Background

Delirium is an acute and complex neuropsychiatric disorder that commonly affects hospitalized patients [1]. Delirium is the most common neuropsychiatric disorder in acute care, with prevalence estimates of 10–30% within the hospitalized population [2, 3]. Patients with delirium are often hospitalized for longer and have a poorer survival prognosis than those without delirium, and the risk of developing delirium generally increases with the severity of illness [4]. Identifying and implementing effective strategies to mitigate the risk of delirium are essential to reducing long-term morbidity related to critical illness [5].

Delirium and other neuropsychiatric disorders are often similar in presentation, often present concurrently, and patients may have more than one type of disorder [6]. There is difficulty distinguishing these disorders: for example, severe hypoactive delirium [7] can be confused with depression, and hyperactive delirium [8] can be confused with mania. Patients should be evaluated for delirium and psychiatric disorders to not to miss important medical problems; a lack of recognition for pre-existing and new psychiatric disorders during an acute care admission may contribute to poor patient mental health and increased severity of psychiatric disorder symptoms [9,10,11].

A Clinical Practice Guideline (CPG) is a methodological statement aimed at providing guidance to clinicians and their patients for specific medical circumstances and conditions [12]. CPG use can support the reduction of financial cost from inappropriate care and can improve clinical decision-making and quality of care for clinicians and patients [13]. A recent systematic review of the quality of CPGs for delirium in acute care evaluated the quality of guideline recommendations focusing on knowledge translation resources and the practical application and monitoring of guideline implementation [14]. The most recent included CPG in this review was published in 2013 thereby necessitating an updated review to identify progress and/or gaps in our evidence base on this topic.

The first objective of our systematic review is to add an analysis, synthesis, and quality assessment of the available CPGs on recommendations for the identification and management of anxiety, depression, and PTSD in adults with delirium in acute care. The second objective is to provide an updated synthesis and quality assessment of CPG recommendations for delirium identification and management in acute care.

Methods

This systematic review was registered on PROSPERO (CRD42021237056) prior to data abstraction and reported using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [15] (Additional file 1: Appendix 1).

Identification and selection of studies

We performed systematic searches of MEDLINE, EMBASE, PsycINFO and CINAHL from inception to February 12, 2021, to identify eligible CPGs. We additionally performed a comprehensive search of the grey literature using the 21 sites listed on the Canadian Agency for Drugs and Technologies in Health (CADTH) Grey Matters Lite tool [16] from inception to February 12, 2021. Search strategies for each database are included in Additional file 2: Appendix 2. A librarian (N.D.) performed an independent PRESS review [17] of the EMBASE search strategy. CPG reference lists were screened for additional guidelines relevant to the review which may have been missed by the search. No limits (e.g., date, language) were applied to any search.

Study eligibility

Two reviewers (T.G.P., S.J.M.) assessed record title and abstract eligibility independently and in duplicate. Guidelines eligible for inclusion were written in English, were issued or endorsed by a national or international scientific society or government organization and had a primary focus on the diagnosis and/or management of delirium in any acute care setting. Guidelines eligible for inclusion contained at least one recommendation on the diagnosis, prevention, or management of delirium presented within guideline text, tables, figures, algorithms, and/or decisions paths. Guideline recommendation(s) on delirium must: (1) be accompanied with an explicit level of confidence (i.e., the GRADE system [18]); and (2) explicitly discuss one or more interventions for the recognition or management of delirium. Comparison of these recommended interventions to other interventions was not required. Guidelines with additional recommendations pertaining to anxiety, depression, or PTSD were of special interest. Any year of publication, publishing region and guideline development process were of interest. The most updated versions of guidelines were included in the review. Title and abstracts were advanced for full-text review if both reviewers agreed independently and in duplicate that they satisfied one or more of the eligibility criteria. Full-text guidelines were included in the review if both reviewers agreed independently and in duplicate that they met all the criteria for inclusion. Discrepancies were handled through discussion with a third reviewer (K.D.K.).

Data extraction and synthesis

Data extraction was completed independently and in duplicate by two reviewers (T.G.P. and S.J.M.). Data extracted from included guidelines consisted of guideline name, author(s), development group, country, language(s), target population(s), evidence consensus method(s) and psychiatric disorder(s). Narrative synthesis of recommendations for diagnosis and management of delirium and of psychiatric disorders was completed after data extraction for relevant guidelines. Recommendations for diagnosis and management of any symptoms of anxiety, depression, and PTSD in delirium were included, in addition to recommendations for the management of pre-existing or new diagnoses of these disorders.

Guideline quality assessment using AGREE II

The Appraisal of Guidelines for Research and Evaluation II (AGREE II) Instrument [19] is designed to assess the methodological quality and reporting foundation of a CPG. The AGREE II tool is composed of a total of 23 items in 6 Domains and 2 overall global rating items. All items are rated on a seven-point scale (1 = no information or poorly reported, 7 = reporting quality is exceptional and meets all criteria); except for the second global rating item which asks if the rater would recommend the guideline for use (yes, yes with modification, or no). The six Domains include: Domain 1, Scope and Purpose (assesses the overall guideline goal and target population as well as the health questions); Domain 2, Stakeholder Involvement (targets the participants involved in the guideline development group and how the guideline represents the perspectives of users); Domain 3, Rigor of Development (assesses how the evidence was gathered, expressed, and how it will be updated); Domain 4, Clarity of Presentation (reviews guideline organization and language); Domain 5, Applicability (assesses if implementation is feasible, the economic consequences, and if there are specific strategies for implementation); and Domain 6, Editorial Independence (measures the level of independence from funding institutions and competing interests of guideline developers). Two reviewers (T.G.P. and S.J.M.) scored all included guidelines independently and in duplicate, and any discrepancies were resolved through discussion. As suggested by the AGREE II tool [19], six scaled Domain scores for each guideline were calculated by summing all individual scores in the Domain and scaling as a percentage of the maximum possible score for that Domain. For each guideline, the average overall score was determined by taking the mean of the first overall global rating item (“Rate the overall quality of this guideline from 1 to 7”) after both reviewers assigned it a score. For the overall assessment, reviewers used criteria from all Domains to judge if they would recommend the guideline for use using “yes”, “yes, with modification”, or “no”.

Results

Results of search

Initial database search, grey literature search and reference list screening resulted in a total of 10,774 records (Fig. 1). After duplicates were removed, 7611 records were screened by title and abstract, and 7314 were excluded. Full-text review was performed on 197 records and following this, 170 records were excluded due to: not being a full guideline (e.g., not providing specific sections such as references) (n = 149), not being endorsed by a national or international scientific society or government organization (n = 12), not available in English (n = 2), not focused on delirium (n = 2), full-text unavailable (n = 1), and being a duplicate (n = 4). The final review included 27 CPGs.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Study characteristics

Guideline characteristics are presented in Table 1. Included guidelines were published from 1999 to 2020, with two guidelines being updates of previous versions: Jacobi et al. published in 2002, updated from 1994 [20], and Devlin et al. published in 2018, updated from 2013 [21]. The most recent versions of guidelines were included in the review. Countries where guidelines were developed included Canada (n = 10), United States of America (n = 7), United Kingdom (n = 7), Australia (n = 1), Europe (n = 1), Germany (n = 1), India (n = 1), and Switzerland (n = 1). The most common guideline target populations were healthcare professionals working in general (e.g., general hospital) acute care settings (n = 12; 44%) and ICU healthcare professionals (n = 4; 15%). Informal consensus (i.e., discussion) was the most commonly used consensus method during guideline development (n = 12; 44%), however, this was often employed alongside formal tools such as GRADE [18], AGREE [19], or the development method and guideline classification schemes described by Shekelle et al. [22]. While all (n = 27) guidelines focused on delirium, 20 guidelines (74%) included recommendations on delirium only, whereas seven guideline (26%) provided additional recommendations for anxiety, depression, and PTSD in delirium.

Table 1 Summary of included guideline characteristics
Full size table

Quality of included guidelines

Individual and cumulative guideline Domain scores assessed by the AGREE II tool are presented in Table 2. The lowest cumulative Domain was Domain 6, Editorial Independence, with a mode of 4% and a range of 4–92%. The highest cumulative Domain was Domain 4, Clarity of Presentation, with a mode of 92% and a range of 47–100%. The guideline by McNeill et al. [23] scored the highest in all Domains compared to all other guidelines: Domains 1–4 and 6 scored 92% while Domain 5 scored 71%. In addition, this guideline received an overall score of 6, and both reviewers recommended the guideline for use.

Table 2 AGREE II domain percentages for guidelines containing recommendations for delirium or psychiatric disorders in delirium
Full size table

Summary of recommendations for delirium

A synthesis of recommendations for delirium stratified by recognition and prevention is presented in Table 3. Specific sub-populations more at risk for delirium were mentioned in fewer than 60% of included guidelines: approximately half (n = 14; 52%) of included guidelines were designed to recognize and prevent post-operative delirium; recommendations specific to older adults (defined as > 65 years of age) were presented in 15 guidelines (56%).

Table 3 Synthesis of recommendations for delirium. Refer to Additional file 3: Appendix 3 for footnote details
Full size table

Recommended delirium assessment tools included the Confusion Assessment Method (CAM) [24], Confusion Assessment Method-Intensive Care Unit (CAM-ICU) [25], the 4 “A”s Test (4AT) [26, 27], and the Intensive Care Delirium Screening Checklist (ICDSC) [28]. Approximately half (n = 14; 52%) of included guidelines explicitly recommended using validated tools for delirium assessment. The most common diagnostic criteria for delirium was the fourth or fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV or DSM-5), mentioned in 11 guidelines (41%). General recommendations for the recognition of delirium included training for healthcare professionals to increase awareness of delirium and delirium treatment options [29,30,31]. One guideline recommended referring patients with delirium to a trained mental health professional for further evaluation [32].

Recommendations for delirium prevention included pharmacological and non-pharmacological interventions once reversible medical causes of delirium (e.g., infection, pain) were addressed. Non-pharmacological interventions promoted patient orientation (i.e., using clocks, using a calendar, avoiding unnecessary room changes); social contact (i.e., friend and family visits); and comfort (i.e., avoiding unnecessary catheterization, monitoring nutrition and hydration, ensuring working hearing and visual aids) [23]. Adequate analgesia was recommended using a non-opioid medication first, and if an opioid was needed the minimum effective dose was recommended [29] with an opioid rotation in place [30]. Physical restraint was only recommended in exceptional circumstances, when a patient was a risk to themselves or others [33], and a restraint protocol should be used and routinely re-evaluated [34]. Pharmacological management for delirium most commonly included treatment with a benzodiazepine, a first, second, or third-generation antipsychotic [35], or a cholinergic drug [36]. Cancer therapy and medications exacerbating delirium (e.g., benzodiazepines, phenytoin) [33, 35] were recommended to be deprescribed. One guideline provided recommendations for imaging new onset (i.e., incident) delirium using a computed tomography (CT) head scan without contrast. Further evaluation of delirium with suspected brain abnormalities can be performed with contrast enhanced magnetic resonance imaging [37].

Summary of recommendations for anxiety, depression, and PTSD in delirium

A synthesis of recommendations for guidelines that report on anxiety, depression and PTSD in delirium is presented in Table 4. Two out of the seven guidelines (29%) [23, 38] provided recommendations specific to older adults > 65 years of age.

Table 4 Synthesis of recommendations for guidelines reporting on anxiety, depression, and post-traumatic stress disorder in delirium
Full size table

Four out of the seven guidelines (57%) provided recommendations specific to the management and prevention of anxiety in delirium. Andersen et al. [32] recommended screening based on validated tools such as the Beck Anxiety Inventory (BAI) [39], Generalized Anxiety Disorder (GAD-7) [40], Hospital Anxiety and Depression Scale (HADS) [41], or Spielberger State-Trait Anxiety Inventory (STAI) [42]. Non-pharmacological interventions were recommended for management of environmental and medical causes of anxiety [43]; however, no specific interventions were recommended. Pharmacological recommendations included the use of a benzodiazepine [43], while minimizing anxiolytics until delirium has resolved to limit delirium aggravation [36]. General treatment considerations included treating reversible medical causes of symptoms first [32], and treating delirium first over other psychiatric disorders [36]. Jacobi et al. indicated that pharmacological sedation should only be used after reversible medical causes were treated and analgesia was provided [20]. Andersen et al. recommended different treatment pathways based on the severity of symptoms of anxiety (e.g., mild symptomatology may require referral to supportive care services while severe symptomatology may require high intensity psychological and pharmacological intervention) [32].

Five out of the seven guidelines (71%) provided recommendations specific to the recognition and management of depression in delirium. Risk factors for depression included patient factors (i.e., cognitive decline or dementia, social isolation, personal or family history of depression or mood disorder) [23] and previous use of certain medications (i.e., antihypertensives, antimicrobials, and analgesics) [38]. Tools recommended for depression screening included the Sig: E. Caps [44], Cornell Scale for Depression [45], and Geriatric Depression Scale [46]. A variety of non-pharmacological interventions for depression were presented and these included: psychotherapy, exercise, electroconvulsive therapy, aromatherapy, and light therapy [23, 38]. All five guidelines mentioned pharmacological interventions and these included antidepressants such as selective serotonin reuptake inhibitors (SSRI) and tricyclic antidepressants; however, selection was recommended to be based on a medication with the fewest drug interactions [47] and with limited duration of therapy during delirium [36] as it may worsen delirium symptoms. General treatment considerations included first treating reversible medical causes, differentiating psychiatric symptoms from normal grieving process for those with cancer or advanced disease, and referring patients with a risk of suicide for an in-depth assessment with a mental health specialist [23]. One guideline recommended that an individualized treatment plan should be developed based on levels of depression severity, and the patient should be monitored for changes in behavior and followed-up every two weeks after intervention [32].

A single guideline out of the seven (14%) provided a recommendation specific to PTSD in delirium, and referred to using pharmacological sedation for agitation only after treating reversible medical causes of symptoms and providing adequate analgesia [20].

Discussion

In this systematic review, we provide an updated synthesis of CPGs for the diagnosis and management of delirium in patients admitted to an acute care setting. This review also provides a new analysis of CPGs with recommendations for the diagnosis and management of anxiety, depression, or PTSD in adults with delirium in acute care. Recommendations for the diagnosis of psychiatric disorders in delirium included screening with tools and assessment of risk factors, while recommendations for the management of psychiatric disorders in delirium included treating reversible medical causes first, utilizing non-pharmacological and pharmacological interventions, and monitoring the patient for changes in cognition. This review highlights the lack of published CPGs for healthcare professionals to recognize and prevent psychiatric disorders in adults with delirium using evidence-based practice.

Recommendations, if present, were often vague (e.g., recommending pharmacological agents without specifying by type or name of drug, lack of specific diagnosis criteria or flowcharts). This may be due to the lack of studies with a low risk of bias (i.e., randomized controlled trials [RCTs]) being performed on the prevention and treatment of psychiatric disorders in adults with delirium in acute care; or a deficit in updating guidelines to reflect current evidence-based practice. Clinical practice is altered constantly as new information is acquired, and new studies are performed [48]. In this review, 70% of included guidelines were published before 2019, and many have no explicit update information or criteria (AGREE II Domain 3, item 14). Up to date clinical practice guidelines are needed to ensure current evidence-based practice and better patient outcomes. Ideally, guideline developers should strive to improve not only the CPG update procedure and search terms used for evidence acquisition described in the AGREE II Domain 3, Rigour of Development but all AGREE II Domains with the lowest scores: Domain 5, Applicability; and Domain 6, Editorial Independence. Guideline developers should ensure that financial implications of guideline implementation have been considered and that clear auditing criteria are described to satisfy the Applicability Domain. Lastly, for the Editorial Independence Domain, the reporting of guideline developers’ competing interests and explicit funding statements should be included.

Healthcare professionals have reported feeling underprepared when dealing with patients with psychiatric disorders in the ICU [49] and may benefit from CPG recommendations to guide patient care. Some guidelines presented screening tools to facilitate the recognition of psychiatric disorders in delirium. Early identification of psychiatric disorders in delirium using standardized screening tools is important in the implementation of preventative interventions [50], which could contribute to a lowered psychological burden [51]. For the prevention of anxiety, depression, and PTSD in delirium, both pharmacological and non-pharmacological interventions were discussed. Non-pharmacological interventions (i.e., music therapy, mind-body interventions by family members or healthcare professionals, counselling) have been previously used in the ICU to reduce ICU-related distress without the need for sedative drugs [52]. Pharmacological interventions for anxiety or depression should be carefully assessed, as antidepressants and benzodiazepines are linked to worsened delirium [53, 54]. In response, guidelines recommended waiting until delirium has resolved to prescribe pharmacological interventions for anxiety or depression [36].

Our work aimed to update a 2017 systematic review on the quality of CPGs for use in the treatment of delirium within acute care [14]. Similarly, to this review, we identified many guidelines not kept up to date, as well as CPGs that achieved low AGREE scores in the domains of Applicability and Editorial Independence. However, our review reveals newfound differences in guideline target users as well as patient population, such that CPGs identified in our review were targeted to groups other than healthcare professionals (e.g., family caregivers, health system leaders, and policy makers). In addition, we found a higher proportion of CPGs that included recommendations for the general hospitalized population. This may represent the increased knowledge and recognition of delirium among members of the clinical care team (apart from healthcare professionals) in the acute care setting over the last 10 years.

The information provided in this systematic review must be taken in the context of relevant limitations. Most guidelines included in this review originated from Canada, the United States of America, or the United Kingdom, and were written primarily in English. Approaches to delirium recognition and treatment, in addition to the use of protocols for the management of delirium, are known to vary across countries [55]; guidelines included in our study may under-represent perspectives from other countries. While most clinicians understand CPGs to be helpful tools, their use in practice may be limited by the inflexibility of certain recommendations in specialized or unusual cases [56]. As well, to ensure reliability, two authors completed study screening, data extraction, and guideline quality assessment independently and in duplicate. Considering that reliability statistics in systematic reviews are sensitive to 'true prevalence' in the data—if the true prevalence of a population is high or low, agreement expected by chance increases and the magnitude of kappa goes down—reliability statistics were not conducted. Additionally, guidelines that were excluded from this review in the full-text stage due to lack of endorsement or availability of full-text may have provided additional recommendations that were missed.

This synthesis of guideline recommendations for delirium and for anxiety, depression, and PTSD in delirium provides a succinct guide for healthcare professionals. Based on the existing literature, current recommendations for the diagnosis and management of delirium and of anxiety, depression, and PTSD in delirium are minimal, but may help inform patient care. More recommendations for the diagnosis and management of psychiatric disorders in delirium are needed, and the conduct of RCTs for interventions of interest could facilitate higher quality evidence-based recommendations.

Conclusion

The evidence base is minimal for clinical practice guidelines that report on the diagnosis and management of symptoms of possible psychiatric disorders in adult patients with delirium in the acute care setting. Patients with delirium that display symptoms of psychiatric disorders may require specific evaluation during their hospitalization to ensure that psychiatric disorders are identified, and management plans are developed for this patient population in the follow-up period.