Complications of Prone Ventilation in ARDS

Abstract & Commentary

Synopsis: Of nine patients with severe ARDS and refractory hypoxemia following multisystem trauma who were managed in the prone position, four experienced serious complications related to the positioning. Although this experience may represent a "worst case" scenario, it illustrates the potential hazards of this adjunctive approach to ventilatory support in ARDS.

Source: Offner PJ, et al. J Trauma 2000;48:224-228.

Offner and colleagues at denver health medical Center report their experience with nine patients managed in the prone position during mechanical ventilation for acute respiratory distress syndrome (ARDS) following severe, multisystem trauma during a 12-month period. The patients were identified retrospectively from among a prospective cohort of patients with severe multisystem trauma admitted to the intensive care unit (ICU) of a level 1 trauma center. Standard assessment methods were used to evaluate severity of illness and to diagnose ARDS. The nine patients reported included two women; their mean age was 29 years, and all had ARDS following blunt thoracic trauma. Mean duration of mechanical ventilation prior to prone positioning was 11 ± 1.7 days. All nine of the patients had been placed in the prone position because of refractory hypoxemia that could not be managed satisfactorily with positive end-expiratory pressure (PEEP) adjusted according to a standardized approach, along with a lung-protective ventilatory strategy aimed at maintaining safe inspiratory plateau pressures. Patients were placed in the prone position by four to six caregivers, including surgeons, nurses, and respiratory therapists, with special care given to all tubes and lines.

Four patients experienced serious adverse effects related to prone positioning. Four of the nine patients had midline incisions from exploratory laparotomy, and wound dehiscence occurred in one patient. Severe facial and upper chest wall pressure necrosis developed in two patients. The fourth patient experienced a cardiac arrest upon first being moved into the prone position.

Oxygenation, as measured by the ratio of arterial PO2 to inspired oxygen fraction (PaO2/FIO2) initially improved in six of the eight patients in whom it was assessed, from 75 ± 7 to 147 ± 27 mmHg (P = 0.04). Offner et al do not report the duration of prone positioning, nor the ultimate outcomes (i.e., ICU or hospital mortality) in their patients. They conclude that although prone positioning may improve arterial oxygenation in patients late in the course of severe ARDS following multisystem trauma, it has the potential for serious complications.


As demonstrated by numerous studies, prone positioning often improves arterial PO2 in patients with ARDS. However, to date there have been no published clinical trials showing a positive effect on survival. Data collection in a large-scale randomized, controlled trial in more than 50 Italian ICUs has been completed, although the results have not yet been released. Thus, prone positioning is currently in the same category as several other measures used in treating patients with severe ARDS that may improve oxygenation in the short term but that have not been shown to affect the ultimate outcome. Included in this category are inhaled nitric oxide, high-frequency jet ventilation, pressure control inverse-ratio ventilation, and various approaches to determining optimum PEEP.

Only one multicenter randomized, controlled trial in patients with ARDS has shown different outcomes with different approaches to mechanical ventilation. The results of that NIH-sponsored study have recently been presented internationally and are to be published in the May 4th edition of the New England Journal of Medicine. The article has been released to the press prior to publication, and is available on the Internet at This study showed that a lung-protective ventilatory strategy using assist-control ventilation and small tidal volumes with limited static inspiratory pressures and permissive hypercapnia improved survival by 25% as compared to the same mode with larger tidal volumes and higher inflation pressures. A previous, smaller study, from a single institution, showed that a lung-protective ventilatory strategy using pressure control ventilation and based on pressure-volume curves obtained during therapeutic paralysis significantly reduced mortality in patients with ARDS (Amato MB, et al. N Engl J Med 1998;338:347-354).

A substantial amount of potentially important information was not included in the paper by Offner et al. Although we are told that two patients were excluded from prone positioning during the same period because of open abdominal wounds following damage-control laparotomy, the proportion of patients turned prone to the total number of patients with severe ARDS during the study period is not provided. It would also have been instructive to know how long the nine reported patients were kept prone, and how many of them survived.

Although it is not stated explicitly, the nine patients included in this series likely represent early experience with prone positioning at Offner et al’s institution. As they point out, these patients were late in the course of severe ARDS, and likely were more predisposed to skin breakdown and wound dehiscence because of poor tissue turgor and extensive third-space edema. These issues notwithstanding, this paper makes an important point. Despite deliberate, extensive efforts to prevent dislodgement of tubes and other adverse effects, prone positioning led to serious complications in several of the patients in which it was used. Particularly in the absence of evidence that this maneuver has a favorable effect on the ultimate outcome in patients with ARDS, clinicians should use caution and make sure that other aspects of management that might also improve oxygenation are used optimally.

Prone positioning in patients with severe ARDS:

a. decreases the incidence of ventilator-associated pneumonia.

b. improves ultimate survival by 25%.

c. frequently improves arterial oxygenation.

d. All of the above

e. None of the above