[Prognostic value of serum sodium variability within 72 hours and lactic acid combined with severity score in patients with sepsis]

Abstract

Objective: To explore the predictive value of serum sodium variability within 72 hours, lactic acid (Lac), sequential organ failure assessment (SOFA) and acute physiology and chronic health evaluation II (APACHE II) in predicting the 28-day prognosis of sepsis patients.

Methods: The clinical data of patients with sepsis admitted to the department of intensive care unit (ICU) of the Affiliated Qingdao Municipal Hospital of Qingdao University from December 2020 to December 2021 were retrospectively analyzed, including age, gender, previous medical history, temperature, heart rate, respiratory rate, systolic pressure, diastolic pressure, white blood cell count (WBC), hemoglobin (Hb), platelet count (PLT), C-reactive protein (CRP), pH value, arterial partial pressure of oxygen (PaO₂), arterial partial pressure of carbon dioxide (PaCO₂), Lac, prothrombin time (PT), activated partial thromboplastin time (APTT), serum creatinine (SCr), total bilirubin (TBil), albumin (Alb), SOFA, APACHE II score, and 28-day prognosis. Multivariate Logistic regression was used to analyze the risk factors of death in sepsis patients. Receiver operator characteristic curve (ROC curve) was used to analyze the predictive value of serum sodium variability within 72 hours, Lac, SOFA, APACHE II alone and in combination on the prognosis of patients with sepsis.

Results: A total of 135 patients with sepsis were included, 73 survived and 62 died at 28 days, with 28-day mortality of 45.93%. (1) Compared with the survival group, SOFA, APACHE II, Lac and serum sodium variability within 72 hours in the death group were significantly higher [SOFA: 10.00 (8.00, 12.00) vs. 6.00 (5.00, 8.00), APACHE II: 18.00 (16.00, 21.25) vs. 13.00 (11.00, 15.00), Lac (mmol/L): 3.55 (2.90, 4.60) vs. 2.00 (1.30, 2.80), serum sodium variability within 72 hours: 3.4% (2.6%, 4.2%) vs. 1.4% (1.1%, 2.5%)], the differences were statistically significant (all P < 0.01). (2) Multivariate Logistic regression showed that SOFA, APACHE II, Lac, serum sodium variability within 72 hours were independent risk factors of prognosis in patients with sepsis [SOFA: odds ratio (OR) = 1.479, 95% confidence interval (95%CI) was 1.114-1.963, P = 0.007; APACHE II: OR = 1.163, 95%CI was 1.009-1.340, P = 0.037; Lac: OR = 1.387, 95%CI was 1.014-1.896, P = 0.040; serum sodium variability within 72 hours: OR = 1.634, 95%CI was 1.102-2.423, P = 0.015]. (3) ROC curve analysis showed that SOFA, APACHE II, Lac and serum sodium variability within 72 hours had certain predictive value for the prognosis of sepsis patients [SOFA: the area under the ROC curve (AUC) = 0.858, 95%CI was 0.795-0.920, P = 0.000; APACHE II: AUC = 0.845, 95%CI was 0.776-0.913, P = 0.000; Lac: AUC = 0.840, 95%CI was 0.770-0.909, P = 0.000; serum sodium variability within 72 hours: AUC = 0.842, 95%CI was 0.774-0.910, P = 0.000]. The combined predictive value of the four indicators (AUC = 0.917, 95%CI was 0.870-0.965, P = 0.000) was higher than that of any single indicator, and has higher specificity (79.5%) and sensitivity (93.5%), indicating that the combined index has higher predictive value for the prognosis of sepsis patients than any single index.

Conclusions: SOFA, APACHE II, Lac, serum sodium variability within 72 hours are independent risk factors for 28-day death in patients with sepsis. The combination of SOFA score, APACHE II score, Lac and serum sodium variability within 72 hours has higher predictive value for prognosis than single index.