Description of the condition
Critical illness requiring admission to an intensive care unit (ICU) continues to increase in frequency around the world. As advances in health care are realized, more patients are surviving their stay in ICU but the implication of this is that there is an increase in the number of patients experiencing challenges during the recovery phase. During their ICU admission, patients experience extreme physical and psychological stressors including critical illness, delirium, fear, lack of privacy, noise, pain, sedation administration, sleep deprivation, and the abnormal ICU environment (Garrouste-Orgeas 2012; Kiekkas 2010; Merilaginen 2010). These experiences impact on a patient’s recovery from critical illness, which can be a complex and protracted process (Adamson 2004). Within this recovery period, patients may experience both physical (for example neuropathy, reduced mobility, and breathlessness) and psychological disorders (for example anxiety, depression and post-traumatic stress) (Cuthbertson 2007).
Psychological disorders, as well as anxiety and depression symptomatology, are commonly reported in patients and their caregivers after ICU admission. However, not every patient in ICU will develop psychological symptoms or a disorder; many individuals will be resistant or resilient to the ICU effects. Many who show distress will return quickly to normal function and some with a psychological disorder will follow a recovery trajectory (Layne 2007). Cross sectional and cohort studies have reported anxiety and depression conditions in patients recovering from ICU admission at a higher rate than the general population, at between 24% and 45% at six weeks (Myhren 2009), three months (Sukantarat 2007) and one year (Rattray 2005) after ICU admission. Anxiety and depression conditions often co-exist with post-traumatic stress disorder (PTSD) (Samuelson 2007). PTSD is a serious disorder that follows the experience of a traumatic event and causes significant impairment in daily life (American Psychiatric Association 1994). The experience of the stressor generates feelings of intense fear, horror, helplessness, threat to life and physical integrity for the individual or someone to whom they have close affectional ties (American Psychiatric Association 1994).
In addition to anxiety, depression and PTSD, ICU survivors have often reported the absence of factual memory and the occurrence of delusional memories, including hallucinations or nightmares, throughout their recovery period (Myhren 2009). ICU-related delusional memories are estimated to be present in around 30% to 70% of patients (Jones 2001; Ringdal 2009; Samuelson 2007), are often persecutory in nature, and tend to be recalled with high vividness and in substantial detail (Kiekkas 2010). The direct cause of these delusional memories is unknown but is thought to be related to a combination of medication (including adrenaline, corticosteroids, opiates, benzodiazepines and sedative drugs such as propofol), sleep deprivation, and critical illness (Jones 2001). The literature surrounding the relationship between recall of absent, traumatic or delusional memories and psychological disorders is mixed, with different authors finding positive (Jones 2001; Rattray 2010; Samuelson 2007; Schelling 2003) and negative associations (Granja 2008; Myhren 2009). The association between delusional memories and the psychological distress of ICU survivors has been mainly attributed to the strong vividness with long duration and high emotional content of these memories when compared with memories of real events (Ringdal 2009).
Research is now focusing on improving the long-term holistic health outcomes of ICU survivors. Psychological distress, including anxiety, depression and PTSD symptomatology, compromises the recovery of ICU survivors and has been increasingly identified as a serious problem. The challenge lies with clinicians and researchers to develop strategies to effectively manage and treat this psychological distress alongside and following life-saving physical treatment to maximise a patient's recovery.
Description of the intervention
One strategy that has been developed and implemented by clinical staff to treat the psychological distress prevalent in ICU survivors is patient diaries. Patient diaries provide a record of events which occur throughout a patient’s admission to the ICU. Following a time-line design, they provide a background to the cause of the patient’s ICU admission and an on-going narrative outlining day-to-day activities. Diversity of practice exists throughout ICUs in implementing patient diaries, including variation in structural, content and process elements.
Emerging in Scandinavia in the 1970s to 1980s (Egerod 2011a), multiple authors have outlined the introduction and evaluation of patient diaries both within their local ICUs and internationally. Patient diaries are generally written prospectively and addressed personally to the individual patient. ICU staff provide an overall structure for the diary, with a cover and sometimes a preprinted introduction and glossary of terms and equipment (Akerman 2010; Egerod 2007; Egerod 2011b). Diaries are generally structured with a summary outlining the reason and event of admission to ICU, daily entries, and a final note on discharge or transfer from the ICU (Egerod 2007).
Primary authorship is predominantly the responsibility of the bedside ICU nurse. Some ICUs encourage the participation of the patient’s family, reporting the diaries as a potential focus for family empowerment and family-centred care (Hale 2010; Roulin 2007). Current practice surrounding the provision of patient diaries to the patients is variable. ICUs differ between putting the diaries on the end of the bed when transferring a patient out of ICU to delivering a coordinated system of follow-up and support for the patients and their families (Akerman 2010; Egerod 2007; Roulin 2007).
How the intervention might work
Personal diaries are used by individuals to reflect on significant aspects of their lives and serve as a vehicle for construction, reconstruction and narration of stories (Egerod 2009). Patient diaries differ from personal diaries in that they are not first-person accounts. Nurses, hospital staff, family or friends vicariously write for the patient while the patient is unable to write due to altered state of consciousness, weakness or physical impairment (Egerod 2009).
Patients’ perceptions of intensive care are variable, often with very little or indeed nothing at all being remembered (Rattray 2010). For many patients their memories are unpleasant, fragmentary or frightening in nature (Rattray 2010). The aim of patient diaries is to provide ICU survivors with an accurate and informative collection of events, improving the memory recall of factual information. Delusional memories have been associated with anxiety, depression, post-traumatic stress symptomatology (Jones 2001; Rattray 2005) and poor health-related quality of life (Granja 2008). The aim of a diary is to provide a coherent narrative of the illness period, clarifying gaps in memory and diminishing the impact or dominance of imagined occurrences and hallucinations (Egerod 2011a). It has also been suggested that diaries can be used by relatives to encourage the healing process, after their own vicarious traumatic experience or as a basis for discussion about the patient’s illness experience (Egerod 2011a).
In comparison to this therapeutic view on patient diaries, there is however considerable concern regarding the method of providing this information and their use to reflect and reconstruct memories, thereby acting as a debriefing tool. Debriefing is a psychological treatment intended to reduce the psychological morbidity that arises after exposure to trauma (Rose 2002). It involves promoting some form of emotional process, catharsis or ventilation by encouraging recollection, ventilation or reworking of the traumatic event (Rose 2002). Since the 1990s debriefing has come under intense scrutiny, and a Cochrane review in 2002 (Rose 2002) found no evidence that single session individual psychological debriefing interventions prevented the onset of PTSD or reduced psychological distress. In addition to the lack of evidence, the majority of criticism is levelled at the timing of the debriefing, suggesting that during the immediate period after stress there is a substantial risk of causing retraumatization and inhibiting the individuals’ ability to normally process the traumatic event (Bledsoe 2002). Providing sensitive and private information without a supportive process could potentially cause significant psychological harm, negatively impacting a patient’s recovery.
Why it is important to do this review
Annual estimates suggest that more than 20 million patients require treatment in ICUs worldwide in order to manage critical illnesses, injuries or exacerbations of chronic conditions (Adhikari 2011). The combined after-effects of critical illness and the ICU experience have been linked to short and long-term psychological compromise, which can significantly impair psychological and physical patient recovery (Garrouste-Orgeas 2012; Kiekkas 2010). This results in a significant emotional, physical and financial burden to patients, families and society. Clinicians have developed and used patient diaries as a tool to treat psychological distress. However, it has not been established whether this is an effective practice or whether it may have an adverse psychological impact due to individual patient factors, author emphasis, or the method of feedback support or lack thereof.
The objective of this review is to assess the effect of patient diaries on recovery, in comparison to standard ICU treatment, in patients recuperating from admission to an ICU and their caregivers or families.
Criteria for considering studies for this review
Types of studies
We will include all randomized controlled trials (RCTs) and controlled clinical trials (CCTs) that have evaluated the effectiveness of patient diaries for their impact on recovery after admission to ICU. CCTs refer to quasi-randomized studies where although the trial involves testing an intervention and control, concurrent enrolment and follow-up of intervention and control-treated groups, the method of allocation is not considered strictly random (see Box 6.3a, Lefebvre 2011). We will include studies irrespective of publication status, year of publication or language. We will exclude non-randomized studies such as cohort studies because of the increased potential for bias. We will also exclude cross-over trials as this methodology is not suitable for evaluating an intervention that must be given at a specific time point.
Types of participants
We will include all patients who are admitted to an ICU and their family members or caregivers. The characteristics describing family members and caregiver participants will be defined by the study investigators. We will include patients irrespective of age, country and critical illness severity.
Types of interventions
The primary intervention under investigation is patient diaries provided by an ICU. We will include any RCT or CCT in which the presence or absence of patient diaries is the only difference between treatment groups. For the purpose of this review, patient diaries are defined as a prospectively written collection of events which occur during the ICU stay, authored by staff or relatives, or both (Garrouste-Orgeas 2012; Nydahl 2010).
Types of outcome measures
- Risk of post-traumatic stress disorder (PTSD) in patients recovering from admission to ICU, as assessed using a structured clinical interview (American Psychiatric Association 1994).
- Risk of anxiety in patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the Hospital Anxiety and Depression Scale (HADS) (Zigmond 1983).
- Risk of depression in patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the HADS (Zigmond 1983).
- Risk of memory recall of ICU in patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the Intensive Care Unit Memory Tool (ICU-MT) (Jones 2000).
- Post-traumatic stress symptomatology in patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the post-traumatic stress disorder-related symptoms screening tool (PTSS–14) (Twigg 2008).
- Post-traumatic stress symptomatology in caregivers or family members of patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the PTSS–14 (Twigg 2008).
- Risk of anxiety in caregivers or family members of patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the HADS (Zigmond 1983).
- Risk of depression in caregivers or family members of patients recovering from admission to ICU, as assessed using a tool with established reliability and validity such as the HADS (Zigmond 1983).
- Carer or family member satisfaction, as described by the study investigator.
- Health-related quality of life in patients recovering from admission to ICU, as assessed using a tool with established reliability and validity.
- Costs, as described by the study investigator; including implementation and healthcare utilization costs.
Search methods for identification of studies
We will search:
- the Cochrane Anaesthetic Group (CARG) Trials Register (latest update);
- the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (latest issue);
- Ovid MEDLINE (1950 to present);
- Ovid EMBASE (1980 to present);
- PsycINFO (1950 to present);
- Published International Literature on Traumatic Stress (PILOTS) database (1971 to present);
- EBSCOhost CINAHL (1982 to present); and
- Web of Science Conference Proceedings Citation Index - Science and Social Science and Humanities (1990 to present).
The search strategy which is detailed in Appendix 1 will be adapted to search the Cochrane Anaesthetic Group Trials Register, CENTRAL, Ovid MEDLINE, Ovid EMBASE and EBSCOhost CINAHL. We will combine the Ovid MEDLINE search with the Cochrane highly sensitive search strategy for identifying randomized trials in MEDLINE and EMBASE: the sensitivity and precision-maximizing version (2008 revision) (Lefebvre 2011). There will be no restrictions on the basis of date, language or publication status. We will also search the following clinical trial registers:
- Australian and New Zealand Clinical Trials Register (www.anzctr.org.au);
- Clinical Trials.gov (www.clinicaltrial.gov);
- Current Controlled Trials (www.controlled-trials.com/mrct);
- Hong Kong Clinical Trial Register (www.hkclinicaltrials.com);
- Clinical Trials Registry - India (www.ctri.in);
- UK Clinical Trials Gateway (www.controlled-trials.com/ukctr/); and
- World Health Organization (WHO) Clinical Trials Registry Portal (www.who.int/trialsearch).
Searching other resources
We will handsearch bibliographies of all retrieved and relevant publications identified by these strategies for further studies. We will contact experts in the field to ask for information relevant to this review.
Data collection and analysis
Selection of studies
We will combine the results of the searches and exclude duplicate records. Two review authors (AU and LA) will independently assess titles and abstracts of retrieved studies for relevance. After this initial assessment we will retrieve full versions of all potentially eligible studies. The same two review authors will then independently check the full papers for eligibility. We will resolve discrepancies between review authors through mutual discussion and, where required, consult a third independent review author (RB). For transparency we will publish a summary of the selection of studies, including excluded studies and reasons for exclusion, using the PRISMA flowchart (Liberati 2009).
Data extraction and management
We will extract the details from eligible studies and summarize them using a data extraction sheet (see Appendix 2). The data extraction sheet has been developed in conjunction with the CARG. We will pilot test the form for the first two studies. Two review authors (AU and LA) will extract data independently and then cross check for accuracy and agreement. We will resolve, where necessary, any discrepancies though discussion and arbitration with a third review author (RB). We will include studies that have been published in duplicate once only. If there are any missing data from the papers then we will make attempts to contact study authors to retrieve missing information. See Appendix 2 for more information regarding the data to be extracted.
Assessment of risk of bias in included studies
Two authors (AU and LA) will independently assess each eligible study for quality and bias using the ‘Risk of bias assessment tool’ described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will resolve disagreements by discussion and if we cannot reach a consensus a third author (RB) will arbitrate. The bias tool addresses six specific domains, namely sequence generation, allocation and concealment, blinding, incomplete outcome data, selective outcome reporting, and other issues which may potentially bias the study (Higgins 2011). We will express judgements as ‘low risk’, ‘high risk’, or ‘unclear risk’ of bias. We will conduct sensitivity analyses to determine whether excluding studies at high risk of bias affects the results of the meta-analysis. We will report the ’Risk of bias’ table as part of the table ‘Characteristics of included studies’ and present a ’Risk of bias summary’ figure which will detail all of the judgements made for all included studies.
Measures of treatment effect
For dichotomous outcomes, we will calculate the risk ratio (RR) plus 95% confidence interval (CI). For continuous outcomes, we will calculate the mean difference (MD) plus 95% confidence interval. It is likely that different tools may be used to measure the same outcome (for example PTSD, anxiety). We will collect data only from those studies where scales have been validated and are self-reported or completed by an independent rater or relative (not the investigator). We will use the standardized mean difference (SMD) as the summary statistic in any meta-analysis of such data.
Unit of analysis issues
We do not predict any unit of analysis issues. It is expected the patient, family member or caregiver will be the unit of analysis for all RCTs or CCTs.
Dealing with missing data
We will, whenever possible, contact the original investigators to request missing data. If the data are not retrievable, we will make explicit the assumptions of methods used to cope with missing data. That is, if the data are assumed to be missing at random we will analyse only the available data; but if missing values are assumed to have a particular value we will detail the method used to assign the value. Additionally, if missing data are evident, we will perform a sensitivity analysis to assess how sensitive results are to reasonable changes in the assumptions that are made, and we will address the potential impact of missing data on the findings of the review in the discussion section.
Assessment of heterogeneity
We will consider clinical, methodological and statistical heterogeneity. We will undertake an assessment of the comparability of the studies prior to meta-analysis. If appropriate, we will pool data using meta-analysis with RevMan 5.1. Due to the predicted clinical heterogeneity, we will use a random-effects model. We will investigate the degree of statistical heterogeneity, that is the variation between the true intervention effects underlying the different studies, by a combination of methods (Higgins 2002). This involves visual inspection of the meta-analytic model and interpretation of the Chi
Assessment of reporting biases
We will report each outcome separately. We will use funnel plots to assess reporting biases if there are at least 10 studies included (Sterne 2011). We will undertake an observation of small-study effects assessment if required.
Initially we will conduct a structured narrative summary of the studies reviewed. We will enter quantitative data into RevMan 5.1 and analyse the data using the RevMan analysis software. If the results are clinically heterogeneous, studies will be meta-analysed using both the fixed-effect model and random-effects model and we will explore any differences between these two estimates.
Subgroup analysis and investigation of heterogeneity
We do not predict any subgroup analysis for this review.
We will perform sensitivity analyses to exclude trials at high risk of bias, such as quasi-randomized trials. We will compare random-effects model and fixed-effect model estimates of each outcome variable.
Summary of findings
We will use the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence associated with specific outcomes: incidence of PTSD in patients recovering from admission to ICU; incidence of anxiety in patients recovering from admission to ICU; incidence of depression in patients recovering from admission to ICU; incidence of memory recall of ICU in patients recovering from admission to ICU; post-traumatic stress symptomatology in patients recovering from admission to ICU; and post-traumatic stress symptomatology in caregivers or family members of patients recovering from admission to ICU in our review and construct a 'Summary of findings' (SoF) table using the GRADE software. The GRADE approach appraises the quality of a body of evidence based on the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. The quality of a body of evidence considers within study risk of bias (methodologic quality), the directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.
We thank Jane Cracknell (Managing Editor, Cochrane Anaesthesia Review Group) and Karen Hovhannisyan (Trials Search Co-ordinator, Cochrane Anaesthesia Review Group) for their assistance in the preparation of this protocol. We would also like to thank Mathew Zacharias (content editor); Nathan Pace (statistical editor); Louise Rose, Ingrid Egerod and Megan Prictor (peer reviewers) for their help and editorial advice during the preparation of this protocol for the systematic review.
Appendix 1. MEDLINE (OvidSP) search strategy
1. ((patient* or caregiver*) adj5 (diaries or diary or (narrat* adj3 (coherent or outlining)))).af. or ((exp Patients/ or exp Caregivers/) and exp Narration/)
2. ((critical* adj3 ill*) or ((intensive care unit* or ICU) adj5 (recover* or delusional memor* or psychological distress or anxiety or depression or PTSD or bedside nurs* or family or caregiver* or recuperate*))).af. or exp Intensive Care Units/ or exp Critical Care/ or exp Critical Illness/
3. 1 and 2
Appendix 2. Data extraction form
Data collection form
Intervention review – RCTs only
1. General Information
2. Study Eligibility
DO NOT PROCEED IF STUDY EXCLUDED FROM REVIEW
3. Population and setting
5. Risk of Bias assessment
See Chapter 8 of the Cochrane Handbook
Provide overall data and, if available, comparative data for each intervention or comparison group.
7. Intervention groups
Copy and paste table for each intervention and comparison group
Copy and paste table for each outcome.
Copy and paste the appropriate table for each outcome, including additional tables for each time point and subgroup as required.
Dichotomous outcome 1
11. Other information
Contributions of authors
Conceiving the review: all authors
Co-ordinating the review: Amanda Ullman
Undertaking manual searches: Amanda Ullman
Screening search results: Amanda Ullman and Leanne Aitken
Organizing retrieval of papers: Amanda Ullman
Screening retrieved papers against inclusion criteria: Amanda Ullman and Leanne Aitken
Appraising quality of papers: Amanda Ullman, Leanne Aitken and Robyne le Brocque
Abstracting data from papers: Amanda Ullman, Leanne Aitken and Robyne le Brocque
Writing to authors of papers for additional information: Amanda Ullman
Providing additional data about papers: Amanda Ullman
Obtaining and screening data on unpublished studies: Amanda Ullman and Leanne Aitken
Data management for the review: Amanda Ullman
Entering data into Review Manager (RevMan 5.1): Amanda Ullman
RevMan statistical data: Amanda Ullman and Stephen MacGillivray
Other statistical analysis not using RevMan: Amanda Ullman
Interpretation of data: all authors
Statistical inferences: Stephen MacGillivray
Writing the review: all authors
Securing funding for the review: N/A
Performing previous work that was the foundation of the present study: all authors
Guarantor for the review (one author): Amanda Ullman
Person responsible for reading and checking review before submission: Amanda Ullman
Declarations of interest
The authors have no conflicts of interests to declare.
Sources of support
- NH&MRC Centre of Research Excellence in Nursing Interventions, Griffith University, Australia.
- Research Centre for Clinical and Community Practice Innovation, Griffith University, Australia.
- Princess Alexandra Hospital, Woolloongabba, Australia.
- School of Nursing and Midwifery, University of Dundee, UK.
- Department of Psychiatry, NHS Tayside, Perth, UK.
- School of Medicine, University of Queensland, Australia.
- Centre of National Research on Disability and Rehabilitation Medicine, University of Queensland, Australia.
- Social Dimensions of Health Institute, University of Dundee, UK.
- No sources of support supplied