The long-term psychological effects of daily sedative interruption on critically ill patients
This review may be edited
Posted By: Sharon McKinley
Posted Date: 21st July, 2004
Title: The long-term psychological effects of daily sedative interruption on critically ill patients
Authors: Kress JP, Gehlbach B, Lacy M, Pliskin N, Pohlman AS, Hall JB.
Reference: Am J Respir Crit Care Med. 2003 Dec 15;168(12):1457-61
Link: Click here for a direct link to the paper. A password may be required for access to fulltext.
Abstract: Critically ill patients often receive sedatives, which may delay liberation from mechanical ventilation and intensive care unit discharge. Daily interruption of sedatives alleviates these problems, but the impact of this practice on long-term psychological outcomes is unknown. We compared psychological outcomes of intensive care unit patients undergoing daily sedative interruption (intervention) with those without this protocol (control). Assessments using (1) the Revised Impact of Event Scale (evaluates signs of posttraumatic stress disorder [PTSD]), (2) the Medical Outcomes Study 36 item short-form health survey, (3) the State-Trait Anxiety Inventory, (4) the Beck Depression Inventory-2, (5) and the Psychosocial Adjustment to Illness score (overall quality of adjustment to current or residual effects of illness) were done by blinded observers. The intervention group had a better total Impact of Events score (11.2 vs. 27.3, p=0.02), a trend toward a lower incidence of PTSD (0% vs. 32%, p=0.06), and a trend toward a better total Psychosocial Adjustment to Illness score (46.8 vs. 54.3, p=0.08). We conclude that daily sedative interruption does not result in adverse psychological outcomes, reduces symptoms of PTSD, and may be associated with reductions in posttraumatic stress disorder.
Are the Results Valid?
1. Was the assignment of patients to treatments randomized? ( Was allocation concealment maintained?)
Yes, mostly, it appears. There were 2 groups of patients entered into this longer term followup of an earlier study of the in-hospital outcomes of the intervention. The first group was patients who were randomised to the ‘daily wake up’ or ‘usual treatment’ group in the original study. The second group could have been patients who were in the ICU at the same time (contemporaneous) who were not entered into the original study, who therefore would not have been randomised. However the authors state that they recruited up to the time of the publication of the first study, and that randomisation was difficult in their ICU after the short term benefits of ‘daily wake up’ were known. This is ambiguous and important as there was serious imbalance in numbers in the control and intervention groups between the original and the contemporaneous patients in this study.
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
2a. Was followup complete?
No. Of the 105 patients screened, only 30% overall were included in this follow up, mainly because of deaths and patients not being able to be reached rather than patients choosing not to be in the study.
2b. Were patients analyzed in the groups to which they were randomized?
3. Were patients, health workers, and study personnel blind to treatment?
Patients were blinded to which study group they had been in, as were the psychologists performing the psychological evaluation. It may have been evident to the families of patients’ who were being woken up daily that this was occurring, but the informed consent of patients or family proxies was waived, so patients probably did not know which they had been in during their treatment in ICU. The staff caring for the patients during their sedation administration in ICU were not told what the outcome measures were in the original study in an attempt at blinding group allocation, but it would have been obvious which patients were being woken up daily. This could have influenced their care of the patients in the different groups.
4. Were the groups similar at the start of the trial?
The groups were similar at baseline, except for duration of mechanical ventilation and length of ICU stay, both of which were improved outcomes in the original study.
5. Aside from the experimental intervention, were the groups treated equally?
This is not stated in this paper or that on the original study, except that there were standard protocols for the infusion of sedative midazolam or propofol, depending on which agent the patient was randomised to. Adherence to the protocols is not reported. As the staff caring for the patients in ICU likely knew which group they were in, this is a possible source of unconscious bias. Perhaps more importantly, patients who appeared to be awake and communicating in the ‘daily wake up’ group are likely to have received more communication and information than those in the control group. This could in turn effect their recollections of being in ICU and response to it afterwards.
What are the Results?
1. How large was the treatment effect?




% difference

Recall of ICU



+ 0.8

Recall of awakening in ICU



+ 26.3

Impact of Events Scale


(range 0-60, lower better)



- 41.0

Avoidance subscale



- 7.4

Intrusive thoughts subscale



- 8.2

PTSD diagnosis

(DSM IV criteria)



- 31.6

Quality of life (SF-36) scores were lower than population norms but similar to others reported in former ICU patients and similar between groups. The lowest scores were for general health, physical functioning and the effect of physical function on role.

Anxiety (Spielberger STAI) was common and moderate to high in both groups.

Depression (Beck Inventory) was mild and similar in both groups. Psychosocial adjustment to illness (PAIS, range 0-100%, lower better), with total scores of 54.3±10.1 in the intervention group and 46.8±12.7 in the control group, and with trends to lower scores in most of the seven subscales.

2. How precise was the estimate of the treatment effect?
For PTSD on the total EIS, the differences between the groups is large and statistically significant, but the standard deviations and confidence intervals for the difference (3.0-29.2) are also quite large. This means that up to 30% of patients could not receive the benefits of ‘daily wake up’ that are reported in this study. The SF-36, STAI and PAIS subscale scores mostly have more precision and some may be statistically significant. However the authors have been conservative in the analyses and rather than analysing the subscales of the SF-36 and PAIS independently have used a multivariate test which may have obscured any differences in subscale scores with large differences and less variability.
Will the Results Help Me In Caring For My Patients?
1. Can the results be applied to my patient care?
There were was a much higher proportion of females in this 6-12 month follow up study than seen in the usual RNSH ICU or tertiary ICUs in Australia, and patients were younger, and had longer LOS and ventilator time than average Australian and RNS ICU patients. However, readers of the article need to compare the data on clinical characteristics of patients in this and the original study with that from their own units for patients who were on mechanical ventilation and continuous sedation.
2. Were all clinically important outcomes considered?
Yes. This study reports on the clinically significant outcomes of how patients feel and function after surviving critical illness requiring treatment in ICU. Sleep disturbances are often reported by such patients after they go home, and the IES contained one question about falling and staying asleep.
3. Are the likely treatment benefits worth the potential harms and costs?
There appear to be no likely harms given the study outcomes; the authors were concerned that ‘daily wake up’ from continuous sedation could have harmful effects on patients’ psychological states during recovery, but this was not demonstrated. Rather some patients who underwent this intervention appear to benefit, and there may less cost if fewer drugs are administered and patients have shorter length of stay in ICU.

What other people had to say about:
The long-term psychological effects of daily sedative interruption on critically ill patients

Add Your Comment
[ Back ]

Any questions or comments please contact Gordon S. Doig
Implemented and designed by John Soer and Gordon Doig
Page last modified on Friday August 10, 2001