The mortality rate for patients with critical illness necessitating ICU admission has been declining (1 , 2) with a commensurate increase in the number of ICU survivors (3). As a result, the long-term sequelae of critical illness, including psychological, cognitive, and physical effects, are becoming more apparent. The term postintensive care syndrome (PICS) has been coined to refer to the broad range of symptoms that patients can experience after surviving an episode of critical illness (4). The psychological domain of PICS includes posttraumatic stress disorder (PTSD), depression, anxiety, and acute stress reactions. Symptoms of PTSD occur in one of five patients in the first 12 months post-ICU and are associated with worse health-related quality of life (HRQoL) (5–7). Further, up to one-third of ICU survivors report symptoms of anxiety (8) and depression (9). Family members and caregivers of ICU patients also report PTSD symptoms related to their loved ones’ ICU admission (10 , 11).
The development of PTSD after intensive care admission appears to have many contributing factors, one of which is memory disturbance (6). Patients with long stays in ICU often have lack of recall of events, and many suffer from delusions and hallucinations (12 , 13). These delusional memories of the ICU experience have been associated with the development of PTSD in ICU survivors (6 , 12 , 14), and thus are a potential focus for intervention.
Patient diaries are an intervention that may help ICU survivors and their families or caregivers in their recovery after critical illness. These ICU diaries are typically kept at the patient bedside and serve as a record of the ICU admission including the events leading to the admission, the patient’s daily status, as well as day-to-day procedures or treatments, and visitors. They are usually written by hospital staff, but may also include patient and family input, and sometimes include photos. The emphasis is on the use of plain, everyday language rather than technical or medical jargon (15). The aim of the diary is generally to provide a clear narrative of the sequence of events throughout the admission. It has been theorized that these diaries may help ICU survivors fill gaps in their memory (16), come to terms with their illness, and “diminish the impact or dominance of imagined occurrences and hallucination” (17), and thereby affect the psychological sequelae of PICS.
Although this intervention is potentially associated with benefits, available research is mostly confounded by small trials, uncontrolled reports or qualitative studies. ICU diaries have been the subject of several reviews already (15 , 17 , 18); however, as yet none have applied statistical analysis to all available research including both randomized controlled trials (RCTs) and cohort studies. A systematic review specifically looking at the effects of diaries in PICS including all available research was required to guide practice and future research.
The primary aim of this systematic review and meta-analysis was to determine whether the use of ICU diaries, compared with no diaries, reduced the rate of PTSD symptoms among ICU survivors and their families. Secondary objectives were to determine whether the use of patient diaries, compared with no diaries, reduced anxiety and depression symptoms, and improved HRQoL among ICU survivors and their family members.
Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Guidelines (19) were used in the preparation of this review. The protocol for our study was prospectively published on International Prospective Register of Systematic Reviews (PROSPERO) (Centre for Reviews and Dissemination 42017082479).
Two authors (P.A.M., R.S.K) independently searched the online databases PubMed, Embase, PsycINFO, and the Cochrane Central Register of Controlled Trials in December 2017 using Medical Subject Headings with search terms aiming to capture those publications which contained variations of “intensive care,” “PTSD/anxiety/depression/quality of life,” and “diary.” Full search strategies can be found in the supplemental data (Supplemental Digital Content 1, http://links.lww.com/CCM/E178). Publication alerts were set to capture relevant newly published studies. This search was supplemented by reviewing all references of relevant articles, consulting with leaders in the field and by searching clinical trials registries including the European Union Clinical Trials Register, ClinicalTrials.gov, and Australia New Zealand Clinical Trials Register. Reasons for exclusion were recorded. Disputes were resolved by discussion or review by a third author (M.R.).
Articles were included if they were original research where an ICU diary intervention group was compared with a group without diaries. RCTs, prospective or retrospective cohort, before-and-after, and case control studies were all included. We included all patients admitted to an ICU for more than 48 hours and their family members, regardless of age, illness severity, or admission category (i.e., medical or surgical). Studies were excluded if they were review articles, commentaries, if there was no control group, or if the patient group was not reflective of the target population (e.g., ICU length of stay < 48 hr). Studies were also excluded if their abstract or study was not fully published in the English language.
A structured template was used for the extraction of data from the included studies. Data collected included study location, publication year, hospital and patient demographics (including preexisting psychological conditions), timing of the diary intervention, length and method of follow-up, and the study-specific outcomes assessed. These data was independently extracted by two authors and compared for concordance.
Risk of Bias
Structured tools were used to assess the methodological quality of included studies. This included a modified version of the “Newcastle-Ottawa Quality Assessment Scale” (20) for cohort studies and the “Cochrane Risk of Bias Tool” (21) for included RCTs. Two authors (P.A.M., R.S.K.) independently assessed the risk of bias of the included studies with discrepancies resolved by blinded independent evaluation by a third author (A.T.).
The data from individual studies were converted to dichotomous outcomes where possible to facilitate pooling of results when different outcome measures were used between studies. Risk ratio (RR) and mean difference were used as the measures of effect for dichotomous and continuous outcomes, respectively. 95% CIs were calculated. Data were pooled using inverse variance weighting in random effects models. The Cochrane-Mantel-Haenszel chi-square test statistic was calculated to assess statistical significance. A p value of less than 0.05 was considered significant. We constructed forest plots to inspect for heterogeneity. The Cochrane Q test and I2 test were performed to objectively assess heterogeneity. Funnel plots and Egger test were used to test for evidence of publication bias. Data analysis was performed using Review Manager 5.3 (22) and Stata 13.0 (23).
Due to the small number of studies eventually included, and the heterogeneity of methods and reporting, some challenges were encountered in pooling the results. For several comparisons, including PTSD in relatives, and anxiety and depression in patients, not all studies which reported these outcomes could be included in the meta-analysis. It was not possible to dichotomize continuous scores such as the Hospital Anxiety and Depression Scale (HADS) using published data. In these instances, the pooled results and the results of the studies which could not be pooled are presented separately.
We identified 1,790 individual titles and abstracts for screening. Fourteen articles were selected for full text review of which seven satisfied all inclusion and exclusion criteria. An additional study was included following search of references, bringing the total number of included studies to eight. Of those excluded, one was an ongoing trial (24), one had no follow-up beyond the ICU stay (25), three had no control group (26–28), one had no original data (29), and in one study the patients included were not the target population (30). See Supplemental Table 4 (Supplemental Digital Content 1, http://links.lww.com/CCM/E178) for details of excluded studies. Figure 1 outlines the flowchart of the search.
Of the eight included trials, three were RCTs (31–33), with the remaining five being observational, cohort or time-series design (34–38). Characteristics of included studies are summarized in Table 1, with further information available in Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
Overview of Methodological Quality
Risk of Bias was assessed using Cochrane Risk of Bias Tool for RCTs, and a modified Newcastle-Ottawa Scale for cohort studies. Two of the RCTs were assessed as “low” risk for six out of seven categories (31 , 32) with the third (33) scoring “unclear” or “high” risk for six out of seven categories. One of the cohort studies (35) received eight out of nine stars (indicating low risk), with the remainder scoring five or six out of nine. The full results of risk of bias assessment are presented in Supplementary Tables 2 and 3 (Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
With the exception of Svenningsen et al (36), all included studies provided at least a brief description of the intervention. In all studies, guidelines were provided to staff to assist in standardizing entries. A summary of patients’ illness and reasons leading to admission was usually included. Two studies (35 , 37) also reported including a list and brief explanation of commonly used medical terms and equipment. Common themes included providing a daily update and a focus on the use of everyday language.
Three studies with a total of 452 patients investigated the effect of ICU diaries on PTSD symptoms (31 , 35 , 37). Garrouste-Orgeas et al (35) assessed PTSD symptoms according to Impact of Event Scale-Revised (39) with a score of greater than 22 defining PTSD. Glimelius Petersson et al (37) and Jones et al (31) both used posttraumatic stress syndrome (PTSS)–14 (40) to assess PTSD symptoms, using a cutoff greater than 45 to define PTSD. Garrouste-Orgeas et al (35) and Jones et al (31) both reported reduction in PTSD with ICU diaries. Glimelius Petersson et al (37) reported increased PTSD in those who received diaries; however, there were significant baseline differences between the two groups, with the patients receiving diaries having a significantly longer duration of mechanical ventilation and ICU length of stay than the control group, which is likely to have introduced bias. Overall, there was no statistically significantly reduction in PTSD with diaries, with a pooled RR 0.75 (95% CI, 0.3–1.73; p = 0.5) (Fig. 2). There was significant heterogeneity for this outcome (p = 0.05; I2 = 66%).
Two studies (32 , 35) (n = 173) reported significant improvement in PTSD symptoms of relatives of ICU survivors with the use of ICU diaries. Jones et al (32) reported mean PTSS-14 scores decreased (indicating improvement) from their baseline by 5 points with ICU diaries, compared with controls where mean scores actually increased by 5 (Mann-Whitney U test p = 0.03). Garrouste-Orgeas et al (35) reported fewer PTSD symptoms in those whose relatives received a diary RR 0.43 (95% CI, 0.27–0.68; p = 0.0004). Due to differences in reporting, these results could not be pooled (Supplementary Fig. 3, Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
Anxiety and Depression.
Three studies (33 , 35 , 38) measured anxiety and depression using the HADS (41). Anxiety and depression were dichotomized in two studies (33 , 35) (n = 88) using HADS score greater than or equal to 8 for defining anxiety or depression in the respective subscales. Significant reductions were observed in both anxiety (RR, 0.32; 95% CI, 0.12–0.86; p = 0.02) and depression scores (RR, 0.39; 95% CI, 0.17–0.87; p = 0.02) of these pooled results (Fig. 3). Fukuda et al (38) also reported a significant mean improvement in HADS scores for both anxiety (7.1 ± 3.8 to 5.7 ± 2.7; p = 0.011) and depression (8.6 ± 5.0 to 7.2 ± 4.3; p = 0.003). However, the published data from this study could not be converted into a dichotomous outcome to be included in pooled analysis. HADS scores are presented in Supplementary Figure 1 (Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
One study (35) reported no significant improvement in anxiety (RR, 0.94; 95% CI, 0.66–1.33; p = 0.72) and depression scores (RR, 0.98; 95% CI, 0.5–1.9; p = 0.95) of 143 relatives with the use of ICU diaries (Supplementary Figs. 4 and 5, Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
Two studies (n = 541) used the Short Form (SF)–36 (42) to assess HRQoL. Bäckman et al (34) reported significant improvement in SF-36 up to 3 years post intervention. Svenningsen et al (36) reported on a subset of their study on ICU patients with delirium who received a diary. They reported significant improvement for the general health (GH) subsection of the SF-36. When pooled with the Bäckman et al (34) results, this demonstrated a mean increase of 11.46 (95% CI, 5.87–17.05; p ≤ 0.0001) in the GH section of the SF-36 scores at 6 months post ICU discharge. Other subsections were not reported in full by Svenningsen et al (36). The forest plot is presented in Supplementary Figure 2 (Supplemental Digital Content 1, http://links.lww.com/CCM/E178).
No studies reported on HRQoL in relatives.
There were an insufficient number of studies to enable us to comment on publication bias from the funnel plots for all studied outcomes (Supplementary Fig. 7, Supplemental Digital Content 1, http://links.lww.com/CCM/E178). Egger test was not significant for the primary outcome.
A summary of the key findings are presented in Table 2, prepared using GradePro Guideline Development Tool (43).
Despite a large number of publications pertaining to ICU diaries, few studies were included in this review and our results have to be interpreted with this limitation in mind. We found that ICU diaries were associated with an improvement in anxiety and depression in patients, and PTSD in families or relatives. We did not find a statistically significant effect for PTSD in patients.
For our primary outcome, PTSD, both Garrouste-Orgeas et al (35) and Jones et al (31) reported significant reduction in PTSD in their respective trials. In contrast, Glimelius Peterssen et al (37) reported increased PTSD in patients with a diary. This is potentially explained by several factors: their criterion for starting a diary was expected ICU stay of 3 or more days, but not all patients meeting this criterion received a diary, and the reasons for this are not explained fully. The ICU diaries group had almost double the ICU length of stay (9.7 vs 4.9 d) and more received invasive mechanical ventilation (90% vs 41%) than the control group. Finally, they only followed up their cohort to 2 months, despite acknowledging the 2002 Brussels Roundtable (44) which suggested follow-up should be for at least 6 months post ICU discharge. For comparison, the most significant results of Garrouste-Orgeas et al’ s (35) trial were not seen until the 12-month follow-up.
Anxiety, depression, and HRQoL all showed significant improvement with the use of patient diaries. Ultimately, quality of life is the most patient-centered outcome available. Surprisingly few studies addressed this as an outcome measure, and those that did had some methodological limitations—Bäckman et al (34) had significant losses to follow-up, particularly in those that did not receive diaries, and in the Svenningsen et al (36) study the focus was on the effect of delirium on PTSD, with the use of diaries reported as an observation. Thus, although in the present review HRQoL was significantly improved by ICU diaries, this is an area which would benefit from further investigation and more robust study design to enhance this conclusion. These are highly patient-centered outcomes of significant interest to patients, families, carers, clinicians, and the broader community, and thus would be useful to investigate in the future.
Two individual studies (32 , 35) demonstrated significant reduction in PTSD for relatives; however, the same significant results were not seen with anxiety and depression. This differential effect of diaries is plausible as the clarification and demystifying function of diaries may mitigate the trauma of having watched a relative suffer through critical illness and hence reduce PTSD, but not necessarily anxiety or depression.
Our study has a number of limitations. It was limited by the quality of included studies, which overall were of moderate quality with substantial risk of bias. Many of the studies were observational and included small patient numbers. Further, although several studies reported significant findings, we were unable to pool some results due to different reporting methods. As a result, we were only able to include three or fewer studies for meta-analysis for each outcome. Therefore, the risk of bias and inaccurate conclusions is high. Further, the length of follow-up ranged from hospital discharge (38) to 36 months post discharge (34), and all results were pooled together which is a cause of heterogeneity. Not all outcomes of interest to patients and families were reported, particularly functional outcomes. Our search was limited to English publications, and unpublished data were not sought.
The only previous systematic review on this topic included only RCTs and hence did not include a large proportion of the ICU diaries literature. Our search was extensive and identified papers not mentioned in previous reviews. Our results are consistent with, and advance upon the findings from previous reviews (5 , 18) which suggested that ICU diaries may reduce psychological morbidity after ICU admission for both patients and their families.
There are a number of ongoing trials which will add to the evidence base for ICU diaries. The ICU-Diaries study (24) is a multicenter RCT currently underway in France, and a further RCT (45) in Denmark is exploring if a diary written by relatives has similar effects on psychologic outcomes. Future research should address the effect on quality of life and activities of daily living, as these are likely to be of most value to patients. Uniform or standardized reporting of results as recommended by a recently reported Delphi process (46) would allow for results to be more easily compared and collaborated. We plan to perform an updated meta-analysis once the results of these above-mentioned studies have been published.
The use of diaries for ICU patients improves anxiety, depression, and HRQoL. There is no statistically significant reduction in PTSD. Large prospective studies with comprehensive assessment of patient-centered outcomes are required for a definitive answer.
1. Kaukonen K-M, Bailey M, Suzuki S, et al. Mortality related to severe sepsis and septic shock among critically ill patients in Australia and New Zealand, 2000–2012. JAMA 2014; 311:1308–1316
2. Australian and New Zealand Intensive Care
Society Centre for Outcome and Resource Evaluation. ANZICS CORE Annual Report 2017. 2017Melbourne, Australian and New Zealand Intensive Care
Society Centre for Outcome and Resource Evaluation.
3. Iwashyna TJ, Cooke CR, Wunsch H, et al. Population burden of long-term survivorship after severe sepsis in older Americans. J Am Geriatr Soc 2012; 60:1070–1077
4. Needham DM, Davidson J, Cohen H, et al. Improving long-term outcomes after discharge from intensive care
unit: Report from a stakeholders’ conference. Crit Care Med 2012; 40:502–509
5. Parker AM, Sricharoenchai T, Raparla S, et al. Posttraumatic stress disorder
in critical illness
survivors. Crit Care Med 2015; 43:1121–1129
6. Davydow DS, Gifford JM, Desai SV, et al. Posttraumatic stress disorder
in general intensive care
unit survivors: A systematic review
. Gen Hosp Psychiatry 2008; 30:421–434
7. Wade D, Hardy R, Howell D, et al. Identifying clinical and acute psychological risk factors for PTSD after critical care: A systematic review
. Minerva Anestesiol 2013; 79:944–963
8. Nikayin S, Rabiee A, Hashem MD, et al. Anxiety symptoms in survivors of critical illness
: A systematic review
and meta-analysis. Gen Hosp Psychiatry 2016; 43:23–29
9. Rabiee A, Nikayin S, Hashem MD, et al. Depressive symptoms after critical illness
: A systematic review
and meta-analysis. Crit Care Med 2016; 44:1744–1753
10. Jones C, Skirrow P, Griffiths RD, et al. Post-traumatic stress disorder-related symptoms in relatives of patients following intensive care
. Intensive Care
Med 2004; 30:456–460
11. Sundararajan K, Martin M, Rajagopala S, et al. Posttraumatic stress disorder
in close Relatives of Intensive Care
unit patients’ Evaluation (PRICE) study. Aust Crit Care 2014; 27:183–187
12. Jones C, Griffiths RD, Humphris G, et al. Memory, delusions, and the development of acute posttraumatic stress disorder
-related symptoms after intensive care
. Crit Care Med 2001; 29:573–580
13. Rundshagen I, Schnabel K, Wegner C, et al. Incidence of recall, nightmares, and hallucinations during analgosedation in intensive care
. Intensive Care
Med 2002; 28:38–43
14. Kiekkas P, Theodorakopoulou G, Spyratos F, et al. Psychological distress and delusional memories after critical care: A literature review. Int Nurs Rev 2010; 57:288–296
15. Aitken LM, Rattray J, Hull A, et al. The use of diaries
in psychological recovery from intensive care
. Crit Care 2013; 17:253
16. Ewens B, Chapman R, Tulloch A, et al. ICU survivors’ utilisation of diaries
post discharge: A qualitative descriptive study. Aust Crit Care 2014; 27:28–35
17. Ullman AJ, Aitken LM, Rattray J, et al. Intensive care diaries
to promote recovery for patients and families after critical illness
: A Cochrane systematic review
. Int J Nurs Stud 2015; 52:1243–1253
18. Mehlhorn J, Freytag A, Schmidt K, et al. Rehabilitation interventions for postintensive care syndrome
. Crit Care Med 2014; 42:1263–1271
19. Moher D, Liberati A, Tetzlaff J, et al; PRISMA Group: Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 2009; 6:e1000097
21. Higgins JP, Altman DG, Gøtzsche PC, et al. Cochrane Bias Methods Group; Cochrane Statistical Methods Group: The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ 2011; 343:d5928
22. The Cochrane Collaboration: Review Manager: RevMan [Computer program]. Version 5.3. 2014Copenhagen, The Nordic Cochrane Centre, The Cochrane Collaboration.
23. StataCorp: Stata Statistical Software: Release 13. 2013TX, StataCorp LP, College Station.
24. Garrouste-Orgeas M, Flahault C, Fasse L, et al. The ICU-Diary study: Prospective, multicenter comparative study of the impact of an ICU diary on the wellbeing of patients and families in French ICUs. Trials 2017; 18:542
25. Kadivar M, Seyedfatemi N, Akbari N, et al. The effect of narrative writing on maternal stress in neonatal intensive care
settings. J Matern Neonatal Med 2015; 28:938–943
26. Bäckman CG, Walther SM. Use of a personal diary written on the ICU during critical illness
. Intensive Care
Med 2001; 27:426–429
27. Combe D. The use of patient diaries
in an intensive care
unit. Nurs Crit Care 10:31–34
28. Storli SL, Lind R. The meaning of follow-up in intensive care
: Patients’ perspective. Scand J Caring Sci 2009; 23:45–56
29. Thomas J, Bell E. Lost days–diaries
for military intensive care
patients. J R Nav Med Serv 2011; 97:11–15
30. Kloos JA, Daly BJ. Effect of a Family-Maintained Progress Journal on anxiety of families of critically ill patients. Crit Care Nurs Q 2008; 31:96–107; quiz 108109
31. Jones C, Bäckman C, Capuzzo M, et al. Intensive care diaries
reduce new onset post traumatic stress disorder following critical illness
: A randomised, controlled trial. Crit Care 2010; 14:R168
32. Jones C, Bäckman C, Griffiths RD. Intensive care diaries
and relatives’ symptoms of posttraumatic stress disorder
after critical illness
: A pilot study. Am J Crit Care 2012; 21:172–176
33. Knowles RE, Tarrier N. Evaluation of the effect of prospective patient diaries
on emotional well-being in intensive care
unit survivors: A randomized controlled trial*. Crit Care Med 2009; 37:184–191
34. Bäckman CG, Orwelius L, Sjoberg F, et al. Long-term effect of the ICU-diary concept on quality of life after critical illness
. Acta Anaesthesiol Scand 2010; 54:736–743
35. Garrouste-Orgeas M, Coquet I, Périer A, et al. Impact of an intensive care
unit diary on psychological distress in patients and relatives* Crit Care Med 2012; 40:2033–2040
36. Svenningsen H, Tønnesen EK, Videbech P, et al. Intensive care
delirium - effect on memories and health-related quality of life - a follow-up study. J Clin Nurs 2014; 23:634–644
37. Glimelius Petersson C, Ringdal M, Apelqvist G, et al. Diaries
and memories following an ICU stay: A 2-month follow-up study. Nurs Crit Care 2015; 23:299–307
38. Fukuda T, Inoue T, Kinoshita Y, et al. Effectiveness of ICU diaries
: Improving distorted memories encountered during ICU admission. Open J Nurs 2015; 5:313–324
39. Weiss DS, Marmar CR. Wilson JP, Keane TM. The Impact of Event Scale—Revised. In: Assessing Psychological Trauma and PTSD: A Practitioner’s Handbook. 1997, pp New York, NY, Guildford Press, 399–411.
40. Twigg E, Humphris G, Jones C, et al. Use of a screening questionnaire for post-traumatic stress disorder (PTSD) on a sample of UK ICU patients. Acta Anaesthesiol Scand 2008; 52:202–208
41. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67:361–370
42. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30:473–483
43. GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. McMaster University, 2015 (developed by Evidence Prime, Inc.). Available at: https://gradepro.org
44. Angus DC, Carlet J; 2002 Brussels Roundtable Participants: Surviving intensive care
: A report from the 2002 Brussels Roundtable. Intensive Care
Med 2003; 29:368–377
46. Needham DM, Sepulveda KA, Dinglas VD, et al. Core outcome measures for clinical research in acute respiratory failure survivors: An international modified delphi consensus study. Am J Respir Crit Care Med 2017; 196:1122–1130