Each year more people worldwide become critically ill and require treatment in the intensive care unit (ICU). Apart from the disease itself, the ICU environment exposes patients to extreme stressors, including painful procedures, multiple medications, mechanical ventilation, and inability to communicate. It has been demonstrated that most ICU patients experience episodes of delirium [
The disturbance develops over a short period of time (usually hours to days) and tends to fluctuate during the course of the day, and there must be evidence from the history, physical examination, or laboratory findings that the disturbance is caused by the direct physiological consequences of a general medical condition [
The primary aim of our study was to assess the correlation between ICU-delirium and the prevalence of PTSD-symptoms, other anxiety and depressive symptoms at 2 and 6 months after discharge from ICU. We also wished to assess whether prior mental illness expressed by number of redeemed prescriptions for psychotropic drugs prior to admission was associated with development of ICU-delirium. Finally, we wished to explore the associations between post-ICU PTSD-symptoms, anxiety, and depression.
Adult patients admitted to three mixed ICUs at two university hospitals in Denmark were consecutively included in this prospective observational study. Inclusion criteria were ICU stay >48 hours, age >17 years, and ability to communicate in Danish. Patients were excluded in cases of severe brain damage restricting communication, or PTSD diagnosed prior to admission. Informed consent was obtained during the first week after ICU-discharge and rereached at the start of the structured telephone interviews performed by the first author after 2 and 6 months. Patients’ answers were entered to a database as soon as they appeared. Before each call, the national database was checked to minimize the risk of bothering relatives if the patient had died in the meantime. Patients were called up to five times to make contact. If there was still no contact, the patient was excluded. Patients were excluded from the second interview if they were readmitted to ICU after the first interview.
Demographics, severity of illness during the first 24 hours of ICU-admission using Simplified Acute Physiology Score (SAPSII) [
The Confusion Assessment Method for the ICU (CAM-ICU) was used to detect delirium during the ICU stay [
The 16 PTSD items based on DSM-III-R in the fourth section of The Harvard Trauma Questionnaire (HTQ) were used in the present study [
A short form of The State-Trait Anxiety Inventory (STAI) was used as the ten state anxiety questions were extracted: “Right now I feel… tense/nervous/restless/anxious/guilty/and so forth.” Answers were rated on a four-point Likert scale from “Not at all” (score = 1) to “Very much so” (score = 4). A sum score of 20 or more indicated clinically significant anxiety [
Major Depression Inventory (MDI) was used as a rating scale to measure the degree of depression [
The ICU-Memory Tool (ICU-M) categorised memories in subscales as
Health related quality of life was assessed using the Short Form-36 (SF-36) providing information in eight specific domains: physical function, role physical, bodily pain, general health, vitality, social function, role emotional, and mental health [
The study protocol was approved by the Danish Data Protection Agency and the National Health Service of Denmark. Approval was not required by the Regional Research Ethics Committee. No interventions were performed. Patients were advised of the voluntary nature of the study and their right to withdraw at any time.
The number of participants included in each study was based on sample size calculations as established in a pilot study conducted in July 2009 and on patient characteristics in 2008. Our primary outcome was PTSD, and delirium the exposure. Assuming PTSD to have a prevalence of 22% [
We calculated the associations between post-ICU PTSD-symptoms, anxiety and depression, and the following. Age, gender, severity of illness, duration of mechanical ventilation, and ICU length of stay. ICU-memories of facts, feelings, and delusions. Health related quality of life. Antipsychotic medications administered in ICU.
Data are presented as proportions or as median and 10th and 90th percentile. Groups were compared using Chi-square test, Kruskal-Wallis test, or Wald test. The associations between number of delirium episode and scores at HTD, MDI, and STAI were estimated by linear regression. The associations between PTSD-symptoms, anxiety and depression, and number of memories were estimated by logistic regressions for each type of memories and adjusted only for age due to limited data. All analyses were done for both two and six month’s data. Results were termed significant when
Information was provided for 641 eligible patients, 360 gave consent from January 2010 to Marts 2012, 297 were interviewed after 2 months, and 248 after 2 and 6 months (2 after 6 months only). The flow diagram in Figure
Demographics and outcomes.
Interviewed | Not included |
|
|
---|---|---|---|
|
|
||
Age, median (10; 90 percentile) | 62 (40; 78) | 67 (46; 82) | <0.01 |
Gender | |||
Men, |
166 (56) | 244 (61) | 0.12 |
Woman, |
133 (44) | 153 (39) | |
SAPSII, median (10; 90 percentile) | 34 (19; 59) | 43 (24; 67) | <0.01 |
Type of admission | |||
Medical, |
113 (38) | 202 (51) | <0.01 |
Surgical, |
186 (62) | 195 (49) | |
Delirium | |||
Negative, |
138 (46) | 80 (20) | <0.01 |
Positive, |
161 (54) | 262 (66) | |
Unable to assess, |
0 (0) | 55 (14) | |
IV haloperidol, |
88 (29) | 147 (37) | 0.04 |
Atypical antipsychotics, |
60 (20) | 83 (21) | 0.82 |
ICU stay, median (10; 90 percentile) | 5 (2; 21) | 6 (3; 25) | 0.03 |
Cohort diagram of the study.
Delirium was detected in more than half of the patients (
Preadmission mental illness was estimated by the number of redeemed prescriptions for psychotropic drugs. More delirious patients than nondelirious patients used antidepressants (11% versus 7%), anxiolytics (22% versus 18%), hypnotics (28% versus 23%), and antipsychotics (11% versus 7%) prior to ICU-admission. Although these findings were consistent, they were not statistically significant. Eight patients were medicated for Attention Deficit Hyperactivity Disorder (ADHD), five of these became delirious, and three were not assessable (UTA).
During the ICU stay, intravenous haloperidol was used in 29% of the 299 interviewed patients and in 53% of the 161 delirious patients. Interviewed patients received less haloperidol than noninterviewed, whereas the two groups used similar doses of other antipsychotics (Table
We found no association between delirium and PTSD-symptoms, anxiety, or depression. OR (95% CI) at two months for PTSD was 1,25 (0,60; 2,57) anxiety 1,55 (0,59; 4,05) and depression 1,04 (0,72; 1,51). At six months OR for PTSD was 1,31 (0,63; 2,71), anxiety 1,14 (0,30; 4,35), and depression 1,04 (0,70; 1,53). For more details please see Table
Associations between delirium versus psychometrics.
ICU-delirium | No-delirium |
| |
---|---|---|---|
|
|
||
|
|||
|
|||
PTSD | |||
None, |
146 (92) | 129 (93) | 0.68 |
Present | 12 (8) | 9 (7) | |
Probable, |
9 (6) | 8 (6) | |
Severe, |
3 (2) | 1 (1) | |
HTQ-score, median (10; 90 percentile) | 1.14 (1.00; 1.89) | 1.19 (1.00; 1.88) | |
Anxiety | |||
None, |
145 (92) | 131 (95) | 0.37 |
Present | 12 (8) | 7 (5) | |
STAI-score, median (10; 90 percentile) | 12 (10; 20) | 12 (10; 18) | |
Depression | |||
None, |
142 (90) | 124 (90) | 0.95 |
Present | 16 (10) | 14 (10) | |
Mild, |
6 (4) | 6 (4) | |
Moderate, |
5 (3) | 5 (4) | |
Severe, |
5 (3) | 3 (2) | |
MDI-score, median (10; 90 percentile) | 7 (1; 21) | 6 (1; 20) | |
|
|||
|
|
||
|
|||
|
|||
PTSD | |||
None, |
121 (94) | 114 (97) | 0.44 |
Present | 8 (6) | 4 (3) | |
Probable, |
4 (3) | 1 (1) | |
Severe, |
4 (3) | 3 (3) | |
HTQ-score, median (10; 90 percentile) | 1.06 (1.00; 1.81) | 1.06 (1.00; 1.63) | |
Anxiety | |||
None, |
125 (96) | 114 (97) | 0.85 |
Present | 5 (4) | 4 (3) | |
STAI-score, median (10; 90 percentile) | 11 (10; 17) | 11 (10; 17) | |
Depression | |||
None, |
119 (92) | 106 (90) | 0.50 |
Present | 11 (8) | 12 (10) | |
Mild, |
3 (2) | 6 (5) | |
Moderate, |
2 (2) | 3 (3) | |
Severe, |
6 (5) | 3 (3) | |
MDI-score, median (10; 90 percentile) | 4 (0; 18) | 4 (0; 20) |
Symptoms of PTSD were found in 21 patients after 2 months, 12 of which had experienced delirium. At 6 months of follow-up, PTSD-symptoms were found in 12 patients (of which 8 were previously delirious) (Table
Previously delirious patients were more anxious according to STAI after 2 months (12 versus 7), but after 6 months this was resolved, and 5 versus 4 had symptoms of anxiety (Table
After two months, 30 patients suffered from depression according to the MDI; 12 (4%) had mild, 10 (4%) had moderate, and 8 (3%) had severe depression. In the severely depressed group, we found twice as many previously delirious compared to nondelirious patients (Table
When only considering the delirious patients, there were no differences in the psychometrics by delirium subtype (data not shown).
Neither haloperidol nor sedatives and other psychotropic drugs significantly affected the prevalence of PTSD-symptoms, anxiety, and depression. At both interviews, no statistically significant difference was found between the prevalence of PTSD-symptoms, anxiety or depression, and SAPSII, admission type (medical/surgical), hospital site, length of ICU stay, days with delirium, or days with sedation. For ten of the 21 patients with HTQ scores indicating PTSD at two months, the STAI score showed anxiety, and eight of these had symptoms of depression at the MDI as well.
The association of PTSD-symptoms, anxiety, and depression with memories of ICU depended on the type of memories experienced.
The psychometrics associations to ICU-memories at ICU-Memory Tool.
Memories of feelings | Memories of delusions | Memories of facts | ||||
---|---|---|---|---|---|---|
OR (95% CI) |
|
OR (95% CI) |
|
OR (95% CI) |
| |
|
||||||
PTSD | 1.27 (0.99; 1.63) | 0.06 | 1.06 (0.76; 1.48) | 0.73 | 0.94 (0.83; 1.06) | 0.32 |
PTSD |
1.20 (0.92; 1.57) | 0.17 | 1.00 (0.71; 1.40) | 0.99 | 0.91 (0.80; 1.04) | 0.16 |
Anxiety | 1.29 (0.98; 1.68) | 0.07 | 1.63 (1.16; 2.27) | <0.01 | 0.93 (0.82; 1.07) | 0.32 |
Anxiety |
1.20 (0.90; 1.60) | 0.07 | 1.55 (1.10; 2.18) | 0.01 | 0.90 (0.78; 1.04) | 0.15 |
Depression | 1.28 (1.03; 1.60) | 0.03 | 1.01 (0.75; 1.36) | 0.94 | 0.98 (0.88; 1.10) | 0.77 |
Depression |
1.17 (0.93; 1.48) | 0.18 | 0.92 (0.67; 1.25) | 0.59 | 0.95 (0.84; 1.06) | 0.36 |
|
||||||
|
||||||
PTSD | 1.46 (1.06; 2.01) | 0.02 | 1.43 (0.96; 2.14) | 0.08 | 1.10 (0.91; 1.33) | 0.31 |
PTSD |
1.31 (0.94; 1.82) | 0.11 | 1.26 (0.83; 1.91) | 0.27 | 1.06 (0.88; 1.30) | 0.51 |
Anxiety | 1.02 (0.71; 1.46) | 0.92 | 0.77 (0.41; 1.43) | 0.41 | 0.94 (0.78; 1.13) | 0.50 |
Anxiety |
1.00 (0.68; 1.47) | 0.97 | 0.75 (0.97; 1.40) | 0.36 | 0.93 (0.77; 1.12) | 0.44 |
Depression | 1.08 (0.86; 1.36) | 0.48 | 1.11 (0.80; 1.53) | 0.54 | 1.03 (0.91; 1.17) | 0.64 |
Depression |
1.04 (0.82; 1.33) | 0.75 | 1.06 (0.76; 1.48) | 0.73 | 1.02 (0.89; 1.16) | 0.80 |
OR = odds ratio. CI = confidence interval.
Based on answers from the 279 patients at the interview after 2 months and 240 patients after 6 months, there was a significant decrease in four domains of SF-36 (vitality, social function, role emotional, and mental health) if patients had PTSD-symptoms, anxiety, or depression. Bodily pain and general health (from SF-36) were reduced comparing patients with and without PTSD, anxiety, or depression. This was, however, not significant regarding PTSD-symptoms at 6 months (Table
The psychometrics associations to health related quality of life at the SF-36
|
Physical function | Role physical | Bodily pain | General health | Vitality | Social function | Role emotional | Mental health | |
---|---|---|---|---|---|---|---|---|---|
|
|||||||||
PTSD | |||||||||
None | 261 | 70 (10; 95) | 25 (0; 100) | 82 (25; 100) | 62 (25; 92) | 65 (25; 90) | 100 (50; 100) | 100 (33; 100) | 88 (60; 100) |
Present | 18 | 58 (10; 90) | 0 (0; 75) | 37 (0; 100) | 45 (15; 82) | 35 (5; 55) | 63 (0; 100) | 33 (0; 100) | 50 (28; 76) |
|
0.21 | 0.04 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | |
Anxiety | |||||||||
None | 263 | 70 (10; 95) | 25 (0; 100) | 82 (25; 100) | 62 (25; 92) | 65 (25; 90) | 100 (38; 100) | 100 (33; 100) | 88 (60; 100) |
Present | 16 | 50 (20; 80) | 0 (0; 100) | 42 (12; 100) | 45 (15; 82) | 43 (5; 80) | 50 (13; 100) | 33 (0; 100) | 52 (20; 92) |
|
0.07 | 0.04 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | |
Depression | |||||||||
None | 252 | 70 (10; 95) | 25 (0; 100) | 88 (34; 100) | 64 (30; 92) | 68 (30; 90) | 100 (50; 100) | 100 (67; 100) | 88 (68; 100) |
Present | 27 | 45 (5; 90) | 0 (0; 75) | 37 (0; 100) | 30 (10; 70) | 30 (5; 50) | 50 (13; 88) | 33 (0; 100) | 52 (32; 72) |
|
0.02 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | |
|
|||||||||
|
|||||||||
PTSD | |||||||||
None | 230 | 75 (18; 100) | 63 (0; 100) | 82 (25; 100) | 67 (25; 95) | 70 (28; 90) | 100 (50; 100) | 100 (67; 100) | 92 (62; 100) |
Present | 10 | 63 (15; 100) | 0 (0; 100) | 46 (18; 91) | 46 (20; 86) | 30 (5; 80) | 56 (6; 100) | 0 (0; 67) | 58 (28; 82) |
|
0.76 | 0.09 | 0.06 | 0.07 | <0.01 | <0.01 | <0.01 | <0.01 | |
Anxiety | |||||||||
None | 234 | 75 (20; 100) | 63 (0; 100) | 82 (25; 100) | 67 (25; 95) | 70 (30; 90) | 100 (50; 100) | 100 (67; 100) | 90 (64; 100) |
Present | 6 | 18 (0; 90) | 38 (0; 100) | 33 (0; 82) | 39 (0; 67) | 20 (0; 30) | 50 (13; 50) | 33 (0; 100) | 54 (20; 68) |
|
<0.01 | 0.30 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | |
Depression | |||||||||
None | 220 | 80 (20; 100) | 75 (0; 100) | 82 (25; 100) | 67 (30; 95) | 75 (35; 90) | 100 (50; 100) | 100 (84; 100) | 92 (68; 100) |
Present | 20 | 53 (3; 93) | 13 (0; 100) | 25 (0; 82) | 40 (5; 87) | 20 (5; 85) | 50 (19; 100) | 33 (0; 100) | 40 (18; 82) |
|
0.01 | 0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 | <0.01 |
Our study did not demonstrate an association between ICU-delirium and PTSD-symptoms or anxiety or depression at 2 and 6 months after discharge from ICU. We found that delirium in ICU was common (54%), and though the prevalence of depression, PTSD-symptoms, and anxiety was more common in post-ICU patients than the general population [
Anxiety and depression are often present during the ICU stay and is significantly higher than in ward patients [
In a study from 1999 24% showed signs of anxiety and 14% had depression within 24 hours of ICU-admission in the group of patients who were able to communicate verbally [
Memories of feelings or delusions in ICU affected PTSD-symptoms, anxiety, and depression, whereas factual memories did not. Most of the patients in our study had some recall of the stay in the ICU, in contrast to Jones et al. (2001), who found that most patients had no recall. Jones et al. concluded that ICU-memories might offer some protection against PTSD and anxiety, which is in accordance with our findings [
The local context in our ICU might have influenced the results of our study. The high nurse-patient ratio, free visiting hours, and absence of physical restraints might have had an impact on the incidence of delirium [
The evidence on treatment of ICU-delirium is scarce. In our cohort, the administration of antipsychotic medications in the ICU was common; 29% received haloperidol and 20% received other psychopharmacological agents, often in combination. Based on three randomized studies, the National Institute for Health and Clinical Excellence in the United Kingdom (UK) [
Our study showed that the four domains of mental health in SF-36 (vitality, social function, role emotional, and mental health) were decreased for patients with symptoms of PTSD, anxiety, or depression. It might be assumed that delirium was the cause as seen in other studies [
Participants in our study were younger, healthier, and less frequently delirious than nonparticipants indicating that we might underestimate the prevalence of later PTSD-symptoms. Since most nonparticipants failed to recover, we were unable to assess for PTSD-symptoms in this group. We did not aim to diagnose PTSD in our study but to detect symptoms of PTSD. A strong correlation, however, between PTSD and symptoms of PTSD was found in a self-reported questionnaire [
Our choice of telephone interviews rather than face-to-face encounters was pragmatic. The method is less demanding for patients with post-ICU fatigue and improves participation. In contrast to questionnaires, telephone interviews afford participants to clarify questions and validate answers. Preexisting disease has been estimated as the most important factor on post-ICU health related quality of life [
Delirium was detected in 54% of our sample. Symptoms of PTSD were found in 8% of previously delirious patients after 2 months and in 6% after 6 months. Most patients (93%) had some recall of the ICU stay. We did not find an association between ICU-delirium and PTSD-symptoms, anxiety, and depression at 2 and 6 months after ICU-discharge. Memories of delusions were significantly associated with anxiety after two months. Remaining associations between types of ICU-memories and prevalence of post-discharge symptoms of PTSD, anxiety, and depression were insignificant after adjusting for age. PTSD-symptoms, anxiety, and depression were associated with a significant reduction in most domains of mental health in SF-36. We recommend more attention to delirium prevention, detection, and management in the ICU.
Consider the following. ICU-delirium and post-discharge symptoms of PTSD anxiety or depression did not correlate. Memories of delusion in ICU affected anxiety after 2 months, whereas other psycometrics were not significant affected.
Attention Deficit Hyperactivity Disorder
Confusion Assessment Method for the ICU
Diagnostic and Statistical Manual of Mental Disorders
Harvard Trauma Questionnaire
Intensive care unit
ICU-Memory Tool
Major Depression Inventory
Posttraumatic stress disorder
Posttraumatic stress syndrome 10-question inventory
Richmond Agitation and Sedation Scale
Simplified Acute Physiology Score
Short Form-36
State-Trait Anxiety Inventory
United Kingdom
Unable to assess.
The authors have no competing interests to report.
Helle Svenningsen conducted the study under supervision of Poul Videbech, Else Kirstine Tønnesen, and Ingrid Egerod. Morten Frydenberg led the design of the statistical analysis. Doris Christensen provided clinical support and was the main contact in Eastern Denmark. Helle Svenningsen wrote the initial draft of the paper; all authors have participated in the revision of subsequent drafts and have read and approved the final version of the paper.
The authors wish to thank the Novo Nordic Foundation; Foundation for Psychiatry, Risskov; Foundation of Research in Mental Disorders, Aarhus University; Danish Society for Nursing Research; Directors’ Research Foundation at Hillerød Hospital; Færgemans scholarship; Foundation of Psychiatry promotion, Lippmann Foundation, and Familien Hede Nielsen Foundation, all from Denmark for supporting our study.