Trends and Outcomes in Sepsis Hospitalizations With and Without Atrial Fibrillation: A Nationwide Inpatient Analysis

OBJECTIVES: Atrial fibrillation is frequently seen in sepsis-related hospitalizations. However, large-scale contemporary data from the United States comparing outcomes among sepsis-related hospitalizations with versus without atrial fibrillation are limited. The aim of our study was to assess the frequency of atrial fibrillation and its impact on outcomes of sepsis-related hospitalizations. DESIGN: Retrospective cohort study. SETTING: The National Inpatient Sample databases (2010-2014). PATIENTS: Primary discharge diagnosis of sepsis with and without atrial fibrillation were identified using prior validated International Classification of Diseases, 9th Edition, Clinical Modification codes. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Overall, 5,808,166 hospitalizations with the primary diagnosis of sepsis, of which 19.4% (1,126,433) were associated with atrial fibrillation. The sepsis-atrial fibrillation cohort consisted of older (median [interquartile range] age of 79 yr [70-86 yr] vs 67 yr [53-79 yr]; p < 0.001) white (80.9% vs 68.8%; p < 0.001) male (51.1% vs 47.5%; p < 0.001) patients with an extended length of stay (median [interquartile range] 6 d [4-11 d] vs 5 d [3-9 d]; p < 0.001) and higher hospitalization charges (median [interquartile range] $44,765 [$23,234-$88,657] vs $35,737 [$18,767-$72,220]; p < 0.001) as compared with the nonatrial fibrillation cohort. The all-cause mortality rate in the sepsis-atrial fibrillation cohort was significantly higher (18.4% and 11.9%; p = 0.001) as compared with those without atrial fibrillation. Although all-cause mortality (20.4% vs 16.6%) and length of stay (median [interquartile range] 7 d [4-11 d] vs 6 d [4-10 d]) decreased between 2010 and 2014, hospitalization charges increased (median [interquartile range] $41,783 [$21,430-$84,465] vs $46,251 [$24,157-$89,995]) in the sepsis-atrial fibrillation cohort. The greatest predictors of mortality in the atrial fibrillation-sepsis cohort were African American race, female gender, advanced age, and the presence of medical comorbidities. CONCLUSIONS: The presence of atrial fibrillation among sepsis-related hospitalizations is a marker of poor prognosis and increased mortality. Although we observed rising trends in sepsis and sepsis-atrial fibrillation-related hospitalizations during the study period, the rate and odds of mortality progressively decreased.