Journal of Psychosomatic Research 75 (2013) 539–545

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

Sleep and psychological health during early recovery from critical illness: An observational study☆ Sharon McKinley a,b,⁎, Mary Fien c, Rosalind Elliott a, Doug Elliott a a b c

University of Technology Sydney, Sydney, NSW Australia Northern Sydney Local Health District, Sydney, NSW Australia Agency for Clinical Innovation, Sydney, NSW Australia

a r t i c l e

i n f o

Article history: Received 10 May 2013 Received in revised form 8 August 2013 Accepted 27 September 2013 Keywords: Intensive care Critical illness Sleep Recovery Psychological health

a b s t r a c t Introduction: Intensive care patients often report sleep disruption in ICU and during recovery from critical illness. Objectives: To assess: (i) patients' self-reported sleep quality in ICU, on the hospital ward after transfer from ICU and two and six months after hospital discharge; (ii) whether patients who report sleep disruption in ICU continue to report sleep disruption in recovery and (iii) whether prehospital insomnia, experiences in intensive care, quality of life and psychological health are associated with sleep disruption six months after hospital discharge. Methods: Patients completed self-report measures on sleep quality at five time points: prior to hospitalization, in ICU, the hospital ward, two months and six months after hospital discharge, their intensive care experiences two months after discharge and psychological health and quality of life six months after discharge. Results: Patients (n= 222) were aged (mean ± SD) 57.2±17.2years, 35% female, had mean ICU stay of 5 ± 6 days and BMI of 26 ± 5. Over half the participants (57%) reported poor sleep at six months; for 10% this was at all time points after ICU admission. Prehospitalization insomnia (p=.0005), sleep quality on the ward (p=.006), anxiety (p = .002), and mental (p = .0005) and physical health (p = .0005) were independently associated with poorer sleep quality in survivors six months after ICU treatment. Conclusions: Sleep is a significant issue for more than half of survivors 6 months after ICU treatment. Some influencing factors, such as hospital sleep quality, anxiety, physical health and mental health, are potentially modifiable and should be targeted in recovery programs. © 2013 Elsevier Inc. All rights reserved.

Introduction Critically ill patients often experience sleep disruption in the intensive care unit (ICU). An integrative review of polysomnograph (PSG) data and self-reports revealed significant sleep fragmentation and variable sleep duration [1]. Importantly, even when sleep duration was normal, it was unrestorative and characterised by fragmentation, prolonged stage 1 and 2 and little or no slow wave and rapid eye movement (REM) sleep. These findings were also demonstrated in a recent large study using PSG [2]. Notably, poor sleep quality also occurs during recovery after treatment in ICU, with two surveys suggesting prevalence of 33% [3] to 38% [4]. In follow up of survivors of ICU treatment for the acute respiratory distress syndrome, some patients reported persistent sleep problems

☆ This work was carried out in the Intensive Care Unit, Royal North Shore Hospital, St Leonards, NSW, Australia. ⁎ Corresponding author at: Intensive Care Unit, Level 6, Acute Services Building, Royal North Shore Hospital, St Leonards NSW 2065, Australia. Tel.: +61 2 94632606; fax: +61 2 94632057. E-mail address: [email protected] (S. McKinley). 0022-3999/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpsychores.2013.09.007

disruption six months or more after hospital discharge [5]. Seven patients had abnormal PSG recordings and five had conditioned insomnia (“evidence of conditioned sleep difficulty and/or heightened arousal” [5]) thought to be related to their critical illness [5]. Sleep disturbances six to 12months after ICU have been associated with lower health related quality of life (HRQOL), specifically for bodily pain and mental health (using the Short Form-36 [6] HRQOL instrument) [4]. However it was also reported that the prevalence of poor quality of sleep, 20% overall, did not change from self-reports prior to ICU to those during recovery from critical illness [4]. In a study of the effect of a home-based rehabilitation on physical function and psychological health for survivors of critical illness it was found that more than 30% of patients reported moderate to severe sleeping problems six months after hospital discharge; sleep problems at 26 weeks were independently associated with HRQOL mental health measured on the SF-36 and posttraumatic stress (PTS) symptoms measured with the Impact of Event Scale [7]. Given the few reports on sleep after ICU, there is need to further investigate the sleep of intensive care patients while they are in ICU and during recovery, and factors related to sleep during recovery. The specific aims of this study were to assess: (i) patients' self-reported sleep quality in ICU, on the hospital ward after transfer from ICU and two and six months after hospital discharge; (ii) whether patients

540

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545

who report poor sleep quality in ICU report poor sleep quality in recovery; and (iii) whether prehospital insomnia, psychological health, quality of life and experiences in intensive care are associated with poor sleep quality six months after hospital discharge. Methods Study design This was a prospective observational study of patients from general, cardiothoracic and neurological ICUs of a tertiary referral hospital in Sydney, Australia. Patients completed self report measures on sleep quality prior to current hospitalization (retrosectively), sleep while in ICU, sleep on the hospital ward after transfer from ICU, sleep two and six months after hospital discharge, their intensive care experiences two months after discharge and psychological health and quality of life six months after discharge. Ethics approval was obtained from the Human Research Ethics Committees (HREC) of the Northern Sydney Local Health District and the University of Technology Sydney. Study setting and participants The general, cardiothoracic and neurological ICUs of the study hospital comprised of 36 beds in total and the hospital was a state-wide referral centre for spinal and burns injuries. The ICUs were closed units with an accredited intensive care specialist physician responsible for the management of all patients. The registered nurse (RN) to patient ratio was 1:1 for mechanically ventilated patients and 1:2 for patients requiring high dependency care. The RN performed all nursing care for the patient and was supported by ancillary staff such as patient services assistants and ward clerks. Patients were eligible to participate in the study if they were treated in one of the study ICUs, aged 18years or more, had an ICU length of stay (LOS) of ≥2 nights, were able to give informed consent to participate, were able to complete the study instruments in English, had adequate vision and hearing to complete study instruments and had been cleared for discharge from ICU to a hospital ward. Patients were excluded if they had a known history or evidence of sleep disorder (e.g. obstructive sleep apnea), if significant treatment limitations were in place (e.g. a medical order not to escalate inotropic support) and/or were receiving palliative care or were in isolation nursing for multiresistant organisms. All participants gave written informed consent using the HREC-approved information sheet and consent form after they had been designated ready for ICU discharge; continuing consent was sought verbally at each data collection time point after ICU discharge. Measures and data collection In ICU and prior to transfer to a hospital ward patients completed the Insomnia Severity Index (ISI) [8] reporting on their prehospital sleep quality, a 1–10 likert scale on the quality of sleep at home [9] and the Richards Campbell Sleep Questionnaire (RCSQ) [10] on their previous night's sleep in ICU. After 1–2 nights on the hospital ward they again completed the RCSQ on their previous night's sleep. Two months after hospital discharge former patients completed the Intensive Care Experience Questionnaire (ICEQ) [11] and the Pittsburgh Sleep Quality Index (PSQI) [12]. Six months after hospital discharge participants again completed the PSQI, the Depression, Anxiety and Stress Scales (DASS) [13], the Posttraumatic Stress Checklist for a Specific event (PCL-S) [14] and the SF-36 quality of life survey [15]. The Insomnia Severity Index [8]. The ISI contains seven items scored on a 0–4 scale for insomnia severity, satisfaction with sleep pattern, interference with daily function, impairment of quality of life and distress about sleep problems. Scores range from 0 to 28, with higher scores reflecting worse sleep quality and scores of 15 or more indicative of

moderate to severe insomnia [8]. In the current study Chronbach's alpha was .90. The Richards-Campbell Sleep Questionnaire [10]. The RCSQ consists of five100 mm visual analogue scales for sleep depth, falling asleep, wakefulness, going back to sleep and overall sleep quality. Scores of the five scales are averaged to obtain one score, with higher scores indicating better sleep. The RCSQ was validated using PSG in nonventilated, medical critical care patients, with the average total score having a correlation of 0.58 with the sleep efficiency index from PSG [10]. In the current study Chronbach's alpha values were .91 in ICU and .90 in the ward. The Sleep in the Intensive Care Unit Questionnaire [9]. Only the first two items of the 7-item SICQ are reported here: ‘rate the overall quality of your sleep at home’ (completed in ICU) and ‘rate the overall quality of your sleep in ICU’ (completed in the ward). Subjective sleep quality is assessed using a 1–10 likert scale, with higher scores indicating better sleep. At the time of data collection in the ICU the Richmond AgitationSedation Scale (RASS) [16] was used to assess the level of sedation, the Faces Anxiety Scale [17] was used to assess state anxiety and patients rated pain intensity on a scale of 1 to 10. The Intensive Care Experience Questionnaire (ICEQ) [11] has 24 items with 1–5 likert responses on agreement with and frequency of occurrence of experiences in an ICU, plus three open-ended questions. The ICEQ has four domains: 1) Awareness of surroundings (nine items, scores 9–45); 2) Frightening experiences (six items, scores 6–30); 3) Recall of experiences (five items, scores 5–25); and 4) Satisfaction with care (four items, scores 4–20). Higher scores indicate greater awareness, more Frightening experiences, better Recall of experiences and greater Satisfaction with care. It has satisfactory reliability and demonstrated concurrent and predictive validity in former ICU patients [11]. In the current study Chronbach's alpha values were .83, .73, .76 and .56 respectively for the four domains. The Pittsburgh Sleep Quality Index (PSQI) [12] consists of 10 questions with a mixture of short answers and use of a likert scale of 1–5 to assess sleep quality and habits over the preceding month. Total scores range from 0 to 21. Higher scores indicate worse sleep and scores N 5 are interpreted as poor sleep quality. The PSQI has been reported to provide a specific and sensitive measure of sleep quality in community-dwelling respondents and was rated highly in a review paper for reliability, validity and responsiveness [18]. In the current study Chronbach's alpha values were .71 at two months and .79 at six months. The Depression Anxiety and Stress Scales instrument (DASS-21) [13,19] contains 21 items, seven each for depression, anxiety and stress responded to on a 4-point 0–3 scale, yielding scores from 0 to 21 which are doubled to correspond to the 0–42 range of the original DASS. The scales have established validity and reliability [20]. In the current study Chronbach's alpha values were .91 for stress, .74 for anxiety and .92 for depression. The Post-traumatic Stress Disorder Checklist for a Specific event (PCL-S) [14,21] contains 17 questions that correspond to the Diagnostic and Statistical Manual-IV (DSM-IV) criteria for PTSD exhibited as reexperiencing, avoidance and increased arousal [22]. Respondents rate how much they have been bothered by a symptom in the last month on a 5-point scale (“not at all” to “extremely”). Total scores range from 17 to 85; higher scores are worse. In this study total severity scores are reported on a continuous scale. The PCL-S has established reliability and validity [14,23]. In the current study Chronbach's alpha was .92. The Medical Outcomes Trust Short Form-36 (SF-36) health survey [6] is a well validated, generic health-related quality of life (HRQOL) instrument with eight scales: Physical functioning, Effect of physical function on role (Role-physical), Bodily pain, General health, Vitality, Social functioning, Effect of emotional health on role (Role-emotional) and Mental health. The eight scales are used to calculate Physical

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545

(PCS) and the Mental Component Summary scores (MCS), with the PCS heavily weighted by the scales related to physical functioning and the MCS heavily weighted by scales related to mental health function. The PCS and MCS are standardised by weighting the eight individual dimension scores using coefficients derived from studies in the US general population. Scores range from 0 to 100 for the subscales and the component summary scores. The instrument's use in survivors of critical illness is reported widely, and it is recommended for studying this population [24]. Procedure All patients in the ICUs were screened daily for eligibility on weekdays from June, 2010 to February, 2011 using patient records and in consultation with the nurses caring for the patients. Eligible patients were approached in the ICU prior to transfer to the hospital wards, given verbal and approved written information about the study (why they were being approached, the aims of the study, what it would involve for them and their freedom to decline without any effect on their continued care). Patients wishing to participate were assisted in completing the ISI on their prehospital sleep at home and the RCSQ on their sleep on the previous night; they responded to questions about their current levels of pain on a 1–10 scale and state anxiety on the Faces Anxiety Scale. Clinical and demographic data recorded were age, sex, height, weight and diagnostic category on admission to ICU. One to two nights after transfer to the ward, patients completed the RCSQ on their previous night's sleep and the 10-point scale (10 = excellent) from the SICQ [9] on their sleep at home and their sleep in ICU. After hospital discharge, the two month and six month questionnaires were sent to participants by post with return prepaid envelopes. If questionnaires were not returned within two weeks of the due date, participants were contacted by telephone as a reminder and assisted in completing the questionnaires by phone if they preferred that to mailing them back. Data analysis Descriptive statistics, mean (standard deviation, SD), median (range; interquartile range, IQR) and frequencies and percentages were used to describe the sample and the results of sleep self-reports, intensive care experiences and psychological health. Bivariate comparisons of factors related to sleep quality six months after discharge were made with t-tests, Mann–Whitney U tests and Chi2 or Fisher's exact tests. Factors found in bivariate comparisons to be significantly (p b 0.05) associated with sleep quality were entered into a multiple linear regression (enter method) with PSQI scores (square root transformed) at six months as the dependent variable. Results Two hundred and twenty-two patients agreed to participate in the study and provided data in the ICU. Data were provided on the hospital ward by 199 (90%), at two months by 183 (82%) and at six months by 176 (79%). Fig. 1 summarizes the numbers of patients who satisfied the inclusion criteria, reasons for exclusion and numbers lost to follow up, deceased and withdrawn. The clinical and demographic characteristics of participants on enrollment are shown in Table 1. The average age approached 60 years, the ratio of males to females was 2:1, more than two thirds were postoperative and the most common diagnoses on admission were cardiovascular and neurological. Sedation scores (RASS) showed that almost all patients were interactive, calm and cooperative at the time of providing consent and initial data collection in ICU. Sleep Patients' self-reports of sleep quality are shown in Table 2a for continuous data and in 2b for data categorized as poor or good. Patients retrospectively rated quality of sleep at home at approximately 7/10 while in ICU and again in the ward. Mean prehospital insomnia scores were low but 18% of patients reported features of moderate or severe clinical insomnia. Mean RCSQ scores were slightly less than 50 in ICU and slightly greater than 50 in

541

the ward. A cut point of 70/100 on the RSCQ was used to categorise good and poor sleep, based on an average score of prehospital sleep quality of 7/10 on the SICQ. More than half of patients (52%) reported sleep quality as poor in ICU and the ward. At two months after hospital discharge, 62% of patients reported poor sleep quality, and this dropped slightly to 57% at six months; 40% reported poor sleep at both two and six months after discharge. Poor sleep was evident in 7% of patients across all five time points: clinical insomnia prior to admission and poor sleep quality in ICU, in the ward and at two and six months after hospital discharge (Table 2b). This proportion was higher (10%) for each time point after hospitalization. Of the 41 patients who reported clinical insomnia prior to their hospitalization, 32 (14.4%) first reported poor sleep in ICU, five (2.3%) first in the ward and three (1.4%) first at two months; no patients with prehospital insomnia reported poor sleep first at six months. Intensive care experiences The four domains of the ICEQ based on patients' reports on their experiences in intensive care two months after discharge are shown in Table 3. The distributions of data for Awareness of surroundings and Recall of experiences trended towards the higher end of the scale, that for Frightening experiences trended towards the lower end of the scale and that for Satisfaction with care was evenly distributed. Depression, anxiety, stress, posttraumatic stress symptoms and health related quality of life The mean scores for depression, anxiety, stress and posttraumatic stress symptoms six months after discharge are shown in Table 4. Depression, anxiety and stress scores in this sample exceeded normative sample means, which are 6.34, 4.70 and 10.11 respectively [20]. The mean posttraumatic stress score was below the suggested cut point of 44 for the specific event PCL [14]. The mean (SD) score for the Physical Component Summary on the SF-36 at six months after discharge was 41.6 (11.9) and for the Mental Component Summary was 48.0 (11.3). Factors associated with quality of sleep at six months Comparisons were made between mean (t-test), median (Mann–Whitney U test) or frequency (Chi2 or Fisher's exact test) values for factors potentially associated with sleep according to six month PSQI scores were N5 (poor) and ≤5 (good). There were no significant differences for age, BMI, anxiety, length of mechanical ventilation, length of ICU stay, length of hospital stay, gender, reason for admission to ICU, diagnostic category on admission, RCSQ in ICU, ICEQ Frightening experiences, ICEQ Recall of experiences or ICEQ Satisfaction with care. Factors for which there were differences according to six month PSQI scores are shown in Table 5. Patients were more likely to have poor sleep quality at six months if they had higher prehospital insomnia scores, poorer sleep quality at home, poorer sleep quality on the hospital ward, lower Awareness of surroundings on the ICEQ, lower physical and mental HRQOL on the SF-36, and higher pain scores and slight agitation at the time of data collection in ICU. Depression, anxiety, stress and posttraumatic symptoms at six months were also associated with poorer sleep. The above significant factors (p b .05) and age were entered into the multiple linear regression analysis (enter method) with PSQI scores at six months (square root transformed continuous scores) as the dependent variable to identify independent associations with sleep quality six months after discharge (Table 6). Poorer sleep quality at six months was independently associated with greater prehospital clinical insomnia, lower self-reported sleep quality on the hospital ward, higher anxiety at six months and lower physical and mental HRQOL.

Discussion In this study we set out to add to the small body of knowledge on associations between quality of patients' sleep during critical illness requiring treatment in intensive care and sleep quality during recovery. Associations of sleep quality six months after hospital discharge with patients' experiences in ICU and psychological health and quality of life at six months were investigated. Retrospective reports of sleep at home prior to the index hospitalization were also studied to control for unrecognized pre-existing sleep disturbances on the main results of the study. Almost three-quarters of patients reported poor sleep quality on their last night in ICU, relative to the average overall quality of sleep of 7 out of 10 at home, and almost 70% reported poor sleep quality after one to two nights on the ward. The average RCSQ score in ICU was similar but a little lower than those reported in other recent studies of 300 in medical ICU patients in the United States [25], 51 in surgical ICU patients in Spain [26] and 52 in surgical ICU patients in Taiwan [27]. More than half of patients reported poor sleep both in ICU and on the ward. After hospital discharge, approximately 60% of patients reported

542

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545

Fig. 1. Flow chart of screening, recruitment and patient retention in the study.

poor sleep at two months and approximately 60% at six months, with 40% reporting poor sleep at both posthospital time points. Only one tenth of patients reported poor sleep at all time points after hospitalization: in ICU, on the hospital ward, at two months and at six months after discharge. Thus the notion of continuity of a problem of sleep disturbance commencing in ICU and continuing into recovery appears to be evident in a minority of this sample. This observation is influenced by the finding that some study participants reported having moderate to severe clinical insomnia prior to their hospitalization. Almost one-fifth of patients self-reported prehospital clinical insomnia, after they had recovered sufficiently to be ready for transfer from ICU to a hospital ward. The potential for recall bias is acknowledged but patients were interactive, calm and cooperative when making reports retrospectively, as they always are with the ISI instrument used to collect these data. This finding is consistent with the report of Orwelius et al. that the rate of sleep quality during recovery was similar to that prior to ICU treatment [4], with 13% of patients retrospectively

reporting poor sleep prior to ICU and six months after ICU, while 70% reported good sleep at both times. This finding affects the interpretation of the results of this study and others on the problem of sleep quality in ICU patients. Only a small minority of patients had clinical insomnia prior to hospitalization and at all subsequent time points but there was a difference in the ISI scores between patients with poor or good sleep at six months, and ISI scores were independently associated with poorer sleep at six months. Insomnia is relatively common in the community with a prevalence of 10–20% [28] and is associated with problems such as depression, which may be both a risk factor and a consequence [28]. In the present study we were able to statistically control for its associations with sleep at six months after critical illness and psychological health and quality of life at six months, again suggesting that prehospital sleep quality should be included in future studies of sleep in intensive care patients [4]. Patients' experiences in intensive care, specifically lower Awareness of surroundings, were associated with poorer sleep quality at six months

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545 Table 1 Demographic and clinical characteristics of participants on enrollment (n = 222) Mean

SD

57.2 26 2.5 2.3

17.2 4.4 2.9 1.1

Median

Minimum–maximum

0.5 3 13

0–32 2–56 2–133

N

%

77

35

156

70

Diagnostic category Cardiovascular Respiratory Gastrointestinal Neurological Trauma Sepsis Other

81 13 20 57 24 5 22

37 6 9 26 11 2 10

RASSb score −1 0 1

13 208 1

6 94 1

Age (years) BMIa (kg/m2) Pain (0–10) Anxiety (1–5)

Length of mechanical ventilation (days) ICU length of stay (days) Hospital length of stay (days)

Gender Female Operative Yes

a

BMI — body mass index. b RASS — Richmond Agitation Sedation Scale: −1 = drowsy; 0 = alert and calm; +1 = restless.

Table 2b Sleep self-reports prehospital, in ICU, in the ward and at two and six months after discharge categorized by cut points N

%

Clinical insomnia prehospital (ISI ≥ 15) No or subclinical insomnia Clinical insomnia — moderate or severe

181 41

82 18

RCSQ in ICUa Poor (b70) Good (≥70)

162 60

73 27

RCSQ in warda Poor (b70) Good (≥70)

136 63

68 32

RCSQ in ICU and ward Poor in ICU and ward Poor in ICU or ward Good in ICU and ward

104 77 18

52 39 9

PSQI total sleep quality at two months Poor (N5) Good (≤5)

109 66

62 38

PSQI total sleep quality at six months Poor (N5) Good (≤5)

101 75

57 43

PSQI at two and six months Poor at two and six months Poor at two or six months Good at two and six months

73 64 45

40 35 25

Sleep quality over time Poor — prehospital, ICU, ward, two months and six months Poor — ICU, ward, two months and six months Poor post-hospital only

15 23 35

7 10 16

a

in bivariate analysis, a finding not previously reported. The scores on the ICEQ contrast with the domain scores from the original studies by the instrument's developer (personal communication J Rattray, 23rd February, 2013). In two of Rattray's earlier studies [11,29] scores were worse for Awareness of surroundings (means 23.2 and 30.2), Frightening experiences (20.4 and 15.7) and Recall of experiences (13.2 and 14.2), and similar for Satisfaction with care (51.1 and 15.4). In these earlier studies in the United Kingdom patients completed the ICEQ in hospital just prior to discharge, whereas in this study the ICEQ was completed two months after hospital discharge. It is possible that the better scores for three of the domains in this study relate to the longer time elapsed since the patients' experiences in ICU, but they could also be related to variations in patient clinical characteristics and practices such as sedation regimens over time and between countries. Depression, anxiety and stress at six months were associated with poorer sleep in the bivariate analyses. Scores were higher for the patients in this study than the norms for the DASS instrument, but these higher levels of psychological morbidity are common for Table 2a Self-reports of sleep quality prehospital, in ICU, in the ward and at two and six months after discharge Mean a

ISI total (at home) (0–28) Quality of sleep at home (SICQ) (1–10)b Report in ICU Report in ward RCSQ in ICU (0–100)b Quality of sleep in ICU (SICQ) (1–10)b RCSQ in ward (0–100)b Total PSQI at home (0–21)b Two months Six months a b

Lower scores are better. Higher scores are better.

SD

Minimum–maximum

7.7

7.1

0–28

7.0 7.4 47.2 4.2 54.3

2.4 2.1 28.1 2.1 24.4

1–10 2–10 0–100 1–10 0–99

7.9 7.4

4.5 4.7

1–20 0–21

543

Based on a score 7/10 for sleeping at home on the SICQ.

recovering ICU patients [29–31]. Posttraumatic stress symptoms are also commonly reported in former critically ill patients [32,33]. While the average PTS symptom score was lower in this study than the suggested cut point for the PCL instrument [14], higher scores were associated with poorer sleep in the bivariate analysis but not in the multivariate analysis. The results found for the Physical and Mental Component Summary scores of the SF-36 HRQOL are almost the same as those in two recently reported studies of rehabilitation after critical illness, one in Australia that included this study site and similar ICUs [34] and one in a recent study from the United Kingdom [35]. Both the PCS and MCS were independently associated with quality of sleep six months after ICU treatment in this study. Another recent, small study found that insomnia six months after ICU was associated with SF-36 PCS, but not MCS [36]. It appears that future investigations of sleep or HRQOL after ICU treatment should include both constructs to better elucidate the magnitude and direction of the relationship. The rate of poor sleep quality in three-fifths of respondents at six months contrasts with previous studies in which lower rates were reported [3,4], but is similar to our previous study [7]. Poorer sleep quality at six months after hospitalization in ICU was independently associated with prehospital insomnia, poorer quality of sleep on the hospital ward after ICU, higher anxiety at six months and lower physical and Table 3 Patients' reports on their experiences in intensive care (ICEQ) two months after discharge (N = 183)

a

Awareness of surroundings (9–45) Frightening experiences (6–30)b Recall of experiences (5–25)a Satisfaction with care (4–20)a a b

Higher scores are better. Lower scores are better.

Mean

SD

Minimum–maximum

36.9 13.0 17.2 14.6

5.8 4.7 4.4 2.9

12–45 6–28 5–25 8–20

544

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545

Table 4 Depression, anxiety, stress (DASS) and posttraumatic stress symptoms (PCL-S) six months after discharge (N = 176)a

Depression (0–42) Anxiety (0–42) Stress (0–42) Posttraumatic stress symptoms (17–85) a

Mean

SD

Minimum–maximum

9.17 5.6 9.2 26.4

9.7 6.3 9.7 10.7

0–42 0–32 0–42 1–67

Higher scores are worse on all scales.

mental HRQOL at six months. Some of these results are consistent with previously reported results of studies in the same or similar populations, such as a relationship with prehospital sleep [4,25] and HRQOL [7,37]. Importantly, the findings of this study also extend evidence on the topic, demonstrating that sleep quality six months after discharge was independently associated with sleep quality during hospitalization, as well as anxiety at six months. This is consistent with our previous study where poor sleep quality at six months after discharge was associated with stress and PTSD symptoms [7]. Methodological strengths of this study include a relatively high sample size for this topic area, the repeated measures design and use of a comprehensive battery of valid and reliable self-report instruments measuring sleep quality and quantity and other psychological constructs potentially related to sleep. Limitations include single-centre study setting and potential for reporting bias in the reliance of selfreport instruments, particularly for prehospital insomnia severity, which is known to be over-reported [28], but in this study may have been underestimated in patients' retrospective reports. In reporting the psychological data, no inferences have been made that the instrument scores are diagnostic, only that symptoms were reported. Formal assessment for delirium was not carried out prior to obtaining patient consent and self-report data in the ICU, so there is the possibility that in some patients the data obtained in ICU were unreliable. However all patients approached were cleared for discharge from ICU. Approximately 14% of patients screened were assessed as unable to give informed consent using a systematic approach, less than the rate of delirium detected in formal assessment of delirium in Australian ICUs [38]. We therefore may have underestimated patients' ability to provide reliable self-reports on the questionnaires, making the possibility of unreliable and biased information unlikely. Although some patients had long lengths of ventilation and ICU stay, the medians were low;

Table 5 Bivariate comparisons of factors related to sleep quality six months after discharge (n = 176)

ISI total score (0–28) Home sleep quality (1–10) RCSQ on ward — total (0–100) Awareness of surroundings (9–45) SF-36 PCS (0–100) SF 36 MCS (0–100)

Pain (0–10) Depression Anxiety Stress PCL total

Agitation–sedation (RASS) score −1 0 a b c

t-test. Mann–Whitney U test. Fisher's exact test.

PSQI N 5 (poor)

PSQI ≤ 5 (good)

Mean

SD

Mean

SD

pa

9.1 6.6 52.4 35.7 38.3 43.9

7.3 2.3 24.6 6.1 11.0 11.5

5.5 7.9 60.3 38.6 46.1 53.4

5.5 1.9 23.6 5.1 11.5 8.6

.0005 .0005 .044 .002 .0005 .0005

Median

IQR

Median

IQR

pb

2 10 6 10 26.5

0–5 4–20 2–12 4–18 21–35

0 2 2 2 20

0–4 0–8 0–4 0–8 17–26

.043 .0005 .0005 .0005 .0005

N

%

N

%

pc

1 100

1.0 99.0

8 66

10.8 89.2

.005

Table 6 Independent associations with sleepa six months after discharge - multiple linear regression

ISI total scorec RCSQ ward Anxiety six months SF-36 PCS SF-36 MCS

b

B

SE of B

Beta

.029 −.005 .031 −.023 −.028

.008 .002 .012 .005 .007

.232 −.136 .220 −.300 −.351

p

95% CI of B

.0005 .025 .011 .0005 .0005

.014–.044 −.009 – −.001 .007–.055 −.033– −.012 −.043– −.013

Adjusted R square = .476. The analysis was adjusted for age, RASS score, pain score, RCSQ in ICU, ICEQ Awareness of surroundings, depression and stress at six months, and PCL total score. a Dependent variable – Total PSQI at six months – square root transformed. b Standardised coefficient. c Insomnia Severity Index prehospitalisation.

patients in this study may therefore not be representative of longerterm ICU patients. In conclusion, this study shows that poor sleep quality is experienced by the majority of patients in ICU, on the hospital ward after transfer from ICU, and at two and six months after hospital discharge. A small proportion of patients with prehospital insomnia had poor sleep quality at each time point, and a further small proportion who first experienced problems sleeping in ICU also reported problems in the ward and after discharge. Poor sleep quality six months after discharge was related to prehospital insomnia and poor sleep quality in hospital. Importantly, associations were also noted with state anxiety and poorer quality of life during recovery from critical illness. These issues require confirmation in further investigations in order to develop strategies for improving the recovery of survivors of critical illness treated in intensive care. Conflict of interest The authors declare that none has any conflicts of interest to declare. References [1] Elliott R, McKinley S, Cistulli P. The quality and duration of sleep in the intensive care setting: an integrative review. Int J Nurs Stud 2011;48:384–400. [2] Elliott R, McKinley S, Cistulli P, Fien M. Characterisation of sleep in intensive care using 24-hour polysomnography: an observational study. Crit Care Mar 18 2013;17:R46. [3] Eddleston JM, White P, Guthrie E. Survival, morbidity, and quality of life after discharge from intensive care. Crit Care Med Jul 2000;28:2293–9. [4] Orwelius L, Nordlund A, Nordlund P, Edell-Gustafsson U, Sjoberg F. Prevalence of sleep disturbances and long-term reduced health-related quality of life after critical care: a prospective multicenter cohort study. Crit Care 2008;12:R97. [5] Lee CM, Herridge MS, Gabor JY, Tansey CM, Matte A, Hanly PJ. Chronic sleep disorders in survivors of the acute respiratory distress syndrome. Intensive Care Med Feb 2009;35:314–20. [6] Ware JE. SF-36 health survey update. Spine 2000;25:3130–9. [7] McKinley S, Aitken LM, Alison JA, King M, Leslie G, Burmeister E, et al. Sleep and other factors associated with mental health and psychological distress after intensive care for critical illness. Intensive Care Med Feb 9 2012;38:627–33. [8] Bastien CH, Vallieres A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med Jul 2001;2:297–307. [9] Freedman NS, Kotzer N, Schwab RJ. Patient perception of sleep quality and etiology of sleep disruption in the intensive care unit. Am J Respir Crit Care Med Apr 1999;159:1155–62. [10] Richards KC, O'Sullivan PS, Phillips RL. Measurement of sleep in critically ill patients. J Nurs Meas Fall-Winter 2000;8:131–44. [11] Rattray J, Johnston M, Wildsmith JA. The intensive care experience: development of the ICE questionnaire. J Adv Nurs Jul 2004;47:64–73. [12] Buysse DJ, Reynolds III CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989;28:193–213. [13] Lovibond PF, Lovibond SH. The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behav Res Ther 1995;33:335–43. [14] Blanchard EB, Jones-Alexander J, Buckley TC, Forneris CA. Psychometric properties of the PTSD Checklist (PCL). Behav Res Ther Aug 1996;34:669–73. [15] Ware JE, Snow KK, Kosinski M. SF-36 version 2 health survey: manual and interpretation guide. Lincoln: Quality Metric Incorporated; 2000. [16] Sessler CN, Gosnell MS, Grap MJ, Brophy GM, O'Neal PV, Keane KA, et al. The Richmond Agitation-Sedation Scale: validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med Nov 15 2002;166:1338–44.

S. McKinley et al. / Journal of Psychosomatic Research 75 (2013) 539–545 [17] McKinley S, Madronio C. Validity of the Faces Anxiety Scale for the assessment of state anxiety in intensive care patients not receiving mechanical ventilation. J Psychosom Res May 2008;64:503–7. [18] Devine EB, Hakim Z, Green J. A systematic review of patient-reported outcome instruments measuring sleep dysfunction in adults. Pharmacoeconomics 2005;23: 889–912. [19] Henry JD, Crawford JR. The short-form version of the Depression Anxiety Stress Scales (DASS-21): construct validity and normative data in a large non-clinical sample. Br J Clin Psychol Jun 2005;44:227–39. [20] Lovibond SH, Lovibond PF. Manual for the Depression Anxiety Stress scales. 2nd ed. University of New South Wales: Psychology Foundation Monograph; 1995. [21] Weathers FW, Litz BT, Herman DS, Huska JA, Keane TM. The PTSD checklist: reliability, validity, & diagnostic utility. Annual Meeting of the International Society for Traumatic Stress Studies; October; San Antonio, Texas; 1993. [22] American Psychiatric Association Task Force on DSM-IV. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4th ed. Washington, DC: American Psychiatric Association; 1994. [23] Ventureyra VA, Yao SN, Cottraux J, Note I, De Mey-Guillard C. The validation of the Posttraumatic Stress Disorder Checklist Scale in posttraumatic stress disorder and nonclinical subjects. Psychother Psychosom Jan-Feb 2002;71:47–53. [24] Angus DC, Carlet J. Surviving intensive care: a report from the 2002 Brussels Roundtable. Intensive Care Med Mar 2003;29:368–77. [25] Kamdar BB, King LM, Collop NA, Sakamuri S, Colantuoni E, Neufeld KJ, et al. The effect of a quality improvement intervention on perceived sleep quality and cognition in a medical ICU. Crit Care Med Mar 2013;41:800–9. [26] Nicolas A, Aizpitarte E, Iruarrizaga A, Vazquez M, Margall A, Asiain C. Perception of night-time sleep by surgical patients in an intensive care unit. Nurs Crit Care Jan-Feb 2008;13:25–33. [27] Li SY, Wang TJ, Vivienne Wu SF, Liang SY, Tung HH. Efficacy of controlling night-time noise and activities to improve patients' sleep quality in a surgical intensive care unit. J Clin Nurs Feb 2011;20:396–407. [28] Buysse DJ. Insomnia. JAMA Feb 20 2013;309:706–16.

545

[29] Rattray J, Crocker C, Jones M, Connaghan J. Patients' perceptions of and emotional outcome after intensive care: results from a multicentre study. Nurs Crit Care Mar 2010;15:86–93. [30] Davydow DS, Desai SV, Needham DM, Bienvenu OJ. Psychiatric morbidity in survivors of the acute respiratory distress syndrome: a systematic review. Psychosom Med May 2008;70:512–9. [31] Davydow DS, Gifford JM, Desai SV, Bienvenu OJ, Needham DM. Depression in general intensive care unit survivors: a systematic review. Intensive Care Med May 2009;35:796–809. [32] Davydow DS, Gifford JM, Desai SV, Needham DM, Bienvenu OJ. Posttraumatic stress disorder in general intensive care unit survivors: a systematic review. Gen Hosp Psychiatry Sep-Oct 2008;30:421–34. [33] Griffiths J, Fortune G, Barber V, Young J. The prevalence of post traumatic stress disorder in survivors of ICU treatment: a systematic review. Intensive Care Med 2007;33:1506–18. [34] Elliott D, McKinley S, Alison J, Aitken LM, King M, Leslie GD, et al. Health-related quality of life and physical recovery after a critical illness: a multi-centre randomised controlled trial of a home-based physical rehabilitation program. Crit Care Jun 9 2011;15:R142. [35] Cuthbertson BH, Rattray J, Campbell MK, Gager M, Roughton S, Smith A, et al. The PRaCTICaL study of nurse led, intensive care follow-up programmes for improving long term outcomes from critical illness: a pragmatic randomised controlled trial. BMJ 2009;339:b3723. [36] Myhren H, Ekeberg O, Stokland O. Health-related quality of life and return to work after critical illness in general intensive care unit patients: a 1-year follow-up study. Crit Care Med Jul 2010;38:1554–61. [37] Parsons EC, Kross EK, Caldwell ES, Kapur VK, McCurry SM, Vitiello MV, et al. Post-discharge insomnia symptoms are associated with quality of life impairment among survivors of acute lung injury. Sleep Med Sep 2012;13:1106–9. [38] Elliott D, Aitken LM, Bucknall TK, Seppelt IM, Webb SAR, Weisbrodt L, et al. Patient comfort in the intensive care unit: a muiticentre, binational point prevalence study of analgesia, sedation and delirium management. Crit Care Resusc 2013;15(3):213–9.