Division of Gastroenterology, Perelman School of Medicine.
Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
Eur J Gastroenterol Hepatol. 2021 Dec 1;33(1S Suppl 1):e693-e699. doi: 10.1097/MEG.0000000000002218.
Bacteremia is a common cause of death in patients with cirrhosis and early antimicrobial therapy can be life-saving. Severe liver disease impairs glucose metabolism such that hypoglycemia may be a presenting sign of infection in patients with cirrhosis. We explored this association using granular retrospective data.
We conducted a case-control analysis from 1 January 2008 to 31 December 17 in the University of Pennsylvania Health System. We identified the first blood culture results from all cirrhosis hospitalizations and obtained detailed vital sign and laboratory data in the 24-72 h prior to culture results. We used multivariable logistic regression to develop models predicting blood culture positivity and in-hospital mortality. We repeated these analyses restricted to normothermic individuals. Restricted cubic splines were used to model nonlinearity in the glucose variable.
We identified 1274 cirrhosis admissions with blood culture results (52.7% positive). In adjusted models, minimum glucose 24-72 h prior to blood culture result date was a significant predictor of blood culture positivity. In particular, glucose levels below 100 mg/dL significantly increased the probability of subsequent positive blood culture (e.g. odds ratio 1.89 for 50 mg/dL vs. 100 mg/dL, P = 0.004). This relationship persisted when restricting the cohort to normothermic individuals. Glucose levels <100 mg/dL in patients with bacteremia were also positively associated with in-hospital mortality.
Early hypoglycemia is predictive of subsequently documented bacteremia and in-hospital mortality in patients with cirrhosis, even among normothermic individuals. In patients without other overt signs of infection, low glucose values may serve as an additional data point to justify early antibiosis.
菌血症是肝硬化患者死亡的常见原因，早期抗菌治疗可能是救命的。严重的肝脏疾病会损害葡萄糖代谢，因此低血糖可能是肝硬化患者感染的首发表现。我们使用详细的回顾性数据来探索这种关联。
我们在宾夕法尼亚大学卫生系统（University of Pennsylvania Health System）进行了一项病例对照分析，时间从 2008 年 1 月 1 日至 2017 年 12 月 31 日。我们从所有肝硬化住院患者中确定了第一次血培养结果，并在培养结果前 24-72 小时内获得了详细的生命体征和实验室数据。我们使用多变量逻辑回归来建立预测血培养阳性和院内死亡率的模型。我们在仅包括体温正常个体的人群中重复了这些分析。受限立方样条用于建模葡萄糖变量中的非线性。
我们确定了 1274 例有血培养结果的肝硬化入院患者（52.7%阳性）。在调整后的模型中，血培养结果前 24-72 小时的最低血糖水平是血培养阳性的显著预测因素。特别是，血糖水平低于 100mg/dL 显著增加了随后阳性血培养的可能性（例如，50mg/dL 与 100mg/dL 相比，比值比为 1.89，P=0.004）。当将队列限制在体温正常个体时，这种关系仍然存在。菌血症患者的血糖水平<100mg/dL 也与院内死亡率呈正相关。
即使在体温正常的个体中，肝硬化患者早期低血糖也可预测随后确诊的菌血症和院内死亡率。在没有其他明显感染迹象的患者中，低血糖值可能是额外的数据分析点，可证明早期使用抗生素是合理的。
