Clinical Practice Variability in Temperature Correction of Arterial Blood Gas Measurements and Outcomes in Hypothermia-Treated Patients After Cardiac Arrest

Abstract

Mechanical ventilation in patients treated with mild therapeutic hypothermia (MTH) for the postcardiac arrest syndrome may be challenging given changes in solubility of arterial blood gases (ABGs) with cooling. Whether ABG measurements should be temperature corrected (TC) remain unknown. We sought to describe practice variability in TC at a single institution and explored the association between TC and neurological outcome. We conducted a retrospective cohort study reviewing electronic health records of all patients treated with MTH after cardiac arrest. We examined whether the percentage of TC ABGs relative to total number of ABGs drawn for each subject during hypothermia was associated with the neurological outcome at hospital discharge and 6–12-month follow-up. The cerebral performance category of 1–2 was defined as a favorable outcome in the logistic regression models. 1223 ABGs were obtained during MTH on 122 subjects over 6 years. TC was never used in 72 subjects (59%; no TC group), made available in 1–74% of ABGs in 17 subjects (14%; intermediate TC group), and made available in ≥75% of ABGs in 33 subjects (27%; mostly TC group). Groups differed in the proportion of subjects with shockable presenting rhythms (47% vs. 47% vs. 76%, p=0.02) and admitting ICU (p=0.005). Favorable 6-month outcomes were more common in the mostly TC than no TC group (48% vs. 25%; OR [95% CI]: 2.9 [1.2–7.1]), but not after adjustment (OR 1.5, 95% CI 0.33–6.9). There was substantial practice variability in the temperature correction strategy. Availability of temperature-corrected ABGs was not associated with improved neurological outcomes after adjusting for covariates.