Abstract Su907: Post-Cardiac Arrest Potassium Abnormalities Associated With Unfavorable Functional Outcome

Background: Abnormal potassium levels are common findings in the intensive care unit (ICU) population. We aimed to determine the incidence of dyskalemias at ICU admission and their association with functional outcome in comatose patients resuscitated from cardiac arrest. Hypothesis: We hypothesized that both hypokalemia and hyperkalemia are associated with unfavorable functional outcome. Methods: Pooled data from four randomized clinical trials in comatose post-cardiac arrest patients admitted to ICU after return of spontaneous circulation (ROSC). Reference potassium levels were defined as between 3 and 4.9 mmol/L, as proposed in the Simplified Acute Physiology Score II. Favorable functional outcome was defined as a Cerebral Performance Category of 1 or 2 at 180 days. Results: We included 1133 patients (557 from HYPERION, 346 from TTH48, 120 from COMACARE and 110 from Xe-HYPOTHECA) with a median age of 64 (IQR: 55-72) years and a predominance of males (72%). Overall, 421 (36%) patients had favorable functional outcome. On admission, 221 (19.5%) patients experienced hyperkalemia and 35 (3.1%) patients experienced hypokalemia. More patients in the normokalemia group (364/877, 41.5%) had a favorable functional outcome, as compared to the hypokalemia (11/35, 31.4%) and hyperkalemia (41/221, 18.6%) groups p<0.001). In a mixed-effects logistic regression model including initial rhythm, delay from collapse to ROSC, bystander cardiopulmonary resuscitation, lactate, and urea and with normokalemia as reference group, hyperkalemia was associated with higher odds for unfavorable functional outcome (OR 1.85, 95% CI 1.10-3.12, p=0.02) whereas hypokalemia was not (OR 1.36 95% CI 0.51-3.60, p=0.53). Conclusions: One fifth of unconscious cardiac arrest patients experienced dyskalemia on ICU admission. Hyperkalemia was associated with unfavorable functional outcome at 180 days compared to normokalemia, whereas hypokalemia was not an independent predictor of outcome.