Supine Position Increases Pneumonia Risk in Ventilated Patients

Abstract & Commentary

Synopsis: Medical and respiratory ICU patients who were managed in the semirecumbent (45° head up) position had a lower incidence of nosocomial pneumonia than patients managed supine, especially if they received enteral feedings.

Source: Drakulovic MB, et al. Lancet 1999;354(9193): 1851-1858.

In this study from the university of barcelona, Drakulovic and colleagues randomized intubated, mechanically ventilated patients to be managed in either the supine (0° head elevation) or the semirecumbent position (45° head elevation) to determine the effect of position on development of ventilator-associated pneumonia. The study was done in one respiratory ICU and one medical ICU, and included all patients ventilated during the study period unless they had had recent abdominal surgery, recent neurosurgical intervention, refractory hypotension, or intubation within one month of entry. All management except for body position was at the discretion of the primary physician.

Eighty-six patients completed the trial at the time it was halted by interim analysis—39 who were nursed in the semirecumbent position and 47 who were nursed supine. The groups were well matched in terms of age (mid-60s), gender (75% male), APACHE II score (about 23), and the cause of acute respiratory failure; one-third of the patients in each group had acute exacerbations of chronic obstructive pulmonary disease. Mean duration of mechanical ventilation was similar in the two groups (171 vs 145 hours in semirecumbent vs supine positions, respectively), as was mean ICU stay (9.7 vs 9.3 days).

The clinical suspicion of pneumonia was determined in a standard fashion, and microbiological confirmation was sought by tracheal aspirate, bronchoalveolar lavage (BAL), or bronchoscopic protected specimen brush (PSB) at the discretion of the clinician. Colony-count thresholds for diagnosing pneumonia were those in widespread use: 105 colony-forming units (CFU)/mL for aspirates, 104 CFU/mL for BAL, and 103 CFU/mL for PSB.

The frequency of clinically suspected nosocomial pneumonia was lower in the semirecumbent group than in the supine group (3/39 vs 16/47 patients; 95% confidence interval [CI] for difference 10-42; P = 0.03). A similar difference was found with respect to microbiologically proven pneumonia (2/39 vs 11/47 patients; 95% CI 4-32; P = 0.018). Patients receiving enteral nutrition were more likely to develop pneumonia (odds ratio 5.7, 95% CI 1.5-33; P = 0.013), and 50% of all patients who were managed supine and also received enteral feeding developed pneumonia. There were no statistically significant differences in ICU mortality between the semirecumbent (7/39, 18%) and supine (13/47, 28%) patients; there was also no difference in overall mortality in patients with pneumonia (either suspected or confirmed) vs. those without pneumonia.


Previous studies have shown that mechanically ventilated patients who are managed in the semirecumbent position tend to reflux and aspirate gastric contents less frequently than those managed supine. These are major predispositions to development of ventilator-associated pneumonia, so it stands to reason that the head-up position would be associated with a lower incidence of this complication. However, this is the first randomized, controlled trial to demonstrate this lower incidence.

Other predispositions to ventilator-associated pneumonia include depressed level of consciousness (and the incidence was inversely proportional to Glasgow Coma Scale score in the present study), increased severity of illness, increased duration of mechanical ventilation, and enteral vs. parenteral nutritional support. The clinician has control of only some of these things, and a relatively simple intervention such as elevating the head of the bed is one of them. It is noteworthy that not all ventilated patients are candidates for management in the semirecumbent position, and that even among those patients who met the entry criteria for this study nearly one-third had to be excluded. Nevertheless, in patients with no contraindication to the head-up position, it seems a reasonable thing to do when possible. (Dr. Pierson is Professor of Medicine, University of Washington, Medical Director, Respiratory Care, Harborview Medical Center, Seattle.)