Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial
This review may be edited
Posted By: Gordon S. Doig
Posted Date: 24 Jan 2003
Title: Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial
Authors: Mitra B Drakulovic MD, Antoni Torres MD, Torsten T Bauer MD, Jose M Nicolas MD, Santiago Nogué MD
Reference: The Lancet. Volume 354, Issue 9193, 27 November 1999, Pages 1851-1858
Link: Click here for a direct link to the paper. A password may be required for access to fulltext.
Abstract: Background Risk factors for nosocomial pneumonia, such as gastro-oesophageal reflux and subsequent aspiration, can be reduced by semirecumbent body position in intensive-care patients. The objective of this study was to assess whether the incidence of nosocomial pneumonia can also be reduced by this measure.
Methods This trial was stopped after the planned interim analysis. 86 intubated and mechanically ventilated patients of one medical and one respiratory intensive-care unit at a tertiary-care university hospital were randomly assigned to semirecumbent (n=39) or supine (n=47) body position. The frequency of clinically suspected and microbiologically confirmed nosocomial pneumonia (clinical plus quantitative bacteriological criteria) was assessed in both groups. Body position was analysed together with known risk factors for nosocomial pneumonia.
Findings The frequency of clinically suspected nosocomial pneumonia was lower in the semirecumbent group than in the supine group (three of 39 [8%] vs 16 of 47 [34%]; 95% CI for difference 10·0-42·0, p=0·003). This was also true for microbiologically confirmed pneumonia (semirecumbent 2/39 [5%] vs supine 11/47 [23%]; 4·2-31·8, p=0·018). Supine body position (odds ratio 6·8 [1·7-26·7], p=0·006) and enteral nutrition (5·7 [1·5-22·8], p=0·013) were independent risk factors for nosocomial pneumonia and the frequency was highest for patients receiving enteral nutrition in the supine body position (14/28, 50%). Mechanical ventilation for 7 days or more (10·9 [3·0-40·4], p=0·001) and a Glasgow coma scale score of less than 9 were additional risk factors.
Interpretation The semirecumbent body position reduces frequency and risk of nosocomial pneumonia, especially in patients who receive enteral nutrition. The risk of nosocomial pneumonia is increased by long-duration mechanical ventilation and decreased consciousness.
Are the Results Valid?
1. Was the assignment of patients to treatments randomized? ( Was allocation concealment maintained?)
Yes. The authors describe the use of 'a computer generated list' that was 'generated and disclosed by an independent person.' If done appropriately, this process would result in the maintenance of concealment, althouth exact details of the 'disclosure process' were not described.
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
2a. Was followup complete?
No. Although the investigators report that all patients were followed for at least 72 hours 'after protocol termination' (Figure 1), ICU mortality is only reported in 87 of 90 randomized patients. (One early patient death plus the results reported for the 86 per protocol patients).
2b. Were patients analyzed in the groups to which they were randomized?
No. Although it is reported that one patient randomized to the semirecumbent group 'died during resuscitation 2h after initiation of the protocol', this death was not considered in the comparison of ICU mortality. Likewise, three additional patients, who were inappropriately recruited to the trial and randomized to/ received semirecumbent therapy were withdrawn from the per protocol or efficacy analysis. Their outcomes (pneumonia and ICU mortality) should have been available and analyzed in a true intention to treat analysis.
3. Were patients, health workers, and study personnel blind to treatment?
No. Due to the nature of the intervention, blinding would be impossible however it would have been possible to ajudicate outcomes (suspected pneumonia, pneumonia) in a blinded fashion. This was not done.
4. Were the groups similar at the start of the trial?
The groups look reasonably similar.
5. Aside from the experimental intervention, were the groups treated equally?
It appears that the groups were treated in a reasonably similar fashion.
What are the Results?
1. How large was the treatment effect?
The trial was stopped as a result of an interim analysis.
Per protocol analysis as reported in the paper:
  • clinically suspected ventilator associated pneumonia was reduced from 34% (16/47) in the supine patients to 8% (3/39) in the semirecumbent patients (p=0.003).[N.B.-the p-value reported is obtained from a Pearson chi-square. In this situation, the Fisher's Exact Test would have been more appropriate, which yields a p-value of 0.004].
  • microbiologically confirmed pneumonia was also significantly reduced from 23% (11/47) to 5% (2/39) in the semirecumbent group (p=0.018).[N.B.-the p-value reported is obtained from a Person chi-square. In this situation, Fisher's Exact Test would have been more appropriate, which yields a p-value of 0.031].
    Intention to treat analysis calculated under conservative assumptions (includes all patients who were randomized and assumes that all 3 protocol violation patients developed microbially confirmed pneumonia.)
  • pneumonia rate in the supine group remains at 34% (16/47) while pneumonia rate was still signficantly lower at 14% (6/43) in the semirecumbent group (p=0.03, Fisher's Exact Test)
  • microbiologically confirmed pneumonia in the supine group remains at 23% (11/47) but at 11.6% (5/38) was not significantly lower in the semirecumbent group (p=0.175, Fisher's Exact Test).
    ICU mortality
  • conservative evaluation of ICU mortality shows no significant difference between the two groups (13/47 supine vs. 10/43 semirecumbent, p=0.809).
  • 2. How precise was the estimate of the treatment effect?
    See abstract.
    Will the Results Help Me In Caring For My Patients?
    1. Can the results be applied to my patient care?
    Yes, where appropriate beds are available, the intervention can be implemented in most ICUs.
    2. Were all clinically important outcomes considered?
    ICU readmission rates, hospital LoS and hospital mortality could have been reported.
    3. Are the likely treatment benefits worth the potential harms and costs?
    Although is is widely accepted that small, underpowered trials tend to over-estimate potential treatment benefits, an extremely conservative interpretation of this paper suggests that there is benefit associated with nursing in the semirecumbent position. Known harms and costs are minimal.
    An evidence-based recommendation, which is the building block for an evidence-based guideline, that summarizes this topic can be found in the EBR section.

    What other people had to say about:
    Supine body position as a risk factor for nosocomial pneumonia in mechanically ventilated patients: a randomised trial

    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