A quantitative assessment of how Canadian intensivists believe they utilize oxygen in the intensive care unit

OBJECTIVES: To investigate attitudes and practices regarding oxygen therapy in intensive care units (ICUs) and to devise quantitative descriptive indices. SETTING: Canadian university-affiliated adult ICUs. PARTICIPANTS: Fifty-two medical directors of ICUs in 48 institutions. INTERVENTION: Structured postal questionnaire returned by 48 participants. MEASUREMENTS AND MAIN RESULTS: Attitudes, beliefs, and stated practices relating to oxygen use in ICUs were determined. Novel descriptors S-50min (minutes of oxygen saturation [Sao2] acceptable to >50% of respondents), F-50max (maximum F(IO)2 above which <50% of respondents would increase F(IO)2), and F-50min (minimum F(IO)2 below which <50% of respondents would decrease F(IO)2) were determined. All respondents believed that oxygen toxicity was a concern. Twenty-nine percent of respondents indicated that they did not always assess tissue oxygenation in critical cases. A stepwise reduction in acceptance of progressive desaturation and increasing duration of hypoxemia was found. Presented with a stable patient with Sao2 of 98%, the maximum level of F(IO)2 above which respondents stated that they would not increase the F(IO)2 was 0.41+/-0.17 (mean +/- SD). For stable patients with Sao2 of 85%, the minimum F(IO)2 below which respondents would not reduce F(IO)2 was 0.59+/-0.23 (mean +/- SD). F-50max was 0.8 vs. 0.5 for Sao2 of 80%-85% vs. 85%-90%, respectively; F-50min was 0.6 vs. 0.21 for Sao2 of 90%-95% vs. 95%-100%, respectively. CONCLUSIONS: Considerable variation exists in the attitudes, beliefs, and stated practices relating to the management of oxygen therapy in the ICU. These data are amenable to quantitative description and illustrate the necessity for documentation of actual practice and development of support systems for decision-making in this and similar areas.