible extent.