Delirium-related distress in hospital: a scoping review and concept analysis

Introduction

Delirium is an acute neurocognitive condition experienced by up to 30% of severely unwell adults and children in hospital.^1,2 Delirium is characterised by acute and fluctuating changes in awareness, attention, and cognition^3,4 and can be caused by numerous medical and environmental factors. Patients can feel distressed during and after the delirium, with common emotions being fear, anger and shame.⁵ Family carers can feel scared for the person’s safety and for their own and uncertain about if or when the delirium will resolve⁶; and delirium care is challenging and complex for nursing, medical and allied health staff.⁷

Distress related to delirium is emerging as an important but under-researched concept. The American Psychological Association defines distress more generally as a negative stress response, occurring where individuals become overwhelmed by stressors (such as perceived threats or demands).⁸ Distress can pose a serious health risk for individuals, causing physical and psychological harms.⁸ A recent narrative review examined distress in delirium for patients, their carers and hospital staff.⁹ The authors used a broad conceptualisation, incorporating related anxiety, depression and post-traumatic stress disorder, and concluded that supportive interventions require more thorough understanding and clinical knowledge of delirium-related distress. Similarly, a systematic review of delirium-related distress in the intensive care unit (ICU) found delirium-related distress was a holistic experience, encompassing “emotional, cognitive, physical, relational, spiritual and situational dimensions” (p. 11).¹⁰ The authors here concluded that a consensus definition would inform the development of outcome measures for future research. Emotional distress is included in new core outcome sets (COS) for delirium in intensive, acute, palliative and residential aged care settings.^11–14 The most recent publication in this suite of work described a consensus study of measurement methods for the COS pertaining to delirium trials in intensive care units (ICUs); in this, consensus on anxiety and depression measures was achieved but not for acute stress and Post-Traumatic Stress Disorder.¹² The authors also noted that distress in delirium went beyond just anxiety and depression.

Thus, to date there is no consensus definition of delirium-related distress, no complete and cohesive understanding of its nature or mediators, and no consensus on its measures. Undertaking a concept analysis of delirium-related distress would build understanding and support the development of a definition, enabling future research to develop specific assessment and outcome measures and interventions. The objective of this scoping review was therefore to review published literature on delirium-related distress using the concept analysis method¹⁵ to develop a cohesive understanding of this concept that informs future research.

Methods

The study was a scoping review¹⁶ and concept analysis,¹⁵ with an a priori protocol registered on the Open Science Framework.¹⁷ The review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR).¹⁸

The scoping review methodology was chosen to enable a comprehensive overview of available literature. Concept analysis was undertaken to clarify the defining characteristics of delirium-related distress by reviewing available literature. Here, ‘attributes’ were the core distinguishing features of delirium-related distress¹⁵; ‘antecedents’ were factors occurring before it¹⁵; and ‘consequences’ were outcomes.¹⁵ Empirical referents, or ways that delirium-related distress is measured, were also identified; as were protective factors, i.e. those that reduced or relieved delirium-related distress.¹⁹Finally, we constructed four cases (model, borderline, related and contrary) to further clarify the concept.¹⁵

Search Strategy

The search was conducted in MEDLINE, Embase, CINAHL, PsycInfo and Scopus. The first search occurred on 19 June 2023 and updated on 14 October 2024. The search terms used for Scopus were delirium; alcohol withdrawal delirium; psychological distress; anxi*; depress*; fear; worry; patient safety; pain; shame; guilt; psychological stress; agitation; restlessness; hospital*; inpatient*; palliative; critical care; intensive care. The search strategy was built with the help of a university librarian and used MeSH (Medical Subject Heading)²⁰ terms wherever possible to include studies that may have used different terminology for the same concepts. For example, MeSH terms were used for delirium and psychological distress in all database searches.

Sources of evidence screening and selection

Qualitative, quantitative and mixed-methods studies were eligible if published in English since 1980 (when delirium was added to the Diagnostic and Statistical Manual²¹) and focused on patient, carer or staff experiences of delirium-related distress in any hospital setting. Screening used Covidence Systematic Review Software. Duplicates were removed. Abstracts were reviewed by the lead author, with 20% double-checked by co-authors.

Data extraction and synthesis

Data were extracted by the lead author into an Excel spreadsheet, which was reviewed by the authorship team. Data extracted included study aims, methods, demographics and results, and the antecedents, consequences, attributes, protective factors and empirical referents of delirium-related distress. Data were synthesised according to the concept analysis headings, including protective factors and empirical referents, and case studies were constructed.

Results

From the 14,243 titles and abstracts screened, 572 had full text review, and 57 articles were included for analysis (Figure 1). Tables 1a, 1b and 1c present the characteristics of included studies, categorised as quantitative (n=23), qualitative (n=27), and mixed methods (n=7) studies. Fourteen studies were published in America,^22–35 13 in the United Kingdom,^{12–14,36–45} seven in Australia,^46–52 five in Canada,^53–57 four in Japan,^58–61 three in Sweden,^62–64 and Denmark,^65–67 two in China,^68,69 Portugal,^70,71 and Poland,^72,73 and one each in India,⁷⁴ Taiwan,⁷⁵ Italy⁷⁶ and Korea.⁷⁷

Figure 1.PRISMA Diagram

Twenty-one studies were conducted in general hospital wards^{13,26,29,30,33,37,40,43,45–47,51,52,56,61,63,64,72,74,77}; 18 in Intensive Care Units (ICU) or with ICU survivors^{12,13,27,38,42,44,53,57,58,62,65,67–71,73,75}; 18 in palliative care and hospice units^{14,22,25,28,31,32,34–36,39,48–50,54,55,59–61}; four in oncology^24,35,48,61; two in geriatric wards^48,50; two in psychiatry^23,48; and one in a specialised delirium unit.⁵²

Overall, there were 6388 participants: 2991 patients (mean age 70 years), 1808 family carers (mean age 58 years), and 1589 hospital staff (mean age 43 years: 72% nurses). Of the studies that reported patients’ primary condition or treatment focus, these included cancer,^{23–25,31,32,34,35,49,59,61} cardiac surgery,⁶³ stroke,⁷² COVID-19²⁶ and dementia.⁵⁶

Of the twenty-three quantitative papers, six compared the outcomes of patients who did and did not develop delirium: three in intensive care unit settings,^53,58,71 two in palliative care or hospice,^22,25 and one in a general hospital ward.²⁶

Key findings on antecedents, attributes, consequences and protective factors of delirium-related distress are described below, and summarised in Table 2 and Figure 2.

Figure 2.Synthesised findings of the concept analysis

Legend: P = patient, FC = family carer, C = clinician

Antecedents

Patients identified a lack of prior knowledge about delirium^40,73,76; and vivid memories or recall of delirium⁵¹ as factors impacting their distress. Carers reported that feeling unprepared for challenging behaviours associated with delirium influenced their own distress.⁵⁶ Hospital staff identified lack of time^48,54; poor understanding of delirium⁵⁰; high care acuity and inadequate patient-to-staff ratios⁴⁸ as antecedents.

Attributes

Attributes mainly reflected psycho-emotional outcomes during delirium. Fear was experienced by up to 73% of patients during delirium-related distress,⁷⁴ and was caused by graphic dreams, hallucinations, and nightmares^{32,34,38,43,47,63,73,74}; agitation and restlessness^31,71; and limited information on delirium.⁶³ Similarly, up to 34% of carers reported fear,^24,53 and worry about caring for the patient.^56,59

Patients described a depressed mood as feelings of loneliness or disconnection^51,57,63; powerlessness^33,51,57; or depersonalisation.^64,76 Carers described depressed mood related to an inability of the patient to recognise them,^40,66 feeling like a burden on the patient and others,⁵⁹ or the experience of witnessing the delirium.^32,52

Anxiety for patients was feeling ‘curious and worried’ about the dying process,³⁴ or concern for the effects of delirium on their families.⁴⁰ For carers, anxiety was related to worry that the delirium was dangerous for the patient.⁶⁶

Stress during an episode of delirium-related distress was identified by up to 50% of carers,²⁹ related to younger age of the carer,⁷¹ delirium duration and severity.³⁷ Carers with higher levels of stress preferred deeper levels of sedation for the patient.³⁵ Hospital staff felt stressed when they could not relieve delirium symptoms or perceived the patient to be uncomfortable,^39,54,55,61 this stress decreased with higher delirium knowledge.⁶⁸ An Australian study found 78% of clinicians (nurses, physicians and pharmacists) gave antipsychotics to patients with delirium with the intent of reducing distress.⁴⁷ A preference for deeper sedation was related to palliative care nurses’ own delirium-related distress.³⁵ Antipsychotics and benzodiazepine use increased nurse’s delirium-related distress if it did not effectively reduce patient distress.³⁹ Nurses in Australia also reported feelings of ambivalence around antipsychotic and benzodiazepine use; feeling the need to over-sedate to minimise family anxiety from witnessing patient’s distressing symptoms, which opposed providing person-centred care and a ‘good death.’⁴⁸

Frustration was common for patients with delirium who perceived they could not communicate with the people during their hallucinations.^43,64

Consequences

Psycho-emotional outcomes were also reported after delirium. Patients showed depressed mood and lower wellbeing and sleep scores compared to patients who did not develop delirium.²⁵ Delirium was an independent risk factor for depression and anxiety after three months,^65,72 PTSD symptoms after twelve months,²⁶ and psychiatric symptoms effecting daily life.⁵⁸

Patients reported feeling fear after delirium.^34,63,67 Patients also reported feeling ashamed,⁶⁴ and a need to apologise for their behaviour.^50,63,64

Anxiety for patients manifested as avoiding falling asleep to prevent re-living hallucinations through dreams,³⁴ avoiding health check-ups for fear they may go back to hospital and experience delirium again,⁵¹ caution when taking medication,⁴⁰ and dismissing the severity of the delirium episode with health professionals and family.⁴³ Carers also reported concern that the patient may have a repeat episode.^27,32,52 Carers felt as though they were a burden on the patient,^33,52,60,75 or felt that they should have done more, especially in instances when the patient died.⁴⁹

Frustration was present for patients who experienced decreased function following delirium.⁶⁷

Regarding longer term psycho-emotional impacts of delirium for patients, a small number of studies examined outcomes at various time points. One survey study in a general hospital ward²⁶ found that delirium was independently associated with worsening 12-month PTSD (aOR = 3.44, 95% CI: 1.89, 6.28) and depression (aOR = 2.18, 95% CI: 1.23, 3.84) symptoms. An observational study in ICU⁵⁸ found that delirium was an independent risk factor for psychiatric symptoms or disability effecting activities of daily life (ADL) at three months (aOR: 3.38; 1.10-10.38) and twelve months (aOR 8.28; 1.48-46.46). A survey study in ICU⁶⁵ found no association between delirium and PTSD symptoms, anxiety or depression at two and six months post ICU discharge. Finally, a longitudinal study in a hospital ward involving patients with stroke⁷² found that delirium was an independent risk factor for anxiety three months post-stroke (OR = 2.83, 95%CI 1.25–6.39, p = 0.012), and for apathy after three (OR = 3.84, 95%CI 1.31–11.21, p = 0.014) and 12 months (OR = 4.95, 95%CI 1.68–14.54, p = 0.004) post-stroke.

Protective Factors.

Feeling safe was a protective factor against delirium-related distress. For example, having family members or friends present who could provide reassurance, increased feelings of safety^51,55–57 and comfort.^39,69,73 However, while often beneficial to the patient’s emotional wellbeing, family and friends reported witnessing delirium could be stressful.³⁹ Nurses also contributed to patients feeling safer, with some patients reporting memories of kindness from nurses.⁴³

Patients reported that having an opportunity to talk about delirium after an episode was beneficial,^57,73 as was reassurance that their hallucinations were not real,⁵⁷ and when carers and nurses were supportive and empathetic.⁷³ Talking to nurses and other hospital staff however was only helpful for patients if they had not been a feature of their hallucinations.⁴² Carers also reported receiving emotional support from nurses and other hospital staff would have reduced their distress.⁶⁶

Empirical Referents

Empirical referents identified by this review were the Delirium Experience Questionnaire,²³ Delirium-O-Meter,⁷⁶ and Delirium Burden Questionnaire.^30,46 Levels of distress for patients, family carers and staff were variable,^{23,31,35,70,71,74,76} and in the three studies that reported the proportion of patients with delirium who experienced no distress, this ranged from none to one quarter of patients.^31,70,74 See Table 3 for a summary of these measures.

A series of studies from the Del-COrS group^12–14 aimed to determine a core outcome set (COS) for delirium. Authors identified emotional distress as a core outcome important to survivors of delirium and their families, with emotional wellbeing rated as 97% critical.¹⁴ A follow-up consensus study¹² confirmed that the Hospital Anxiety and Depression Scale (HADS) was useful to measure anxiety and depression, but not stress or post-traumatic stress symptoms; and that emotional distress should be measured up to 12 months post-delirium; however, the intervals for measurement were not agreed upon.

Constructed Cases

Four constructed cases are presented in Table 4: a model case highlighting defining attributes of delirium-related distress; a related case which does not include all defining attributes; a borderline case which closely resembles delirium-related distress but is not immediately evident as such; and a contrary case which provides divergent attributes not consistent with delirium-related distress.

Discussion

The objective of this scoping review and concept analysis was to develop a cohesive understanding of delirium-related distress to inform future research. Psychological distress, as defined by the American Psychological Association, is “a set of painful mental and physical symptoms…”.⁷⁸ The Del-COrS characterised emotional distress as “including anxiety, depression, acute stress and posttraumatic stress disorder.” (p. 1542)¹³ Our results indicate that a broader and more nuanced definition for delirium-related distress is required. For example, we found that delirium-related distress involved aspects of depression, anxiety, stress, and frustration, with long lasting effects. Of note, earlier definitions lack fear, which in our review was found to be a common attribute and consequence of delirium-related distress. We have also shown that lack of prior knowledge of delirium and stressful hospital environments can worsen related distress.

Importantly, family and friends brought comfort and feelings of safety to the patient. However, family visiting could also be distressing when relationships were strained or patient’s behaviour changed.³⁹ Similarly, while the presence of nurses could be increase patients’ feelings of safety, this was not the case when nurses had been a feature of the patient’s hallucinations.⁴² Our study highlights the value of engagement with families and listening to patients, as patients reported that they valued supportive staff and being able to talk about their experience. This aligns with previous research highlighting the importance of kindness and empathy for patients and their families.¹⁰

Empirical referents reported in the results were measurement tools focused on aspects of delirium, not specifically on the aspects of delirium-related distress discussed in this study. The Delirium-O-Meter⁷⁶ was the only measure included identified in this review with an item related to delirium-related distress; namely, ‘anxiety/fear’ in the carer version. Based on the findings of our review, future measures of delirium-related distress could include domains such as fear, depression, stress, anxiety and frustration.

The findings of this concept analysis echo a previous systematic review and meta-synthesis in the ICU setting,¹⁰ in which delirium-related distress was described as a whole person experience, with elements of distress being emotional, cognitive, physical and spiritual. Our study extends this finding to the broader hospital setting.

Our review draws attention to subsequent healthcare avoidance and health anxiety, such as avoiding General Practitioner visits and caution taking medication. These behaviours are reported in cancer⁷⁹ and select mental health conditions⁸⁰ but less are recognised in delirium and warrant future study.

Strengths and limitations

This scoping review and concept analysis is limited by the lack of antecedents and consequences in the literature. Antecedents were most notably lacking for family carers; and consequences for all groups, as the included studies mostly focused on the immediate emotional experience. The literature on protective factors was scarce for all groups. No studies involving children were found. Further work is needed to clarify these gaps. Another limitation is that this review did not examine cultural differences in the experience of delirium-related distress; as it might manifest and be described differently in different cultures,⁸¹ this could be a focus of future research. Similarly, we did not quantitively compare differences between patients with delirium who did and did not experience distress. Quantitative observational studies of the protective factors would inform new interventions.

A strength of this scoping review was its broad inclusion of literature, while the concept analysis method allowed for a structured and in-depth exploration of delirium-related distress. However, the breadth of this review is perhaps countered by less detailed analysis. Some included papers provided only summarised brief descriptions of delirium-related distress and lack of primary data prevented deeper analysis.

In the current review, only 20% of full text papers were screened by at least two authors. Due to the large amount of included literature, as well as limited frame for the review which is part of a doctoral program (KC), it was not possible to double screen every full text paper, which may have increased the risk of random error in the screening process.

Conclusions

This scoping review and concept analysis synthesised 57 studies relevant to delirium-related distress for patients, carers and staff across a range of countries and hospital settings. Antecedents included aspects of the delirium and influence of the hospital environment. Attributes identified were largely psycho-emotional in nature, and consequences involved psycho-emotional effects and health avoidance. Feeling safe through the presence of family, friends and nurses and other hospital staff was identified as a protective factor for delirium-related distress. Future research could attend to building the evidence base for antecedents, consequences and protective factors of delirium-related distress by applying concept analysis methodology to direct accounts from patients, carers and hospital staff. Further studies are required to develop and measure new interventions for delirium-related distress in hospital populations.

Acknowledgements

The team wish to acknowledge the valuable contributions of Sarah Su, Senior librarian at UTS.

Ethics statement

N/A

Funding sources

KC is a recipient of the Research Training Program (RTP) scholarship, provided by the Australian Government Department of Education through the University of Technology Sydney.

Declarations of interests

None

Author contributions

KC – Conceptualisation, methodology, investigation, analysis, writing – original draft, visualisation

MD - Conceptualisation, methodology, validation, writing – review & editing

DD - Conceptualisation, methodology, validation, writing – review & editing

AH - Conceptualisation, methodology, validation, writing – review & editing