Chokengarmwong Nalin, Yeh Daniel Dante, Chang Yuchiao, Ortiz Luis Alfonso, Kaafarani Haytham M A, Fagenholz Peter, King David R, DeMoya Marc, Butler Kathryn, Lee Jarone, Velmahos George, Januzzi James Louis, Lee-Lewandrowski Elizabeth, Lewandrowski Kent
From the Department of Anesthesiology, King Chulalongkorn Memorial Hospital (N.C.), Thai Red Cross Society (N.C.), Department of Anesthesiology (N.C.), Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; Division of Trauma, Emergency Surgery and Surgical Critical Care (N.C., D.D.Y., L.A.O., H.M.A.K., P.F., D.R.K., M.D., K.B., J.L., G.V.), Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, Department of Medicine (Y.C.), Massachusetts General Hospital and Harvard Medical School; Division of Cardiology (J.L.J.), Massachusetts General Hospital and Cardiometabolic Trials, Harvard Clinical Research Institute; Department of Pathology (E.L., K.L.), Massachusetts General Hospital; and Harvard Medical School (E.L., K.L.), Boston, Massachusetts.
J Trauma Acute Care Surg. 2017 Sep;83(3):485-490. doi: 10.1097/TA.0000000000001552.
New onset atrial fibrillation (AF) in critically ill surgical patients is associated with significant morbidity and increased mortality. N-terminal pro-B type natriuretic peptide (NT-proBNP) is released by cardiomyocytes in response to stress and may predict AF development after surgery. We hypothesized that elevated NT-proBNP level at surgical intensive care unit (ICU) admission predicts AF development in a general surgical and trauma population.
From July to October 2015, NT-proBNP concentrations were measured at ICU admission. Abnormal NT-proBNP concentrations were defined by age-adjusted cut-offs. We examined the relationship between the development of AF and demographics, clinical variables, and NT-proBNP level using univariate analysis and a multivariable logistic regression model.
Three hundred eighty-seven subjects were included in the cohort, none of whom were in AF at ICU admission. The median age was 63 years (52-73 years), and 40.3% were women. The risk of developing AF was higher for abnormal versus normal NT-proBNP (22% vs. 4%; p < 0.0001). Using optimal derived cutoffs (regardless of age), the risk of developing AF was 2% for NT-proBNP less than 600 ng/L, 15% for NT-proBNP of 600 ng/L to 1,999 ng/L, and 27% for NT-proBNP of 2,000 ng/L or greater. Multiple logistic regression analysis identified three independent predictors for new-onset AF: age, older than 70 years (odds ratio [OR], 3.7, 95% confidence interval [CI], 1.5-9.3), history of AF (OR, 25.3; 95% CI, 9.6-67.0), and NT-proBNP of 600 or greater (OR, 4.3; 95% CI, 1.3-14.2). When none or only one predictor was present, AF incidence was less than 1%. When all three predictors were present, AF incidence was 66%. For subjects 70 years or older but no history of AF, AF incidence was 12.8% when NT-proBNP was 600 or greater compared with 0% when NT-proBNP was less than 600. For subjects younger than 70 years with a history of AF, AF incidence was 44.4% when NT-proBNP was 600 or higher compared to 0% when NT-proBNP was less than 600.
Elevated NT-proBNP at ICU admission in general surgical and trauma patients is predictive of AF development in the first 3 ICU days. Addition of NT-proBNP measurement to known risk factors can improve predictive power and identify patients who might potentially benefit from evidence-based prophylactic treatment for AF.
危重症外科患者新发房颤（AF）与显著的发病率和死亡率增加相关。N末端B型利钠肽原（NT-proBNP）由心肌细胞在应激反应时释放，可能预测术后房颤的发生。我们假设手术重症监护病房（ICU）入院时NT-proBNP水平升高可预测普通外科和创伤患者房颤的发生。
2015年7月至10月，在ICU入院时测量NT-proBNP浓度。异常NT-proBNP浓度通过年龄校正的临界值定义。我们使用单因素分析和多变量逻辑回归模型研究房颤发生与人口统计学、临床变量及NT-proBNP水平之间的关系。
该队列纳入387名受试者，他们在ICU入院时均未发生房颤。中位年龄为63岁（52 - 73岁），40.3%为女性。NT-proBNP异常者发生房颤的风险高于正常者（22%对4%；p < 0.0001）。使用最佳推导临界值（不考虑年龄），NT-proBNP低于600 ng/L时发生房颤的风险为2%，NT-proBNP为600 ng/L至1999 ng/L时为15%，NT-proBNP为2000 ng/L或更高时为27%。多变量逻辑回归分析确定了新发房颤的三个独立预测因素：年龄大于70岁（比值比[OR]，3.7，95%置信区间[CI]，1.5 - 9.3）、房颤病史（OR，25.3；95% CI，9.6 - 67.0）以及NT-proBNP为600或更高（OR，4.3；95% CI，1.3 - 14.2）。当无或仅有一个预测因素时，房颤发生率低于1%。当三个预测因素均存在时，房颤发生率为66%。对于70岁及以上但无房颤病史的受试者，NT-proBNP为600或更高时房颤发生率为12.8%，而NT-proBNP低于600时为0%。对于有房颤病史的70岁以下受试者，NT-proBNP为600或更高时房颤发生率为44.4%，而NT-proBNP低于600时为0%。
普通外科和创伤患者ICU入院时NT-proBNP升高可预测ICU前3天房颤的发生。将NT-proBNP测量添加到已知危险因素中可提高预测能力，并识别可能从基于证据的房颤预防性治疗中潜在获益的患者。
