• Samuel M Brown, Abhijit Duggal, Peter C Hou, Mark Tidswell, Akram Khan, Matthew Exline, Pauline K Park, David A Schoenfeld, Ming Liu, Colin K Grissom, Marc Moss, Todd W Rice, Catherine L Hough, Emanuel Riv... more ers, B Taylor Thompson, Roy G Brower, and National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) Prevention and Early Treatment of Acute Lung Injury (PETAL) Network. less
• 1Pulmonary and Critical Care Medicine, Intermountain Medical Center/University of Utah School of Medicine, Salt Lake City, UT.2Department of Critical Care, Respiratory Institute, Cleveland Clinic, Clevelan... more d, OH.3Division of Emergency Critical Care Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA.4Division of Pulmonary and Critical Care Medicine, Baystate Medical Center, Springfield, MA.5Oregon Health & Science University, Portland, OR.6Ohio State University Medical Center, Columbus, OH.7Division of Acute Care Surgery, University of Michigan, Ann Arbor, MI.8Biostatistics Center, Massachusetts General Hospital, Boston, MA.9Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado School of Medicine, Aurora, CO.10Division of Allergy, Pulmonary, and Critical Care Medicine, Vanderbilt University School of Medicine, Nashville, TN.11Division of Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, WA.12Henry Ford Hospital, Detroit, MI.13Division of Pulmonary and Critical Care Medicine, Massachusetts General Hospital, Boston, MA.14Pulmonary and Critical Medicine, Johns Hopkins University School of Medicine, Baltimore, MD. less
• Crit. Care Med. 2017 Aug 1; 45 (8): 1317-1324.

ObjectivesIn the contemporary ICU, mechanically ventilated patients may not have arterial blood gas measurements available at relevant timepoints. Severity criteria often depend on arterial blood gas results. Retrospective studies suggest that nonlinear imputation of PaO2/FIO2 from SpO2/FIO2 is accurate, but this has not been established prospectively among mechanically ventilated ICU patients. The objective was to validate the superiority of nonlinear imputation of PaO2/FIO2 among mechanically ventilated patients and understand what factors influence the accuracy of imputation.DesignSimultaneous SpO2, oximeter characteristics, receipt of vasopressors, and skin pigmentation were recorded at the time of a clinical arterial blood gas. Acute respiratory distress syndrome criteria were recorded. For each imputation method, we calculated both imputation error and the area under the curve for patients meeting criteria for acute respiratory distress syndrome (PaO2/FIO2 ≤ 300) and moderate-severe acute respiratory distress syndrome (PaO2/FIO2 ≤ 150).SettingNine hospitals within the Prevention and Early Treatment of Acute Lung Injury network.PatientsWe prospectively enrolled 703 mechanically ventilated patients admitted to the emergency departments or ICUs of participating study hospitals.InterventionsNone.Measurements And Main ResultsWe studied 1,034 arterial blood gases from 703 patients; 650 arterial blood gases were associated with SpO2 less than or equal to 96%. Nonlinear imputation had consistently lower error than other techniques. Among all patients, nonlinear had a lower error (p < 0.001) and higher (p < 0.001) area under the curve (0.87; 95% CI, 0.85-0.90) for PaO2/FIO2 less than or equal to 300 than linear/log-linear (0.80; 95% CI, 0.76-0.83) imputation. All imputation methods better identified moderate-severe acute respiratory distress syndrome (PaO2/FIO2 ≤ 150); nonlinear imputation remained superior (p < 0.001). For PaO2/FIO2 less than or equal to 150, the sensitivity and specificity for nonlinear imputation were 0.87 (95% CI, 0.83-0.90) and 0.91 (95% CI, 0.88-0.93), respectively. Skin pigmentation and receipt of vasopressors were not associated with imputation accuracy.ConclusionsIn mechanically ventilated patients, nonlinear imputation of PaO2/FIO2 from SpO2/FIO2 seems accurate, especially for moderate-severe hypoxemia. Linear and log-linear imputations cannot be recommended.

26 keywords...

hide…