Serum anion gap predicts in-hospital and 1-year mortality in cardiac care unit patients

Abstract

Abstract Background Anion gap (AG) has been used as a prognostic indicator in critical care setting as well ascardiovascular fields such as coronary artery and aortic disease. However, there were few studies about the association of AG in cardiac care unit (CCU) for in-hospital mortality and none of them for the 1-year mortality. Objective To assess a predictive ability of AG for in-hospital and 1-year mortality in CCU patients in a tertiary hospital in Thailand. Methods This is a retrospective cohort study in the single tertiary care hospital in the Southern Thailand. The data were extracted from the Hospital information System (HIS) database of the hospital. Patients admitted from emergency department (ED) to CCU from July 2015 to December 2019 were included. Anion gap was calculated from serum electrolytes obtained at emergency room or within 1 hour of CCU admission by using the follow formula: [(Na (mmol/L)) − (Cl (mmol/L) + HCO3 (mmol/L))]. The associations between anion gap and in-hospital and 1-year mortality were analyzed in the logistic and Cox regression model respectively. Results A total of 1271 patients were included in this study. There were higher prevalence of diabetesmellitus and chronic kidney disease (CKD) in patients with wide AG than in normal AG. Hemoglobin level,blood urea nitrogen and serum creatinine were higher whereas mean Braden score was lower in the wide AG group than normal AG group. The 1-year survival rate of patients with AG ≥14 was significantly lower than patients with AG <14 (Figure 1). The survival rate of patients with AG ≤12 was not significantly different from that between 12 and 14, since survival curves of the two groups were overlapping. Figure 2 illustrated strong linear relationships between baseline AG and in-hospital (A) and 1-yr mortality risk (B). The c-statistics of AG for in-hospital (A) and 1-yr mortality (B) were 0.822 (95% CI 0.763–0.878) and 0.787 (95% CI 0.75–0.824) respectively. In multivariable analysis, AG gap was an independent predictor for both in-hospital (odd ratio 1.13, 95% CI 1.07–1.19, p<0.ehab724.15231) and 1-year mortality (HR 1.09, 95% CI 1.07–1.12, p<0.0001) after adjusted for age, sex, diabetes, CKD, chronic lung disease, history of cancer, prior stroke, Braden score and cardiogenic shock. Conclusions Anion gap, costless fifth electrolyte - calculated from 3 out of 4 routine basic electrolytes (Na KCl HCO3), was a strong independent predictor of in-hospital and 1-year mortality in CCU patients. Moreover, in the future, it may lead to web-based risk calculator development for CCU patients. Funding Acknowledgement Type of funding sources: Public hospital(s). Main funding source(s): Prince of Songkla University Hospital KM curve of all cause mortalityLinear and ROC curve