Zhongheng Zhang1, Bin Zheng2, Nan Liu3,4, Huiqing Ge5, Yucai Hong6. 1. Department of Emergency Medicine, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, No 3, East Qingchun Road, Hangzhou, 310016, Zhejiang, China. zh_zhang1984@zju.edu.cn. 2. Department of Surgery, 2D, Walter C Mackenzie Health Sciences Centre, University of Alberta, Edmonton, AB, Canada. 3. Duke-NUS Medical School, National University of Singapore, Singapore, Singapore. 4. Health Services Research Centre, Singapore Health Services, Singapore, Singapore. 5. Department of Respiratory Care, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, China. 6. Department of Emergency Medicine, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, No 3, East Qingchun Road, Hangzhou, 310016, Zhejiang, China.
Abstract
PURPOSE: Protective mechanical ventilation based on multiple ventilator parameters such as tidal volume, plateau pressure, and driving pressure has been widely used in acute respiratory distress syndrome (ARDS). More recently, mechanical power (MP) was found to be associated with mortality. The study aimed to investigate whether MP normalized to predicted body weight (norMP) was superior to other ventilator variables and to prove that the discrimination power cannot be further improved with a sophisticated machine learning method. METHODS: The study included individual patient data from eight randomized controlled trials conducted by the ARDSNet. The data was split 3:1 into training and testing subsamples. The discrimination of each ventilator variable was calculated in the testing subsample using the area under receiver operating characteristic curve. The gradient boosting machine was used to examine whether the discrimination could be further improved. RESULTS: A total of 5159 patients with acute onset ARDS were included for analysis. The discrimination of norMP in predicting mortality was significantly better than the absolute MP (p = 0.011 for DeLong's test). The gradient boosting machine was not able to improve the discrimination as compared to norMP (p = 0.913 for DeLong's test). The multivariable regression model showed a significant interaction between norMP and ARDS severity (p < 0.05). While the norMP was not significantly associated with mortality outcome (OR 0.99; 95% CI 0.91-1.07; p = 0.862) in patients with mild ARDS, it was associated with increased risk of mortality in moderate (OR 1.11; 95% CI 1.02-1.23; p = 0.021) and severe (OR 1.13; 95% CI 1.03-1.24; p < 0.008) ARDS. CONCLUSIONS: The study showed that norMP was a good ventilator variable associated with mortality, and its predictive discrimination cannot be further improved with a sophisticated machine learning method. Further experimental trials are needed to investigate whether adjusting ventilator variables according to norMP will significantly improve clinical outcomes.
PURPOSE: Protective mechanical ventilation based on multiple ventilator parameters such as tidal volume, plateau pressure, and driving pressure has been widely used in acute respiratory distress syndrome (ARDS). More recently, mechanical power (MP) was found to be associated with mortality. The study aimed to investigate whether MP normalized to predicted body weight (norMP) was superior to other ventilator variables and to prove that the discrimination power cannot be further improved with a sophisticated machine learning method. METHODS: The study included individual patient data from eight randomized controlled trials conducted by the ARDSNet. The data was split 3:1 into training and testing subsamples. The discrimination of each ventilator variable was calculated in the testing subsample using the area under receiver operating characteristic curve. The gradient boosting machine was used to examine whether the discrimination could be further improved. RESULTS: A total of 5159 patients with acute onset ARDS were included for analysis. The discrimination of norMP in predicting mortality was significantly better than the absolute MP (p = 0.011 for DeLong's test). The gradient boosting machine was not able to improve the discrimination as compared to norMP (p = 0.913 for DeLong's test). The multivariable regression model showed a significant interaction between norMP and ARDS severity (p < 0.05). While the norMP was not significantly associated with mortality outcome (OR 0.99; 95% CI 0.91-1.07; p = 0.862) in patients with mild ARDS, it was associated with increased risk of mortality in moderate (OR 1.11; 95% CI 1.02-1.23; p = 0.021) and severe (OR 1.13; 95% CI 1.03-1.24; p < 0.008) ARDS. CONCLUSIONS: The study showed that norMP was a good ventilator variable associated with mortality, and its predictive discrimination cannot be further improved with a sophisticated machine learning method. Further experimental trials are needed to investigate whether adjusting ventilator variables according to norMP will significantly improve clinical outcomes.
Authors: Todd W Rice; Arthur P Wheeler; B Taylor Thompson; Jay Steingrub; R Duncan Hite; Marc Moss; Alan Morris; Ning Dong; Peter Rock Journal: JAMA Date: 2012-02-05 Impact factor: 56.272
Authors: Tài Pham; Ary Serpa Neto; Paolo Pelosi; John Gerard Laffey; Candelaria De Haro; Jose Angel Lorente; Giacomo Bellani; Eddy Fan; Laurent Jean Brochard; Antonio Pesenti; Marcus Josephus Schultz; Antonio Artigas Journal: Anesthesiology Date: 2019-02 Impact factor: 7.892
Authors: J R Hotchkiss; L Blanch; G Murias; A B Adams; D A Olson; O D Wangensteen; P H Leo; J J Marini Journal: Am J Respir Crit Care Med Date: 2000-02 Impact factor: 21.405
Authors: Marcelo B P Amato; Maureen O Meade; Arthur S Slutsky; Laurent Brochard; Eduardo L V Costa; David A Schoenfeld; Thomas E Stewart; Matthias Briel; Daniel Talmor; Alain Mercat; Jean-Christophe M Richard; Carlos R R Carvalho; Roy G Brower Journal: N Engl J Med Date: 2015-02-19 Impact factor: 91.245
Authors: Matthias Briel; Maureen Meade; Alain Mercat; Roy G Brower; Daniel Talmor; Stephen D Walter; Arthur S Slutsky; Eleanor Pullenayegum; Qi Zhou; Deborah Cook; Laurent Brochard; Jean-Christophe M Richard; Francois Lamontagne; Neera Bhatnagar; Thomas E Stewart; Gordon Guyatt Journal: JAMA Date: 2010-03-03 Impact factor: 56.272
Authors: John G Laffey; Giacomo Bellani; Tài Pham; Eddy Fan; Fabiana Madotto; Ednan K Bajwa; Laurent Brochard; Kevin Clarkson; Andres Esteban; Luciano Gattinoni; Frank van Haren; Leo M Heunks; Kiyoyasu Kurahashi; Jon Henrik Laake; Anders Larsson; Daniel F McAuley; Lia McNamee; Nicolas Nin; Haibo Qiu; Marco Ranieri; Gordon D Rubenfeld; B Taylor Thompson; Hermann Wrigge; Arthur S Slutsky; Antonio Pesenti Journal: Intensive Care Med Date: 2016-10-18 Impact factor: 17.440
Authors: Roy G Brower; Paul N Lanken; Neil MacIntyre; Michael A Matthay; Alan Morris; Marek Ancukiewicz; David Schoenfeld; B Taylor Thompson Journal: N Engl J Med Date: 2004-07-22 Impact factor: 91.245
Authors: Ary Serpa Neto; Rodrigo Octavio Deliberato; Alistair Ew Johnson; Tom J Pollard; Leo A Celi; Paolo Pelosi; Marcelo Gama de Abreu; Marcus J Schultz Journal: Intensive Care Med Date: 2019-10-08 Impact factor: 17.440
Authors: Robin Paudel; Christine A Trinkle; Christopher M Waters; Lauren E Robinson; Evan Cassity; Jamie L Sturgill; Richard Broaddus; Peter E Morris Journal: Am J Med Sci Date: 2021-09-28 Impact factor: 2.378
Authors: Michiel T U Schuijt; Liselotte Hol; Sunny G Nijbroek; Sanchit Ahuja; David van Meenen; Guido Mazzinari; Sabrine Hemmes; Thomas Bluth; Lorenzo Ball; Marcelo Gama-de Abreu; Paolo Pelosi; Marcus J Schultz; Ary Serpa Neto Journal: EClinicalMedicine Date: 2022-04-16
Authors: Liam Weaver; Anup Das; Sina Saffaran; Nadir Yehya; Timothy E Scott; Marc Chikhani; John G Laffey; Jonathan G Hardman; Luigi Camporota; Declan G Bates Journal: Ann Intensive Care Date: 2021-07-13 Impact factor: 6.925
Authors: Tommaso Pettenuzzo; Annalisa Boscolo; Alessandro De Cassai; Nicolò Sella; Francesco Zarantonello; Paolo Persona; Laura Pasin; Giovanni Landoni; Paolo Navalesi Journal: Crit Care Date: 2021-07-15 Impact factor: 9.097