Duration of prone position sessions: a prospective cohort study

Abstract

Background: Prone position (PP) is highly recommended in moderate-to-severe ARDS. However, the optimal duration of PP sessions remains unclear. We searched to evaluate the time required to obtain the maximum physiological effect, and to search for parameters related to patient survival in PP.

Methods and results: It was a prospective, monocentric, physiological study. We included in the study all prone-positioned patients in our ICU between June 2016 and January 2018. Pulmonary mechanics, data from volumetric capnography and arterial blood gas were recorded before prone positioning, 2 h after proning, before return to a supine position (SP) and 2 h after return to SP. Dynamic parameters were recorded before proning and every 30 min during the session until 24 h. 103 patients (ARDS 95%) were included performing 231 PP sessions with a mean length of 21.5 ± 5 h per session. They presented a significant increase in pH, static compliance and P_aO₂/F_iO₂ with a significant decrease in P_aCO₂, P_plat, phase 3 slope of the volumetric capnography, P_etCO₂, V_D/V_T-phy and ΔP. The beneficial physiological effects continued after 16 h of PP and at least up to 24 h in some patients. The evolution of the respiratory parameters during the first session and also during the pooled sessions did not find any predictor of response to PP, whether before, during or 2 h after the return in SP.

Conclusions: PP sessions should be prolonged at least 24 h and be extended in the event that the P_aO₂/F_iO₂ ratio at 24 h remains below 150, especially since no criteria can predict which patient will benefit or not from it. Trial registration The trial has been registered on 28 June 2016 in ClinicalTrials.gov (NCT02816190) (https://clinicaltrials.gov/ct2/show/NCT02816190?term=propocap&rank=1).

Keywords: Acute respiratory distress syndrome; Acute respiratory failure; Capnography; Mechanical ventilation; Prone position; Protective ventilation.

Fig. 1

Evolution of ΔV_D/V_T-phy, ΔS_III, ΔP_etCO₂ and ΔC_dyn for each parameter (session 1), from 0 h (just before prone positioning) to 24 h of prone position; Ⓐ at sessions’ end (dotted line); Ⓑ 2 h after return in supine position (dotted line). *p < 0.05 (global time effect comparison)

Fig. 2

Evolution of ΔV_D/V_T-phy, ΔS_III, ΔP_etCO₂ and ΔC_dyn in responder sessions for each parameter (all sessions), from 0 h (just before prone positioning) to 24 h of prone position; Ⓐ at sessions’ end (dotted line); Ⓑ 2 h after return in supine position (dotted line). *p < 0.05 (global time effect comparison)

Fig. 3

Evolution over five consecutive prone position sessions of ΔC_dyn, ΔP_etCO₂, ΔS_III and ΔV_D/V_T-phy, in patient #20 (dead) and patient #35 (survivor), from 0 h (just before prone positioning) to the end of each prone position session