Driving pressure

Reviewed and revised 11 April 2024; co-authored by Jack Iwashyna (@iwashyna) who provided the original analysis of the Amato et al (2015) study

OVERVIEW

Driving pressure (DP or ΔP) is defined as the distending pressure above the applied positive end-expiratory pressure (PEEP) required to generate tidal volume (V_T) (Roca et al, 2023)

• ΔP has been suggested by Amato et al (2015) to be the key variable for optimisation when performing mechanical ventilation in patients with acute respiratory distress syndrome (ARDS)
• ΔP is the ratio of tidal volume to (static) respiratory system compliance ; i.e. ΔP = V_T/C_RS
• ΔP can be calculated at the bedside as plateau pressure minus positive end-expiratory pressure, ie. ΔP = P_plat – PEEP

RATIONALE

Clinical need

• Protective lung ventilation strategies and ‘open lung’ approaches are associated with less ventilator-induced lung injury (VILI), improved oxygenation and improved outcomes
• Important components of these strategies all decrease stress on the lung:
  • lower tidal volumes
  • lower plateau pressure
  • higher PEEP
• However, clinical trials have found conflicting responses to the manipulation of these separate components of lung protection and when optimization of one component negatively affects another it is often unclear at the bedside which is preferred
• ΔP is easily measured/ calculated.

Physiology

• Normalised target tidal volumes to predicted body weight (PBW), as per the ARDSNet ventilation strategy, does not take into account the varying proportion of lung that is not available for ventilation in ARDS (‘baby lung’ concept)
• The decrease in available lung for ventilation manifests as a decrease in C_RS
• If two lungs are the same size, but the first lung has lower C_RS, a delivered V_T calculated according to PBW will cause more mechanical stress in the first lung than the more compliant second lung
• Thus normalizing V_T to C_RS and using the ratio as an index to indicate the “functional” size of the lung may provide a better predictor of outcomes in patients with ARDS than V_T alone
• This ratio is termed the driving pressure (ΔP = V_T/C_RS) and can be routinely calculated for patients who are not making inspiratory efforts as ΔP = P_plat – PEEP

EVIDENCE IN ARDS

Amato et al, NEJM 2015

• Retrospective analysis of previously prospectively collected patient-level RCT data
• n = 3,562 patients from 9 major previously reported ARDS trials
• Methods
  • Driving pressure was calculated as ΔP =P_plat – PEEP (if no inspiratory effort) and was averaged over the first 24h post-randomization
  • A customized risk-adjustor was used to control for differences in risk-of-death to see which variable best explained the mortality pattern
  • the first analysis used a standard risk analysis with multivariate adjustments and multilevel mediation analysis to ask: “can changes in driving pressure explain who received the mortality benefits of alternative ventilation strategies?”
  • In a second analysis, patients were divided into quintiles differing in combination of PEEP, P_plat and ΔP, testing to see which was more consistently associated with trends in mortality
• Outcomes
  • In the first analysis, driving pressure (ΔP) was a better predictor of mortality than Compliance or V_T: “ΔP mediated 75% of the benefits due to treatment-group assignment in the V_T trials (P=0.004 for the average causal mediation effect) and 45% of these benefits in the PEEP trials (P = 0.001).”
  • In the second analysis, across quintiles:

Pplat	PEEP	ΔP	Mortality
rising	same	rising	rising
rising	rising	same	same
same	rising	falling	falling

• Commentary and criticisms:
  • Although an elegant epidemiologic examination of physiology, a secondary re-analysis of RCTs does not equal an RCT!
  • Calculated ΔP is a post-randomization variable, so if making inspiratory efforts or P_plat contains information about response to therapy, then ΔP is confounded
  • If the findings are valid, does this mean we should set P_plat as high as we want as long my ΔP is low? Conversely, does this mean a low P_plat but with quite low PEEP harms patients?
  • The study analysed a physiological target, not an intervention – introducing an intervention to achieve the target may not improve mortality if the target is not part of the causal pathway for mortality or the intervention causes additional harm (Fan and Rubenfeld, 2017).

Aoyamo et al, 2018

• Systematic review and meta-analysis to determine the risk of mortality for higher vs. lower driving pressure in mechanically ventilated patients with ARDS.
• Included 7 studies (6,062 patients) with low risk of bias.
• Findings:
  • Higher driving pressure associated with significantly higher mortality (pooled risk ratio 1.44; 95% CI 1.11-1.88).
  • Similar results for studies with driving pressure cutoffs of 13-15 cm H2O (pooled risk ratio 1.28; 95% CI 1.14-1.43).

De Jong et al, 2018

• Retrospective single-center study of prospectively collected data of all ARDS patients admitted consecutively to a mixed medical-surgical adult ICU.
• Compared obese (body mass index ≥ 30 kg/m²) and non-obese patients.
• Findings:
  • No difference in mortality rate at day 90: 47% (95% CI, 40-53) in non-obese, 46% (95% CI, 36-56) in obese patients.
  • Driving pressure at day 1 significantly lower in non-obese survivors (11.9 ± 4.2 cmH₂O) than non-survivors (15.2 ± 5.2 cmH₂O, p < 0.001) – associated with P_plat, C_rs, and ΔP.
  • In obese patients, driving pressure at day 1 not significantly different between survivors (13.7 ± 4.5 cmH₂O) and non-survivors (13.2 ± 5.1 cmH₂O, p = 0.41) at day 90.
• Conclusion: In contrast to non-obese ARDS patients, driving pressure was not associated with mortality in obese ARDS patients.

Terry et al, 2018

• Retrospective observational cohort study (2016-2018) of mechanically ventilated patients (n=3,204) in medical and surgical ICUs comparing different BMI classes (Normal/overweight, obese, severely obese)
• Findings:
  • No differences in in-hospital mortality, ventilator-free days, or ICU length of stay among all three BMI groups.
  • Severe obesity associated with higher DP (RR 1.51) and Ers (RR 1.31).
• Conclusion: Driving pressure lacked prognostic value in mechanically ventilated obese patients.

Romano et al, 2020

• Randomized, controlled, nonblinded pilot trial with 31 ARDS patients on invasive mechanical ventilation and driving pressure ≥13 cm H₂O.
• Interventions:
  • Driving pressure-limited group: Ventilated with volume-controlled or pressure-support modes, tidal volume titrated to 4-8 ml/kg of predicted body weight (PBW), aiming for a driving pressure of 10 cm H₂O or the lowest possible (daily measurement).
  • Control group: Ventilated according to ARDSNet protocol, using a tidal volume of 6 ml/kg PBW (allowed to be set down to 4 ml/kg PBW if plateau pressure >30 cm H₂O).
• Results:
  • Baseline driving pressure: 15.0 cm H₂O in both groups.
  • Driving pressure (first hour to third day) significantly lower in driving pressure-limited group (mean difference 4.6 cm H₂O; 95% CI 6.5 to 2.8; P < 0.001).
  • Tidal volume in driving pressure-limited group kept lower than control group (mean difference 1.3 ml/kg PBW; 95% CI 1.7 to 0.9; P < 0.001).
  • No statistically significant differences in severe acidosis incidence (pH < 7.10) within 7 days.
• Conclusion: a clinical trial of a driving pressure-limited ventilation strategy in ARDS patients is feasible.

DROP study, 2020

• Prospective longitudinal study (n=122 patients with ARDS) comparing driving pressure and absolute PaO₂/FiO₂ratio in determining the best PEEP level
  • PEEP increased until plateau pressure reached 30 cmH₂O, then decreased at 15-min intervals to 15, 10, and 5 cmH₂O.
  • Best PEEP by PaO₂/FiO₂ ratio (PEEPO₂): Highest PaO2/FiO2 ratio obtained.
  • Best PEEP by driving pressure (PEEPDP): Lowest driving pressure.
• Findings:
  • Mean PEEPO₂ value: 11.9 ± 4.7 cmH2O.
  • Mean PEEPDP value: 10.6 ± 4.1 cmH2O.
  • Difference: 1.3 cmH2O (95% CI 0.4–2.3; one-tailed P value 0.36).
  • Only 46 PEEP levels were the same with both methods (37.7%; 95% CI 29.6–46.5).
  • PEEP level ≥ 15 cmH₂O in 61 (50%) patients with PEEPO₂ and 39 (32%) patients with PEEPDP (P = 0.001).
• Conclusion: Best PEEP level varies depending on the method chosen, with PEEPDP level lower than PEEPO₂ level. No evidence provided of impact on patient outcomes.

Costa et al, 2021

• Data from pooled database of ARDS patients (n=4,549 patients) in 6 randomized clinical trials and 1 observational cohort study was used to assess the impact of mechanical power on mortality in ARDS patients compared to primary ventilator variables (ΔP, V_T, RR).
• Findings:
  • ΔP, RR, and mechanical power were all significant predictors of mortality.
  • ΔP impact on mortality was four times larger than RR.
  • The formula (4 × ΔP) + RR was at least as informative as mechanical power
• Conclusion: supports hypothesis that both low ΔP and low RR are important variables for optimisation of mechanical ventilation in ARDS patients.

Goligher et al, 2021

• Secondary analysis of five randomized trials using Bayesian multivariable logistic regression to assess interaction between V_T strategy and respiratory system elastance (E_rs) on 60-day mortality.
• Findings: High posterior probability (93%) that mortality benefit varies with E_rs; greater absolute risk reduction in mortality with higher E_rs.
• Conclusion: Findings support prioritizing ΔP over V_T in lung-protective ventilation strategies for ARDS.

EVIDENCE IN PATIENTS WITHOUT ARDS

Studies of ΔP in mechanically ventilated patients with obesity have mixed findings

• Sahayeta et al (2019) used data collected during a prospective, observational cohort multicenter study to determine whether ΔP and P_plat at enrollment were associated with hospital mortality among 1132 mechanically ventilated participants. Among those without ARDS, higher ΔP (adjusted OR = 1.36 per 7 cm H₂O, 95% CI 1.14-1.62) and P_plat (adjusted OR = 1.42 per 8 cm H₂O, 95% CI 1.17-1.73) were associated with higher mortality.
• Restrospective analyses by Schmidt et al (2018) and Lanspa et al (2019) both showed no association between ΔP and mortality (n=622 and n= 2641, repectively)

EVIDENCE IN PATIENTS WITH OBESITY

Studies of ΔP in mechanically ventilated patients with obesity have failed to demonstrate prognostic utility.

De Jong et al, 2018

• Retrospective single-center study of prospectively collected data of all ARDS patients admitted consecutively to a mixed medical-surgical adult ICU.
• Compared obese (body mass index ≥ 30 kg/m²) and non-obese patients.
• Findings:
  • No difference in mortality rate at day 90: 47% (95% CI, 40-53) in non-obese, 46% (95% CI, 36-56) in obese patients.
  • Driving pressure at day 1 significantly lower in non-obese survivors (11.9 ± 4.2 cmH₂O) than non-survivors (15.2 ± 5.2 cmH₂O, p < 0.001) – associated with P_plat, C_rs, and ΔP.
  • In obese patients, driving pressure at day 1 not significantly different between survivors (13.7 ± 4.5 cmH₂O) and non-survivors (13.2 ± 5.1 cmH₂O, p = 0.41) at day 90.
• Conclusion: In contrast to non-obese ARDS patients, driving pressure was not associated with mortality in obese ARDS patients.

Terry et al, 2018

• Retrospective observational cohort study (2016-2018) of mechanically ventilated patients (n=3,204) in medical and surgical ICUs comparing different BMI classes (Normal/overweight, obese, severely obese)
• Findings:
  • No differences in in-hospital mortality, ventilator-free days, or ICU length of stay among all three BMI groups.
  • Severe obesity associated with higher DP (RR 1.51) and E_rs (RR 1.31).
• Conclusion: Driving pressure lacked prognostic value in mechanically ventilated obese patients.

PROBLEMS

• Lack of high quality clinical trials demonstrating clinical utility of ΔP-guided therapy
• It is unclear if ‘improving’ driving pressure using an open lung approach (high PEEP and/or lung recruitment manoeuvres) would improve patient outcomes
• Transpulmonary pressure (TPP or ΔP_L), measured using an esophageal manometer  (TPP = Palv – Pes), is likely a better surrogate of lung stress as it excludes contributions from the chest wall (Fan and Rubenfeld, 2017)
• ΔP, V_T, P_plat, and PEEP are physiologically coupled – a change in one may affect the others unpredictably and in a way that may improve mortality (e.g. improved ΔP and V_T with increased PEEP in “recruiters”, but not in “non-recruiters”) (Fan and Rubenfeld, 2017)
• Effects of PEEP adjustment on ΔP may not be apparent over short time frames (e.g. 5-10 min) (Fan and Rubenfeld, 2017)
• Spontaneous respiratory efforts alter ΔP and interfere with accurate titration – true ΔP is likely underestimated in spontaneously breathing patients (Hirshberg and Majercik, 2020)
• daily administration of sedatives and neuromuscular blockers to facilitate ΔP may have clinically significant harms (Hirshberg and Majercik, 2020)
• Maybe less useful in patients with high chest wall elastance (e.g. obesity) as much of the pressure that is applied by the ventilator is needed to distend the chest wall rather than the lung – both P_plat and pleural pressure (P_pl) increase, with no increase in TPP and lung strain (De Jong et al, 2018).

CONCLUSION

• Titrating V_T to prevent ΔP >13 cmH₂O, if has minimal costs in terms of CO₂ clearance, appears to be a reasonable adjunct to a protective lung ventilation approach
• However, the use of driving pressure is yet to be subjected to a high quality randomised controlled trial confirming its clinical utility and safety
• Optimal threshold for ΔP, if any, is unknown (various studies suggests targets in the range of 10-15 cmH₂O).

References and Links

LITFL

• CCC — Protective lung ventilation
• CCC — Open lung approach to ventilation
• CCC — Positive end-expiratory pressure (PEEP)
• CCC — High airway and alveolar pressures
• CCC — Transpulmonary pressure and oesophageal manometry

Journal articles

• Amato MB, Meade MO, Slutsky AS. Driving pressure and survival in the acute respiratory distress syndrome. The New England journal of medicine. 372(8):747-55. 2015. [pubmed] [free full text] [letters]
• Aoyama H, Pettenuzzo T, Aoyama K, Pinto R, Englesakis M, Fan E. Association of Driving Pressure With Mortality Among Ventilated Patients With Acute Respiratory Distress Syndrome: A Systematic Review and Meta-Analysis. Crit Care Med. 2018 Feb;46(2):300-306. doi: 10.1097/CCM.0000000000002838. PMID: 29135500.
• Aoyama H, Yamada Y, Fan E. The future of driving pressure: a primary goal for mechanical ventilation? J Intensive Care. 2018 Oct 4;6:64. doi: 10.1186/s40560-018-0334-4. PMID: 30305906; PMCID: PMC6172758.
• Bellani G, Grassi A, Sosio S, Foti G. Plateau and driving pressure in the presence of spontaneous breathing. Intensive Care Med. 2019 Jan;45(1):97-98. doi: 10.1007/s00134-018-5311-9. Epub 2018 Jul 13. PMID: 30006893.
• Borges JB, Hedenstierna G, Larsson A, Suarez-Sipmann F. Altering the mechanical scenario to decrease the driving pressure. Critical care (London, England). 19:342. 2015. [pubmed] [free full text]
• Costa ELV, Slutsky AS, Brochard LJ, Brower R, Serpa-Neto A, Cavalcanti AB, Mercat A, Meade M, Morais CCA, Goligher E, Carvalho CRR, Amato MBP. Ventilatory Variables and Mechanical Power in Patients with Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med. 2021 Aug 1;204(3):303-311. doi: 10.1164/rccm.202009-3467OC. PMID: 33784486.
• De Jong A, Cossic J, Verzilli D, Monet C, Carr J, Conseil M, Monnin M, Cisse M, Belafia F, Molinari N, Chanques G, Jaber S. Impact of the driving pressure on mortality in obese and non-obese ARDS patients: a retrospective study of 362 cases. Intensive Care Med. 2018 Jul;44(7):1106-1114. doi: 10.1007/s00134-018-5241-6. Epub 2018 Jun 15. PMID: 29947888.
• Fan E, Rubenfeld GD. Driving Pressure-The Emperor’s New Clothes. Crit Care Med. 2017 May;45(5):919-920. doi: 10.1097/CCM.0000000000002386. PMID: 28410313.
• Grieco DL, Chen L, Dres M, Brochard L. Should we use driving pressure to set tidal volume? Curr Opin Crit Care. 2017 Feb;23(1):38-44. doi: 10.1097/MCC.0000000000000377. PMID: 27875410.
• Goligher EC, Costa ELV, Yarnell CJ, Brochard LJ, Stewart TE, Tomlinson G, Brower RG, Slutsky AS, Amato MPB. Effect of Lowering Vt on Mortality in Acute Respiratory Distress Syndrome Varies with Respiratory System Elastance. Am J Respir Crit Care Med. 2021 Jun 1;203(11):1378-1385. doi: 10.1164/rccm.202009-3536OC. PMID: 33439781.
• Hirshberg EL, Majercik S. Targeting Driving Pressure for the Management of ARDS…Isn’t It Just Very Low Tidal Volume Ventilation? Ann Am Thorac Soc. 2020 May;17(5):557-558. doi: 10.1513/AnnalsATS.202002-108ED. PMID: 32356695; PMCID: PMC7193815.
• Lanspa MJ, Peltan ID, Jacobs JR, Sorensen JS, Carpenter L, Ferraro JP, Brown SM, Berry JG, Srivastava R, Grissom CK. Driving pressure is not associated with mortality in mechanically ventilated patients without ARDS. Crit Care. 2019 Dec 27;23(1):424. doi: 10.1186/s13054-019-2698-9. PMID: 31881909; PMCID: PMC6935179.
• Loring SH, Malhotra A. Driving pressure and respiratory mechanics in ARDS. The New England journal of medicine. 372(8):776-7. 2015. [pubmed] [free full text]
• Rezaiguia-Delclaux S, Ren L, Gruner A, Roman C, Genty T, Stéphan F. Oxygenation versus driving pressure for determining the best positive end-expiratory pressure in acute respiratory distress syndrome. Crit Care. 2022 Jul 13;26(1):214. doi: 10.1186/s13054-022-04084-z. PMID: 35831827; PMCID: PMC9281138.
• Roca O, Goligher EC, Amato MBP. Driving pressure: applying the concept at the bedside. Intensive Care Med. 2023 Aug;49(8):991-995. doi: 10.1007/s00134-023-07071-2. Epub 2023 May 16. PMID: 37191695.
• Schmidt MFS, Amaral ACKB, Fan E, Rubenfeld GD. Driving Pressure and Hospital Mortality in Patients Without ARDS: A Cohort Study. Chest. 2018 Jan;153(1):46-54. doi: 10.1016/j.chest.2017.10.004. Epub 2017 Oct 14. PMID: 29037528.
• Terry C, Brinton D, Simpson AN, Kirchoff K, Files DC, Carter G, Ford DW, Goodwin AJ. Elevated Driving Pressure and Elastance Does Not Increase In-Hospital Mortality Among Obese and Severely Obese Patients With Ventilator Dependent Respiratory Failure. Crit Care Explor. 2022 Dec 12;4(12):e0811. doi: 10.1097/CCE.0000000000000811. PMID: 36583205; PMCID: PMC9750660.

FOAM and web resources

• Maryland CC Project — Brower: Driving pressure in ARDS, does it matter?!? (2016)