Critically ill patients' experience of agitation: A qualitative meta-synthesis

Background: Acute episodes of agitation are frequently experienced by patients during critical illness, yet what is not understood is the experience of agitation from the patient and family perspectives. Aims and objectives: To search existing literature, appraise it and then provide a synthesized interpretation to broaden the understanding of patients' and their families' experience of agitation during an adult critical care admission. Design: Qualitative meta-synthesis. Methods: A qualitative meta-synthesis based on a systematic literature search registered with PROSPERO. The search conducted between July and September 2019 was applied to ProQuest, Cumulative Index to Nursing and Allied Health, British Nursing Index, Cochrane Library, Ovid Medline, Web of Science, and PsycINFO databases. We appraised the selected literature and presented a synthesized interpretation. Analysis was based on the approach of Gadamerian hermeneutics. Due to the lack of data identified; the family experiences of agitation could not be addressed within the review. Results: In total, 8 studies were included capturing the experiences of 494 patients, aged between 18 and 92 years, with 225 (45%) women. The analysis generated three core themes: (a) What is real, what is not, (b) loss of communication and dependency, and (c) what helps, what does not. Fear of death, the emotion of anxiety, and feelings of pain alongside transient periods of fluctuating conscious levels provoked a feeling of intense vulnerability. The loss of effective communication and the feeling of dependence incite agitation and distress. Conclusions: The patient's recollection of their critical illness can be completely partially absent and disjointed with uncertainty around what is real what not. Family members observe the full effect of the patient's critical care illness and could be a wealth of untapped information. Relevance to awareness of the critically ill experience highlights possible contributing factors to agitation development,


| BACKGROUND
Agitation, a condition of psychomotor disturbance, is common in the critical care setting, observed in 16% to 52% of the critical care patient population. 1 The patient can show unintentional and purposeless movement stemming from anxiety and accompanied by disorganized thought. 2 An agitated patient can exhibit behaviour such as restlessness or thrashing, placing themselves at risk. 3 The presence of agitation has been linked to the unplanned removal from organ supportive treatments, such as intravenous infusions or ventilatory circuits, higher rates of nosocomial infection, and increased length of both critical care and hospital stay. 2,[4][5][6][7] This reinforces the need for constant observation and vigilance on the part of the healthcare team.
However, when faced with a critically ill, acutely agitated patient, practitioners have limited evidence on which to base decisions about the most appropriate action to take. 8 Additionally, some actions taken may exacerbate agitation rather than reduce its occurrence.
Agitation in critical care remains a complex under-researched health experience. The complexity stems from the fact that there are numerous reasons for agitation and identification of the cause is challenging. The presence of agitation is often linked to delirium; however, agitation is not delirium and agitation without delirium is common. 2 However, the outward manifestation of agitation and hyperactive delirium hold the same clinical challenges. 9 Critical care research has tended to concentrate on delirium, intensive care unit (ICU) syndrome, or post-traumatic stress disorder 1 rather than agitation.
Additionally, over the last decade, sedation practice within critical care has undergone a significant shift to managing patients with lighter sedation regimes. 10 Current evidence and clinical standards support targeted light sedation to decrease the complications related to critical care admission. 11 Yet there has been little consideration of contextual factors and the impact related to the introduction of such a significant change within individual critical care departments. 12 Patients may be more aware of their surroundings, and critical care units are stress-inducing environments with patients reporting feelings of fear, pain, and anxiety, which may induce agitation. 1,11,13 The management of agitation potentially experienced by the patient is, therefore, of paramount importance.
Critically ill patients are difficult to reach population in terms of research as the critical care experience is overwhelming for those affected and, for some, due to illness or impact of sedation, memories of their stay cannot be recollected. 14

| DESIGN AND METHODS
The qualitative meta-synthesis approach was adopted to construct a more detailed narrative without the need to obtain further data from a difficult to reach population. 16 The review team consisted of threeexperienced researchers, two nurses, and one doctor, all with critical care experience and two with experience in qualitative research and conducting literature reviews.

| Methodological framework
The purpose of a meta-synthesis is to offer a coherent description or explanation of a particular phenomenon. 17 The inherent nature of the approach is to facilitate knowledge development. 18 Meta-synthesis is not an accumulation or averaging of results across studies, the aim is patently interpretive. 19 We chose a hermeneutic approach, a process of interpretation. 20 The approach seeks to find a deeper meaning of a phenomena. The concept offered by Gadamer 21 of the hermeneutic circle and the fusion of horizons was used as a frame of reference to guide the meta-synthesis process. The hermeneutic circle refers to the process of repeatedly reading the text while conducting a reflexive dialogue with the text. During this process of exploration, there is a development of an understanding of the phenomena's various aspects, both as they are experienced individually and as they correspond to reveal the whole. 20 We undertook the process of dismantling our understanding, comparing it to other views, and What is known about this topic • The management of a critically ill, acutely agitated patient,   presents a significant challenge and practitioners have   limited evidence on which to base decisions about the   most appropriate action to take. • Additionally, some actions taken may exacerbate agitation rather than ameliorate or reduce its occurrence.
• Agitation in critical care remains a complex underresearched health experience.

What this paper adds
• The management of agitation is a complex health concern, which is under-researched especially from the perspective of the patient and their families. • The occurrence of agitation in critical care requires practitioners to consider ways to humanize care such as improved commutation between patients and staff, involvement in care decision-making, and therapeutic use of touch in highly technical areas such as critical care. reconstructing our understanding in order to produce a new understanding or horizon. The concept of the fusion of horizons offered by Gadamer 21 represents this development of understanding from old horizons to new. We explored the interpretation of the patients' experience of agitation during an adult critical care admission using the interpretations from previous researchers to move our understanding, or horizon, to develop new understanding of experiencing agitation.

| Data generation
The first stage of a meta-synthesis is ensuring the inclusion of studies that are related to the same substantive phenomena (Sandleowski and Barroso, 2007 all studies using qualitative methods for data collection and analysis. Truncation was applied to keywords, "adj" was applied to ensure search terms were adjacent and Boolean operator terms AND/OR plus Medical Subject Headings (MeSH) terms employed including: "critical illness," "critical care," "intensive care units," "experiences," and combination of sets from these categories mentioned earlier. Agitation was not a MeSH term in Medline, however delirium was. Additionally, research into the outward presentation of acute agitation and its management in critical care is under-researched; therefore, the terms delirium and anxiety were also used ( Table 1).
Due to the small body of work in this area, the search was not limited to the abstract or title but across "All text." Grey literature, reviews, editorials, and mixed-method studies were excluded. Studies were included if they referred to the patients' and/or families' experience of agitation and/or delirium during an adult critical care admission. Excluded was any study with a focus on the non-critical care setting, patient population under 18 years of age, care of those living with dementia, or experiences of health care professionals. There was no exclusion based on country of origin; however, non-English text with no available translation was excluded. Within the search, four literature reviews were identified. [24][25][26][27] The reference list of these reviews were explored, and two more studies were identified.

| Study screening
One reviewer performed the initial screening by title and abstract to exclude off-topic papers. The subsequent papers were then screened by two reviewers for potentially relevant text by title and abstract, and if necessary, full text using the inclusion and exclusion criteria. A third reviewer was available to discuss any potential disagreement, but this was not required. The Consolidated Criteria for Reporting Qualitative research checklist was applied to underpin the critical appraisal of the selected studies. One author (S.F.) conducted the critical appraisal and discussed queries with J.Y. and P.D. An overview of the selection process is illustrated in Figure 1. Due to the lack of available qualitative data, the family experiences of agitation could not be addressed as part of the review.

| Data extraction and analysis
The first stage was to read the studies exploring the different voices within the text, the participant, the author, and the review analysts. A qualitative meta-synthesis requires the research team to extract the data presented by each of the studies and separate these from the original interpretations. This process began with a second read-through of the studies and we extracted all raw data, in the form of the presented direct quotes and the researcher's themes from across the studies, maintaining the original wording.
We subsequently created a matrix capturing the first-order constructs, the research participants' original quotes, and secondorder construct in the form of the original researcher's interpretation. 20 This was to ensure the context of the original research was accurately represented. 28 This activity aligned with the ethos of working within the hermeneutic circle, exploring an experience via an understanding of various aspects, where the parts (selected studies) illuminate the whole (phenomena of interest). 20 This also maintained a clear audit trail from the original data to any final interpretative findings. All three researchers had previous professional experience of critical care with two researchers having experience of managing agitation in this clinical setting. Therefore, we directed attention to explore our own preunderstanding of agitation as a distressing episode to experience and observe. Our preunderstanding stemmed from the perspective of ensuring person centred care is enabled. The stance was reflected upon throughout the interpretative process. We interpreted this complete data set constantly moving between first-order construct, second-order constructs, and our preunderstanding. This process resulted in the generation of third-order construct of newly generated analytical themes. We then framed this meta-synthesis in a way that we believed would clarify and illustrate the critically ill patients' experience of agitation to provide an account for clinicians to consider, with the aim of promoting person-centred critical care. Our analysis of the data generated three themes, offering new interpretations of the sample as a whole. 17 For an example of the steps taken using data generated relating to theme 1, see

| Methodological critique of the studies
Appraising qualitative research is controversial as even studies with flaws in methodology can provide valuable information. All the included studies presented their aims and objective. As all included studies were qualitative, approaches to sampling was either purposive or convenience. Five studies did not explicitly state their methodological approach, but this was clear in the description. 2,14,29,30,34 Although purposive sampling was applied, it should be noted that Guttormson's 31 work was secondary analysis of previously generated "My worst nightmare was the light at the end of the tunnel: 'stop breathing and you will get there.' In the other direction was my pregnant daughter was willing me to live. This nightmare has been very disturbing to me ever since." (ID 11) "It was mostly flashbacks from my life where I got younger and younger." (ID 1006) "My son and daughter-in-law came up. I was sent to an apartment where they controlled me with a joystick." (ID 505) "I was visiting some friends. I was lying in the hallway and could see my friends at work, but in reality, I saw the nurses." (ID 387) "I lay in bed and kept sinking-the more I tried to move, the more I sank." (ID 427) "I climbed around in the mountains and had to kill a lot of people to return. It felt so real and frightening. I couldn't get away and I couldn't find the solution." (ID 1151) The  30 "I couldn't speak…terrible…I wanted to explain that I felt like I was dying and didn't get any air, but I couldn't." "I got help…the fact that I got proper help when I was gasping for air…and survived." "When they gave me a small piece of ice to suck on… oh…it was wonderful." "Why did this happen to me? What will happen to me?
Will I live or die?" "I was like a vegetable, without being able to do anything by myself. It made me feel very defenseless and helpless." "It felt so strange…I didn't know if I was alive or dead." "Yes, the worst thing of them all was all the talking… with each other…all the time, it was really annoying and most disturbing…then, just when you have fallen  Third-order  construct  newly  generated  interpretation  for theme 1 asleep, you were roused by loud voices again. Strange, they don't understand that all you want when you're really sick and all, is a little peace and quiet. And why they talk so loud, I don't get it." "All those tubes and lines, irritating, made me kind of crazy…and that thing in my throat, it was nasty." Guttormson 31 "I dreamed I was in a little car and we were shackled down, with our feet and wrist tied. This went on for several days it seemed to me" (ID 3) "you know, I'd wake up. Well like somebody is holding your head underwater or strangulation was a major theme of the scary dreams I had." (ID 39) "I know it's a hallucination, part dream, but at the time you're going through it, to me, it was very real." (ID22) "Now I realise they were bad dreams. Now I realise I was hallucinating." (ID3) "I was locked up and it was like a black cage with a black veil over it. Every time I tried to get out that person would try and stick me with a pitchfork." " The crazy thing now is that this is not possible and cannot be true, the ambulance drops me at the farm and the farmer's wife makes me pray naked." (ID8) "Afraid yes and you didn't understand why, how long did I have to stay imprisoned." (ID6) "I was sitting in a car with a little girl.. I drove off with her then I entered somewhere with her not knowing where I was." (ID20)  15 Clarifying participant representation using identification codes was also difficult as the allocation of participant identification codes was only provided in the studies conducted by Karlsson et al, 34 Guttormson, 31 Van Rompaey et al, 32 Svenningsen et al, 29 and Olsen et al. 14 The majority of studies had two or more independent researchers generating the themes. 14,15,29,32,34 The study conducted by Guttormson 31

| Meta-synthesis findings
In this section, the three developed themes are presented.

| Theme 1: what is real, what is not
In data extracted from six of the eight selected studies, patients recalled experiences linked to the sensations of being drowned, suffocated, or strangled. They used terms such as "terrifying" and "overwhelming," with the resulting feelings of "panic" and "anxiety." 14,[29][30][31][32][33] You know, I'd wake up. Well like somebody is holding your head underwater or strangulations was a major theme of the scary dreams I had. (ID 39 31  The experience of such uncertainty of their reality impacted on the patients' ability to sleep and communicate with the staff. There were also delusions involving the staff leading to patients' distrust, mainly of the nurses.  The anxiety and frustration expressed appeared to stem from an acute sense of dependence and vulnerability.

| Theme 3: what helps, what does not
Data from four of the selected studies did offer insights into strategies that may be helpful to support patients. 14,30,33,34 They also offered reflections of negative behaviours contributing to the feeling of fear, anger, and anxiety.
Three participants across two of the studies expressed the importance of therapeutic touch, hand-holding, feeling cared for, and comforted. 30 34 Potentially, nurses and the wider health care team may not consciously consider the act of simply being present with the patient as a strategy to help alleviate suffering.
The feeling of loss or lack of control has been described in the critical care patient literature. 10,26 Being involved in care decisionmaking when in critical care can make patients feel human and reestablish the feeling of control. 24 One noticeable gap in our data is the involvement of the family in

| CONCLUSION
The main finding of our study is that critically ill patients continue to experience agitation, delusional thoughts, and hallucinations.