Posttraumatic Stress Disorder FollowingCritical Illness

Erin K. Kross, MD, Cynthia J. Gries, MD, MSc,J. Randall Curtis, MD, MPH*

Division of Pulmonary and Critical Care Medicine, Harborview Medical Center,

School of Medicine, University of Washington, 325 Ninth Avenue, Seattle,

WA 98104-2499, USA

Posttraumatic stress disorder (PTSD) is a common psychiatric conditionthat can occur after a traumatic event. PTSD is the fourth most commonpsychiatric illness in the United States with estimates of lifetime prevalenceranging from 5% to 6% in men to 10% to 14% in women [1]. Traumaticevents can provoke fear, helplessness, or horror in response to an eventthat threatens life or safety [1]. Prevalence estimates for adults who are atrisk for PTSD range from 2% to 15% after combat in Vietnam to 14%to 80% after rape [2]. In addition to PTSD, individuals exposed to traumaticevents also are at risk for other psychologic morbidity, such as depression,panic disorder, generalized anxiety disorder, and substance abuse. The bur-den of PTSD can be high and can result in an inability to work or return toprior levels of functioning. Therefore, PTSD results in an increased cost tosociety because of increased health care costs and decreased productivity.

The criteria for PTSD are defined by the Diagnostic and StatisticalManual of Mental Disorders, Fourth Edition (DSM-IV) [3]: a person musthave been exposed to a traumatic event, which involves a perceived or actualthreat to the person’s own life or physical integrity or that of another per-son. The condition is characterized by a constellation of symptoms in threedomains: (1) symptoms of re-experiencing, (2) symptoms of avoidance andemotional numbing, and (3) symptoms of increased arousal. These symp-toms must meet two criteria to satisfy the diagnosis: (1) symptoms mustcause significant impairment in social, occupational, or other important

Crit Care Clin 24 (2008) 875–887

* Corresponding author. Division of Pulmonary and Critical Care, Harborview Medical

Center, University of Washington, 325 Ninth Avenue, Box 359762, Seattle, WA 98104-2499.

E-mail address: jrc@u.washington.edu (J.R. Curtis).

0749-0704/08/$ - see front matter � 2008 Elsevier Inc. All rights reserved.

doi:10.1016/j.ccc.2008.06.002 criticalcare.theclinics.com

876 KROSS et al

functional domains and (2) symptoms must be present for a least 1 monthafter exposure to the traumatic event or events.

ICU treatment for critical illness exposes patients and their families toenormous stress. This stress results from the experience of life-threateningillness and the need for intensive, and often invasive, medical procedures.Survivors of critical illness often report memories of pain and anxiety duringtheir ICU stay [4]. Therefore, the trauma of a critical illness and ICU treat-ment meets the criteria of a traumatic event as stated by the DSM-IV crite-ria of PTSD. A similar experience applies to family members of critically illpatients, as the DSM-IV definition allows for a traumatic, life-threateningevent to be witnessed rather than personally experienced. Family membersof ICU patients may be at risk for PTSD as ICUs are a foreign and oftenfrightening environment for many family members. Family members maybe exposed to invasive monitoring for their loved ones and unfamiliar med-ical procedures and devices. In addition, family members often are asked toassume the role of surrogate decision maker, because the majority ofpatients in ICUs are not able to participate in decisions about withholdingor withdrawing life support. The role of being a surrogate decision makerseems to be associated with additional stress for some family members [5].Therefore, three groups are considered that are at risk for PTSD afterICU treatment. First, there are patients who survive a critical illness andare discharged after ICU care. Second, there are family members of individ-uals who survive critical illness. Finally, there are family members ofindividuals who die during or shortly after their ICU stay. This article-reviews the body of literature regarding adult ICU patients and their fami-lies for prevalence of PTSD after critical illness in these three groups andidentifies risk factors for the development of PTSD and suggestions forclinical implications and further research in this field.

Survivors of critical illness

The bulk of the literature investigating PTSD after critical illness hasexamined survivors of critical illness. Two systematic reviews of the topicrecently have been published; the first looked at PTSD in 16 studies of med-ical ICU patients [2] and the second looked at survivors of all intensive caretreatment [6]. For this review, only studies that evaluated the associationbetween medical ICU admission and PTSD (or posttraumatic stress symp-toms) are included, similar to the review by Jackson and colleagues [2].In that review, studies were required to use quantitative or objective mea-sures of PTSD. Studies of patients in surgical or trauma ICUs were excludedbecause of concern that the surgery or trauma itself may have elicited PTSDsymptoms rather than the ICU experience. Five of these studies were pro-spective cohort studies [7–11] and one was a randomized controlled trial[12]. The remainder of the studies included six retrospective cohort studies[13–18] and four cross-sectional studies [19–22]. Since this review was

877POSTTRAUMATIC STRESS DISORDER

published, there have been two other prospective cohort studies publishedassessing prevalence of PTSD symptoms after critical illness [4,23]. Table 1shows an overview of these studies.

The studies included in the review varied greatly in methodology. Criteriafor entry into each study differed, including varied required lengths of stayin an ICU. Some studies investigated all medical intensive care patients,whereas others were restricted to those who had specific medical conditions,such as acute lung injury or septic shock. Patients in the studies variedregarding severity of illness and need for mechanical ventilation, both ofwhich may affect the experience of ICU treatment. The numbers of subjectincluded in each study ranged from 20 to 143. A few of the studies evaluatedindividuals at multiple time points [7,12,14,15,23]. In these, the initial eval-uation occurred within 2 months of hospital discharge whereas follow-upevaluations occurred at varying time points, up to as many as 8 years afterICU discharge. The remaining studies surveyed patients at only one timepoint, which ranged from 3 months to 13 years after discharge from anICU or hospital. Patient follow-up rates also varied in these studies, rangingfrom 24% to 84% [2].

Several, although not all, of the studies, excluded subjects who had priorpsychiatric or neurologic illness [4,12,13,15,17,20]. The studies also varied inthe method in which PTSD was measured and assessed. Nine of the studiesused only standardized screening tools for assessment, including the Post-Traumatic Stress Syndrome 10-Questions Inventory (PTSS-10), Impact ofEvents Scale (IES), IES-Revised, Davidson Trauma Scale, Trauma Symp-tom Checklist-33, and Experience of Treatment in Intensive Care 7-ItemScale [4,7,8,11–13,16,18,20,21]. In most cases, these tools were administeredin person; however, in two studies they were administered by telephone[18,20]. In this group of studies, the evaluation of PTSD was made solelyon information gathered from these screening tools. Five of the studiesused the Structured Clinical Interview for DSM-IV (SCID) after a screeningtool suggested PTSD [9,14,15,17,19]. In general, these five studies showedlower prevalence of PTSD identified by the SCID. For example, one studyshowed that only 57% of the subjects who screened positive via the PTSS-10were subsequently identified as having PTSD when followed-up with theSCID [19]. Most of these studies, however, did not assess using the SCIDin patients who did not screen positive on a screening tool.

Prevalence of PTSD ranged from 5% in a prospective cohort study thatused the IES [7] to 63% in the placebo group in a retrospective cohort studythat used the PTSS-10 and SCID to measure PTSD (see Table 1) [13].Several of the studies that reported higher prevalence of PTSD (54%,59%, and 63%) had particularly small sample sizes or were subgroup anal-yses of larger populations [9,13,17]. There also was variation in the reportedprevalence based on time of assessment of symptoms. Specifically, preva-lence was highest when assessment occurred close to the time of dischargeand decreased as time from discharge increased [2].

Table 1

Summary of studies assessing the prevalence of posttraumatic stress disorder in survivors of critical illness in the medical intensive care unit

Author and

reference no. Study design

Time between critical

illness and assessment

of posttraumatic

stress disorder Instrument used Sample size Response rate

Prevalence of

posttraumatic stress

disorder

Schelling, et al

1999 [17]

Retrospective

cohort

2 to 9 years PTSS-10, clinical

interview

54 Unknown 38% (18.5% in

treatment group,

59% in control

group)

Schelling, et al

1998 [16]

Retrospective

cohort

6 to 10 years PTSS-10 80 78% 27.5%

Stoll, et al

1999 [15]

Retrospective

cohort

2 time points at least

2 years apart (1 to

19 years after

discharge)

PTSS-10, clinical

interview

52 65% 25%

Eddleston, et al

2000 [10]

Prospective

cohort

3 months Selected PTSD

questions

143 63% 36% with ‘‘distressing

flashback’’

Nelson, et al

2000 [22]

Cross-sectional 6 to 41 months

(mean 19 months)

Seven items pertaining

to PTSD

24 71% 39% with ‘‘bad

memories

or dreams’’

Schelling, et al

2001 [13]

Retrospective

cohort

21 to 49 months PTSS-10, SCID 20 84% 40% (63% in placebo

roup, 11% in

treatment group)

Scragg, et al

2001 [20]

Cross-sectional O5 years IES, TSC-99, ETIC-7 60 56% 15% with PTSD (30%

with posttraumatic s

tress symptoms)

Shaw, et al

2001 [21]

Cross-sectional Unknown IES 20 N/A 35%

Jones, et al

2003 [12]

Randomized

controlled trial

6 weeks and

6 months

IES 115 at 6 weeks,

102 at 6 months

80% 51% (at 6 months)

878

KROSSet

al

Kress, et al

2003 [9]

Prospective cohort Approximately

1 year

IES-R, clinical

interview

32 30% 18.5% (54% from

control group, 0%

from intervention

group)

Cuthbertson, et al

2004 [8]

Prospective cohort 3 months Davidson Trauma

Scale

78 70% 14%

Kapfhammer, et al

2004 [14]

Retrospective

cohort

Median of 8 years PTSS-10, SCID 46 58% 23.9%

Nickel, et al

2004 [19]

Cross-sectional Unknown PTSS-10, SCID 41 N/A 9.76% with PTSD

(17% with

posttraumatic stress

symptoms)

Capuzzo, et al

2005 [7]

Prospective cohort 1 week and

3 months

IES 63 at 3 months 73% 5%

Rattray, et al

2005 [11]

Prospective cohort Hospital discharge,

6 months and

12 months

IES 67 at 6 months,

60 t 12 months

73% 20% with high

avoidance scores,

18% with high

intrusion scores

Deja, et al

2006 [18]

Retrospective

cohort

57 þ/� 32 months PTSS-10 65 49.6% 29%

Girard, et al

2007 [4]

Prospective cohort 6 months PTSS-10 43 24% 14%

Sukantarat, et al

2007 [23]

Prospective cohort 3 and 9 months IES 51 88% 24%–38%

879

POSTTRAUMATIC

STRESSDISORDER

a

880 KROSS et al

Several risk factors for PTSD were identified in these studies, althoughthe methods used varied. Risk factors that were identified by more thanone study include more delusional memories as opposed to factual memories[7,12,19], increased length of stay or duration of mechanical ventilation[8,11,14], greater levels of sedation or neuromuscular blockade [4,9,22],younger age [4,8,11,20], prior mental health history [8,19], and female gen-der [4,10,20]. In one study, greater perceived social support was implicatedas protective against the development of PTSD [18].

In summary, it is difficult to provide a precise estimate of the prevalenceof PTSD among survivors of critical illness, but these studies suggest it issignificantly higher than seen in the general population and probably inthe range of 14% to 50%. Several risk factors for the development ofPTSD have been identified in those who are critically ill, including featuresof the ICU stay that correlate with higher severity of illness and length ofstay and features of patients suggesting that some characteristics put indi-viduals at increased risk.

Family members of survivors of critical illness

Far fewer studies have investigated the prevalence of PTSD in familymembers of individuals who survive their critical illness and these studiesare outlined in Table 2 [5,24]. The largest of these was conducted in Franceby Azoulay and colleagues [5] and was a longitudinal study conducted in21 medical-surgical ICUs in 2003. Family members were eligible for studyif they came to an ICU to visit a patient who had an ICU stay of greaterthan 48 hours. The closest family member was identified and offered partic-ipation in the study. Telephone interviews were conducted 90 days after ICUdischarge or death. Family members completed the IES, the 36-Item ShortForm Health Survey (SF-36), and the Hospital Anxiety and DepressionScale. The investigators considered IES scores greater than 30 as the cutofffor posttraumatic stress reaction (PTSR), indicating a significant risk forPTSD. A total of 459 patients were screened for inclusion. Two hundredeighty-four family members consented and were able to be reached for fol-low-up. Of these, 228 were family members of ICU survivors whereas 56were family members of individuals who died in an ICU.

In this study, 94 (33.1%) family members screened positive for PTSRwith an IES score greater than 30 (see Table 2). When restricting analysisto those whose family members who survived their ICU stay, 28.9% of fam-ily members screened positive for PTSR. Factors independently associatedwith an IES score greater than 30 in family members included female gen-der, being a child of the ICU patient, history of cancer in the ICU patient,higher illness severity scores, feeling that information was incomplete, shar-ing in medical decision making in the ICU, and death of the patient in theICU.

Table 2

Summary of studies assessing the prevalence of posttraumatic stress disorder in family members o survivors of critical illness

Author and

reference no. Study design

Time between

critical illness

and assessment

of posttraumatic

stress disorder

Instrument

used Sample siz Response rate

Prevalence

of posttraumatic

stress disorder

Azoulay,

et al 2005 [5]

Prospective

cohort

90 days IES 284 total,

228 surv ors

62% 33.1% in entire

group; 28.9%

in survivors

Jones, et al

2004 [24]

Randomized

controlled trial

8 weeks, 6 months IES 90 at 8 wee s,

84 at 6 m nths

86% at 8 weeks,

81% at 6 months

49% at 6 months

881

POSTTRAUMATIC

STRESSDISORDER

f

e

iv

k

o

882 KROSS et al

Other smaller studies have shown similar elevated prevalence of PTSD infamily members of ICU survivors. A study by Jones and colleagues [24] wasconducted as part of a randomized controlled trial of a rehabilitation pro-gram after discharge from an ICU. The investigators identified the closestfamily member of 104 recovering ICU patients. PTSD symptoms wereassessed using the IES at 8 weeks and 6 months after ICU discharge. Theinvestigators administered the interviews by telephone and used a cutoffof greater than 19 on the IES as indicative of symptoms of PTSD. Therewas no difference in IES scores between the control and intervention groups.IES scores in both groups were similar at the 8-week and 6-month inter-views. The prevalence of an IES score greater than 19 was 49% at the6-month interview.

One other smaller study looked at PTSD symptoms in family membersduring an ICU stay [25]. This study included 32 family members of ICUpatients. The investigators used the IES to assess PTSD symptoms at twotime points: (1) within 1 week of admission to an ICU and (2) just beforeanticipated ICU discharge. Using a cutoff IES score of 30, they found a prev-alence of PTSD of 81% in family members at the first time point and 59% atICU discharge. These individuals, however, did not meet DSM-IV criteriafor PTSD as the measurements were made within 1 week of admission toan ICU and, therefore, symptoms attributable to the ICU could not havebeen present for the 1-month time period required for the diagnosis ofPTSD by DSM-IV criteria.

In summary, family members of patients who survive critical illness are atincreased risk for PTSD and estimates of the prevalence of PTSD in thesefamily members suggest prevalence in the 25% to 50% range, althoughthese studies are fewer in number. Risk factors are similar to those seen inpatients and include female gender, prior psychiatric illness, and severityof illness of the patients. In addition, poor communication with healthcare providers seems to be associated with increased symptoms of PTSD.

Family members of those who die after critical illness

There are several studies that have looked at PTSD after bereavement notspecific to ICUs. In one such study, widows and widowers were interviewedwith the SCID 2 months after death of a spouse and the investigators foundthat 10% of widows and widowers met criteria for PTSD at that time [26].They found no significant difference between types of death, when compar-ing deaths from chronic illness versus sudden, unexpected death. They didnot discuss hospitalization or ICU admissions, however, before death intheir study.

There are three published studies examining PTSD symptoms in familiesof patients who died (Table 3). The first study to specifically examine theprevalence of PTSD in family members of those who die in an ICU wasthat by Azoulay and colleagues [5] (discussed previously). In that study,

Table 3

Summary of studies assessing the prevalence of posttraumatic stress disorder in family members of patients who die after critical illness

Author and

reference no. Study design

Time between

critical illness and

assess-ment of

posttraumatic

stress disorder Instrument used

Sample

size

Response

rate

Prevalence

of posttraumatic

stress disorder

Azoulay, et al

2005 [5]

Prospective

cohort

90 days IES 56

families

of those

who died

62% 50%

Laurtrette, et al

2007 [27]

Randomized

controlled trial

90 days IES 126 86% 45% in

intervention

group, 67% in

control group

Gries, et al

2008 [28]

Retrospective

cohort

6 to 36

months

PTSD

Checklist-

Civilian

Version

196 71% 15.3%

883

POSTTRAUMATIC

STRESSDISORDER

884 KROSS et al

56 family members of patients who died were interviewed with the IES at 90days after death. Again, the overall prevalence of PTSD, defined by IESgreater than 30, in the total study population (families of those who survivedand those who died) was 33.1%. When looking at the subgroup of familymembers of those who had died, the prevalence of PTSD was 50%. MedianIES scores were higher, as was prevalence of those who had IES scoresgreater than 30, in family members of patients who died in an ICU afterdecisions to limit life-sustaining treatments (60%), and in family memberswho were involved in decisions to limit life-sustaining treatments (80%).

Based on their results, the French group followed up that study with anintervention with the goal of reducing the burden of psychologic symptomsin family members whose loved ones died in an ICU [27]. This studyenrolled family members of 126 patients who died in an ICU in France. Par-ticipants were randomly assigned to the intervention, which included a struc-tured end-of-life care family conference and a bereavement brochure for thefamily, or to usual care. Interviews were conducted using the IES to assessPTSD 90 days after death. They found lower prevalence of PTSD, using IESscores of greater than 30, in the intervention group (45% versus 67%). Theprimary differences between the intervention and control group experienceswere attributed to clinician-family communication. In addition, familymembers in the intervention group spent more time in family conferencesand more of the conference time talking than did the control group.

More recently a study from the United States examined the level of symp-toms of PTSD in family members of 196 patients who died in an ICU usingthe PTSD Checklist-Civilian Version between 6 and 36 months after death[28]. This study found a prevalence of 15.3% and found that the prevalencedecreased with increasing time between death and survey administration.Other factors associated with increased symptoms of PTSD included femalegender, history of medication use for emotions by the family member beforethe critical illness, and a history of the family member seeing a psychiatrist,neurologist, or pain specialist before the critical illness.

In summary, family members of patients who die in an ICU also seem atincreased risk for PTSD in the general population and prevalence estimatesinclude a broad range, from 15% to 65%. Risk factors include female gen-der and prior use of medications for emotions or seeing a psychiatrist beforethe critical illness [28].

Limitations of current research

There are several challenges to studying PTSD in general and particularlyin this population after critical illness. The diagnosis of PTSD requires notonly symptoms of distress but also a precipitating traumatic event. There issignificant comorbidity with PTSD and other psychiatric illnesses [1]. Assuch, it often is difficult to decipher the cause of PTSD symptoms and therelative contribution of PTSD to an individual’s overall level of distress.

885POSTTRAUMATIC STRESS DISORDER

In studies in ICUs in particular, the assessment of PTSD that is related toICU care is challenging. It is difficult to separate the experience of anICU from other aspects of health care and illness. It also is challenging inthis research setting to ensure that the traumatic event of interest is, infact, the ICU experience, and that other traumatic events that may havebeen experienced in the past are not contributing to symptoms at the timeof assessment.

The studies that have looked at PTSD in the setting of critical care havesignificant methodologic limitations. Most of these studies likely overesti-mate the prevalence of PTSD by using only screening tools for diagnosticpurposes. These screening instruments are not designed to definitively iden-tify the presence, absence, or severity of PTSD. Although some of the stud-ies (discussed previously) did follow-up with the SCID, most did not. In thecase of PTSD in family members, all relied on screening tools alone (al-though one recent study used the SCID to assess anxiety, depression, andcomplicated grief in family members of patients who died in an ICU) [29].Most screening tools, including that used most commonly, IES, have notbeen validated in patients who have had critical illness.

Additionally, most of the published studies on this topic have limitedsample sizes. Among studies of ICU survivors, the largest study had 150patients in follow-up whereas most had fewer than 50 patients. There areproblems with low follow-up rates and the potential for nonresponse biasin the samples. Although it is difficult to assess the direction of the bias in-troduced by nonresponse or loss to follow-up, the avoidant symptoms ofPTSD suggest that these studies with significant nonresponse rates may un-derestimate the burden of PTSD.

Finally, there are difficulties generalizing published results to all ICUpatients and their families. In the case of the French study of PTSD in fam-ily members, it is difficult to generalize these results to the United Statespopulation. Prior studies in the United States have shown that family mem-bers of critically ill patients are more satisfied with care when they areinvolved in the decision-making process at the end of life [30], which seemscontrary to the finding by Azoulay and colleagues [5] where there wasincreased prevalence of PTSD in family members involved in end-of-lifedecision making.

Relevance to critical care clinicians

It is important that critical care clinicians be aware that PTSD is a poten-tial morbidity of critical illness for patients who survive their illness and forfamily members of those who survive or die after their critical illness.Although the diagnosis of PTSD can be made only after the incident of crit-ical illness and, therefore, is beyond the scope of practice for most criticalcare clinicians, it may be important to be aware of potential risk factorsfor this morbidity during the course of critical illness and to counsel patients

886 KROSS et al

and families that psychologic distress may occur after an ICU course. In thisway, patients and family members may be more aware of the risk for devel-opment of these symptoms and may be able to seek appropriate counselingand therapy as needed once the critical illness has ended.

Summary

Despite the limitations (described previously), there is consistent evidencein the literature that the prevalence of PTSD in survivors of critical care iselevated above that seen in the general population. Similarly, prevalence forfamily members of survivors and those who died during their ICU stay areelevated. The experience of an ICU, for patients and their families, has thepotential to cause significant psychologic distress. PTSD is an importantsequela of critical illness and should be considered by health care providerswhen providing post-ICU care to survivors and their families and to familiesof those who die in an ICU. Increased recognition, and perhaps screening,may be important as there are effective treatments available [1].

Future studies are needed to further assess the prevalence and burden ofPTSD in these three groups. It would be beneficial not only to screen forPTSD symptoms but also to follow-up positive screening with structuredclinical interviews. With increased awareness of PTSD as a common sequelaof critical illness, future interventions should be investigated to prevent andto treat this adverse outcome in patients and family members. The Frenchstudy by Laurtrette and colleagues [28] is promising, showing decreasedsymptom burden with communication-based intervention for families.Additional interventions should be considered for patients and for familymembers of ICU patients who are not expected to die.

References

[1] Yehuda R. Post-traumatic stress disorder. N Engl J Med 2002;346(2):108–14.

[2] Jackson JC, Hart RP, Gordon SM, et al. Post-traumatic stress disorder and post-traumatic

stress symptoms following critical illness inmedical intensive care unit patients: assessing the

magnitude of the problem. Crit Care 2007;11(1):R27.

[3] Diagnostic and statistical manual of mental disorders, 4th edition: DSM-IV. Washington,

DC: American Psychiatric Association; 1994.

[4] Girard TD, Shintani AK, Jackson JC, et al. Risk factors for post-traumatic stress disorder

symptoms following critical illness requiring mechanical ventilation: a prospective cohort

study. Crit Care 2007;11(1):R28.

[5] Azoulay E, Pochard F, Kentish-Barnes N, et al. Risk of post-traumatic stress symptoms in

familymembers of intensive care unit patients. Am JRespir Crit CareMed 2005;171:987–94.

[6] Griffiths J, Fortune G, Barber V, et al. The prevalence of post traumatic stress disorder in

survivors of ICU treatment: a systematic review. Intensive Care Med 2007;33:1506–18.

[7] CapuzzoM,ValpondiV,Cingolani E, et al. Post-traumatic stress disorder-related symptoms

after intensive care. Minerva Anestesiol 2005;71(4):167–79.

[8] Cuthbertson BH, Hull A, Strachan M, et al. Post-traumatic stress disorder after critical ill-

ness requiring general intensive care. Intensive Care Med 2004;30(3):450–5.

887POSTTRAUMATIC STRESS DISORDER

[9] Kress JP, Gehlbach B, Lacy M, et al. The long-term psychological effects of daily sed-

ative interruption on critically ill patients. Am J Respir Crit Care Med 2003;168(12):

1457–61.

[10] Eddleston JM, White P, Guthrie E. Survival, morbidity, and quality of life after discharge

from intensive care. Crit Care Med 2000;28:2293–9.

[11] Rattray JE, Johnston M, Wildsmit JA. Predictors of emotional outcomes of intensive care.

Anaesthesia 2005;60(11):1085–92.

[12] Jones C, Skirrow P, Griffiths RD, et al. Rehabilitation after critical illness: a randomized,

controlled trial. Crit Care Med 2003;31:2456–61.

[13] Schelling G, Briegel J, Roozendaal B, et al. The effect of stress doses of hydrocortisone dur-

ing septic shock on posttraumatic stress disorder in survivors. Biol Psychiatry 2001;50:

978–85.

[14] Kapfhammer HP, Rothenhausler HB, Krauseneck T, et al. Posttraumatic stress disorder

and health-related quality of life in long-term survivors of acute respiratory distress syn-

drome. Am J Psychiatry 2004;161:45–52.

[15] Stoll C,KapfhammerHP,Rothernhausler HB, et al. Sensitivity and specificity of a screening

test to document traumatic experiences and to diagnose post-traumatic stress disorder in

ARDS patients after intensive care treatment. Intensive Care Med 1999;25(7):697–704.

[16] Schelling G, Stoll C, Haller M, et al. Health-related quality of life and posttraumatic stress

disorder in survivors of the acute respiratory distress syndrome. Crit Care Med 1998;26(4):

651–9.

[17] Schelling G, Stoll C, Kapfhammer HP, et al. The effect of stress doses of hydrocortisone

during septic shock on posttraumatic stress disorder and health-related quality of life in sur-

vivors. Crit Care Med 1999;27(12):2678–83.

[18] Deja M, Denke C, Weber-Carstens S, et al. Social support during intensive care unit stay

might improve mental impairment and consequently health-related quality of life in survi-

vors of severe acute respiratory distress syndrome. Crit Care 2006;10(5):R147.

[19] Nickel M, Leiberich P, Nickel C, et al. The occurrence of posttraumatic stress disorder

in patients following intensive care treatment: a cross-sectional study in a random

sample. J Intensive Care Med 2004;19(5):285–90.

[20] Scragg P, Jones A, Fauvel N. Psychological problems following ICU treatment. Anaesthesia

2001;56(1):9–14.

[21] Shaw RJ, Harvey JE, Nelson KL, et al. Linguistic analysis to assess medically related post-

traumatic stress symptoms. Psychosomatics 2001;42:35–40.

[22] Nelson BJ,Winert CR, BuryCL, et al. Intensive care unit drug use and subsequent quality of

life in acute lung injury patients. Crit Care Med 2000;28:3626–30.

[23] Sukantarat K, Greer S, Brett S, et al. Physical and psychological sequelae of critical illness.

Br J Health Psychol 2007;12(1):65–74.

[24] 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;30:456–60.

[25] Paparrigopoulos T, Melissaki A, Efthymiou A, et al. Shortterm psychologic impact on fam-

ily members of intensive care unit patients. J Psychosom Res 2006;61:719–22.

[26] Zisook S, Chentsova-Dutton Y, Shuchter SR. PTSD following bereavement. Ann Clin Psy-

chiatry 1998;10(4):157–63.

[27] Laurtrette A, DarmonM, Megarbane B, et al. A communication strategy and brochure for

relatives of patients dying in the ICU. N Engl J Med 2007;356:469–78.

[28] Gries C, Engelberg R, Kross E, et al. Risk of PTSD and depression in family members after

the death of a critically ill patient. Am J Respir Crit Care Med 2008;177:A515.

[29] SiegelMD,Hayes E, Vanderwerker LC, et al. Psychiatric illness in the next of kin of patients

who die in the intensive care unit. Crit Care Med 2007;36(6):1722–8.

[30] Gries CJ, Curtis JR, Wall RJ, et al. Family member satisfaction with end-of-life decision

making in the ICU. Chest 2008;133(3):704–12.