A 60 yr old woman was admitted to the ICU with a severe community acquired pneumonia and septic shock. She was invasively ventilated with a lung protective strategy, optimised PEEP and recruitment manouvres as needed. Her refractory hypoxia persisted and so she was probed for 16 hours a day for the first 5 days of her admission. She made slow but steady improvements and was discharged from the ICU 10 days later.

What is the current evidence for proning as a rescue therapy for refractory hypoxia?

Sam Clark

The last twenty years has seen a significant shift in the management of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS), with the development of lung-protective ventilation, permissive hypercapnia, recruitment strategies, and PEEP. These techniques, while life saving are not without morbidity: both supine mechanical ventilation and high concentrations of oxygen are associated with inflammation and fibrosis (1, 2). The prone position (or Proning) theoretically can improve oxygenation and reduce ventilation related lung injury(3). Ventilation-perfusion synchronizes better in sicker patients when prone, gravitational effects on pleural pressures are reduced and consistent transpulmonary pressures lead to balanced recruitment of collapsed regions (3).

Multiple observational studies have demonstrated that oxygenation has improved after proning patients(4), but some authors have questioned whether this had led to the predicted survival benefit (3). Also, it has been less clear from the observational literature whether patients are protected from ventilator induced lung injury and there are complications associated with proning including facial & airway oedema, pressure ulcers, inadvertent extubation and bronchial intubation, which though rare can lead to significant morbidity (4).

Up until recently there have been four larger scale RCTs assessing the effects of proning on patients with ARDS. In all three of these studies the treating physicians were, perhaps understandably, not blinded to the nature of any interventions. The first, published by Gattonini and colleagues in 2001, examined 304 patients from 28 ICUs from Italy and Switzerland with ALI or ARDS (5). The control group followed the American-European consensus guidelines for mechanical ventilation for the time and the treatment group was proned for at least 6 hours per day for period of ten days. The trigger was an PaO2: FiO2 ratio<200 plus PEEP of 5 or more or a PaO2: FiO2 ratio<300 plus PEEP of 10 or more. They found an overall mortality of 23% in ICU with 60% at 6months and a relative risk of 0.84 (CI 0.56-1.27). There was a significant improvement in arterial oxygenation. They concluded that the prone position did improve hypoxia but not mortality. There are a number of potential criticisms of this study, for example being insufficiently powered and including patients, who were on the mild end of the ALI/ARDS spectrum.

Guerin et al, who interestingly designed their study prior to Gattinoni’s results being available, recruited 791 patients from 21 ICUs in France (6). Randomizing them to either standard supine therapy or 8 hours proning. Patients were considered eligible if they had PaO2: FiO2 ratio<300 with no contraindications for proning. They found that there was no significant difference between the groups in their primary outcome: 28-day mortality for prone group was 32.4% and 31.5% in the supine group (CI 0.79-1.19, p=0.77). Days on ventilator were almost identical, and oxygenation was improved in the prone group. Numbers of VAPs were significantly lower in the prone group; however, there were significantly higher numbers of complications. Again, critics questioned the wide inclusion criteria, the potential delay in receiving the intervention and the length of time that the intervention was applied to the patients per day.

Mancebo and colleagues designed a trial that they hoped would answer these questions (2). They recruited 136 patients within 48 hours of developing severe ARDS (American-European Consensus definition of ARDS plus diffuse bilateral infiltrates on chest X-ray). The intervention group was proned for at least 20 hours per day. The study was powered to find a 20% percent improvement in mortality from a baseline of 50% in the supine group (and therefore should have enrolled 200 patients), unfortunately it was aborted early due to a rapid drop off in the numbers of patients enrolled in the last year. Most baseline descriptive statistics were similar except SAPSII scores, which were significantly worse in the prone group. ICU mortality was 58% in the supine group and 43% in the prone group (AR = 15% & RR=25%, p=0.12). This was non significant though advocates pointed to a distinct trend towards a clinical benefit especially given the underpowered nature of the study. A post hoc analysis demonstrated significant benefit of proning, if applied to patients with SAPSII<50, though this needs to be treated with caution.

Taccone et al continued to examine the potential benefits of the prone position for patients with severe hypoxaemia with moderate to severe ARDS (7). It was a multi-centered unblinded RCT with 25 European ICUs. The inclusion criteria specified that PaO2: FiO2 ratio needed to be below 200 indicating significant ARDS with a further severe subgroup with ratio below 100 and had to be within 72 hours of a diagnosis of ARDS. They found no significant difference in overall mortality, or in the separate subgroups. They did find a significant improvement in arterial oxygenation and consequent reduction in inspired oxygen, however, not in PEEP.

Overall, the literature had demonstrated that though oxygenation had improved in nearly every patient subgroup across all the trials there was no consistent benefit in mortality. There had been suggestion from post hoc analysis and meta-analysis of the available data that early instigation of proning for substantial proportion of the day may have been beneficial. There were, however, a number of issues with these studies: firstly there was little consistency in the protocols used between the studies in both for duration and timing of the intervention; secondly, a number of the studies were performed prior to the instigation of current lung protective strategies and patients were being ventilated at up to 10ml/kg in both the supine and prone arms of the earlier trials, which may have mitigated a proportion of the proposed benefit from proning patients. However, meta-analysis of the available data suggested that proning could be beneficial not only from an oxygenation perspective, but also in terms of survival (4).

Most recently, Guerin et al published a RCT that attempted to answer these previous questions (8). They compared 474 (of 3449) with severe ARDS, who had been intubated and ventilated for less than 36 hours. They were assign to either a supine group receiving best practice therapy for preventing ventilator related lung injury including low tidal volumes (6ml/kg), appropriate use of PEEP (mean was 10cmH2O), and plateau pressure less than 30cmH2O, or to a prone group that were proned for at least 16 consecutive hours per day for up to 28 days. The prone group was also given 12-24 hours to be stabilized prior to initiation of any proning within that 36-hour period. The results demonstrated a dramatic reduction in mortality between the prone and the supine groups both at 28 days (16% vs. 32%; hazard ratio 0.39, CI 0.25 to0.63) and 90 days (23.6% vs. 41%; hazard ratio 0.44, CI 0.29 to 0.67) respectively, with almost no significant difference in complications in between the two groups. There was an increase in the number of patients suffering cardiac arrest, but this was in the supine group. This was a well-conducted study that was appropriately powered and had good standardization across both groups for ventilation and weaning using a PEEP-FiO2 table. Interestingly, one reason that this study may have shown such an effect size is that all of the centers enrolled had at least five years experience of using the prone position, and that it may only be “safe” technique in experienced hands. However, while further studies may confirm or alter the conditions or patient groups that may most benefit from proning, Guerin et al have shown the prone position has a place in the management of severe ARDS.

---

References:
(1) Wagstaff A. Oxygen Therapy. IN Bersten A., Soni N., eds Oh’s Intensive Care Manual. Philadelphia: Butterworth-Heinerman/Elsevier: 2009. P315-326.
(2) Mancebo J., Fernandez R., Blanch L., et al. A Multicenter Trial of Prolonged Prone Ventilation in Severe Acute Respiratory Distress Syndrome. Am J Resp Crit Care Med. 2006; 173:1233-9.
(3)Fessler H., & Talmor D. Should Prone positioning be routinely used for lung protection during mechanical ventilation. Respir Care 2010;55(1):88 –96.
(4) Gattinoni L., Carlesso E., Taccone P., et al Prone positioning improves survival in severe ARDS: a pathological review and individual patient meta-analysis Minerva Anestesiologicica 2010. 76;6:p448-454.
(5) Gattinoni L., Tognoni G., Pesenti A., et al. Effect of prone position on the survival of patients with Acute Respiratory failure. NEJM 2001; 345:568-573.
(6) Guerin C., Gaillard S., Lemasson S., et al Effects of Systematic Prone Position in Hypoxemic Acute Respiratory Failure. JAMA 2004; 292(19);p2379-2387.
(7) Taccone p., Pesenti a., Latini R., et al Prone Positioning in patients with moderate and severe acute respiratory distress syndrome. JAMA 2009; 302(18):1977-1984.
(8) Guerin C., Reignier J., Richard J., et al Prone Positioning in Severe Acute Respiratory Distress Syndrome. Engl J Med 2013;368:2159-68.