Is Hyperoxia Harmful After Resuscitation from Cardiac Arrest?
Is Hyperoxia Harmful After Resuscitation from Cardiac Arrest?
Abstract & Commentary
By David J. Pierson, MD, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle. Dr. Pierson reports no financial relationships relevant to this field of study.This article originally appeared in the November 2010 issue of Critical Care Alert. It was peer reviewed by William Thompson, MD. Dr. Thompson is Associate Professor of Medicine, University of Washington. Dr. Thompson reports no financial relationships relevant to this field of study.
Synopsis: In this study from 120 hospitals in the Project IMPACT database, the presence of hyperoxia (arterial PO2 300 mm Hg or higher) in the first 24 hours after resuscitation from cardiac arrest was associated with a worse in-hospital mortality than either normoxia or hypoxia.
Source: Kilgannon JH, et al; for the Emergency Medicine Shock Research Network (EMShockNet) Investigators. Association between arterial hyperoxia following resuscitation from cardiac arrest and in-hospital mortality. JAMA. 2010;303:2165-2171.
Project impact, a proprietary database originally created by the Society for Critical Care Medicine and now maintained by Cerner Corp., collects data from a voluntary consortium of ICUs across America. In this study, Kilgannon et al used data collected from 120 Project IMPACT hospitals between 2001 and 2005 to look for associations between initial arterial PO2 obtained in the ICU and patient outcome following resuscitation from cardiac arrest.
Data from adult patients with non-traumatic cardiac arrest, either out-of-hospital or occurring in hospitalized patients, within 24 hours of ICU admission were examined. The presence of hyperoxia, normoxia, or hypoxia according to the first PO2 entered into the database during the initial 24 hours in the ICU was correlated with in-hospital mortality. Hyperoxia was defined as a PO2 of 300 mm Hg or higher; hypoxia was either a PO2 < 60 mm Hg or PaO2/FIO2 (P/F) < 300 mm Hg; normoxia was either a PO2 > 60 mm Hg or a P/F > 300 mm Hg with PO2 < 300 mm Hg. Statistical attempts were made to control for potential confounders such as age, pre-admission functional status, comorbid conditions, and vital signs.
Of the 8736 eligible patients, 2410 did not have an arterial blood gas recorded within the first 24 hours in the ICU. Among the other 6326 patients, 1156 (18%) had hyperoxia, 3999 (63%) hypoxia, and 1171 (19%) normoxia on the initial ICU blood gas specimen. Hyperoxia was associated with higher in-hospital mortality (63%; 95% confidence interval [CI], 60%-66%) as compared to the normoxia group (45%; 95% CI, 43%-48%) and the hypoxia group (57%; 95% CI, 56%-59%). After correcting for the potential confounders, initial hyperoxia had an odds ratio for death of 1.8 (95% CI, 1.5-2.2). The authors conclude that exposure to hyperoxia following cardiac arrest is an independent predictor of a worse outcome in the form of in-hospital mortality.
Commentary
The findings of this study support the notion that exposure of the brain and other tissues to hyperoxia following return of spontaneous circulation after cardiac arrest is harmful, perhaps through the generation of free oxygen radicals. Based on these findings, the authors call for clinical trials of controlled reoxygenation during the post-resuscitation period. While the results fit nicely with our concept of pathophysiology, there are a couple of troubling issues with the study with respect to its design and the potential generalizability of the findings.
The first issue is how exposure to the variable of interest was identified both in terms of definition and with respect to duration of exposure. The 3 categories of arterial oxygenation, as determined on the first arterial blood specimen recorded after the patient arrived in the ICU, do not correspond to any physiologic categorization I can figure out. Presumably, oxygen free radicals are generated in some relation to tissue oxygen exposure that is, to the number of oxygen molecules to which vulnerable cells are exposed. This should correlate with tissue PO2, which would be approximated most closely by capillary oxygen content, and next best (in the absence of hemoglobin concentration) by arterial PO2 but not the concentration of inspired oxygen or the alveolar-to-arterial PO2 gradient, for which P/F is a surrogate. One would expect that arterial PO2 and not P/F would best assess the variable of interest. In the present study, however, according to the criteria used, patients in the normoxia and hypoxia groups could both have PO2 values as high as 299 mm Hg (PO2 between 60 and 299 mm Hg in the former, and PO2 up to 299 mm Hg in the latter).
This concern could be addressed if the paper reported actual PO2 values in the 3 groups, but it does not. Thus, the magnitude of differences in the variable of greatest interest among the groups, in relation to the reported outcome, is unknown. In addition, in keeping with the hypothesis being investigated, the duration of exposure to hyperoxia would be expected to be an important variable. As it is, the time from resuscitation to identification of oxygenation status is unknown, other than the patient entered the ICU within 24 hours after resuscitation and the blood gas was obtained within 24 hours after that. Presumably, the reasons for this, and for the unusual definitions used with respect to oxygenation, have to do with what was available in the database.
A second concern relates to the potential generalizability of the findings beyond the hospitals that furnished the data. Project IMPACT collects data from a voluntary consortium of ICUs rather than from a defined subset of U.S. ICUs selected by specified criteria. According to the authors, the 120 hospitals furnishing data for the present study were mainly large non-academic community hospitals. How the results of this study might apply to patients and their management in institutions with different demographics is unknown.
These concerns notwithstanding, the findings of the present study support the concept that, after resuscitation from cardiac arrest, adequate oxygenation should be provided to avoid the adverse effects of tissue hypoxia, but excessive exposure to oxygen beyond that necessary for adequate arterial oxygenation should be avoided. The study's results are consistent with the most recent recommendations of the International Liaison Committee on Resuscitation that arterial saturation be maintained between 94% and 96% following resuscitation from cardiac arrest.1
Reference
1. Neumar RW, et al. Post-cardiac arrest syndrome: Epidemiology, pathophysiology, treatment, and prognostication. Circulation. 2008;118:2452-2483.
In this study from 120 hospitals in the Project IMPACT database, the presence of hyperoxia (arterial PO2 300 mm Hg or higher) in the first 24 hours after resuscitation from cardiac arrest was associated with a worse in-hospital mortality than either normoxia or hypoxia.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.