Accuracy of vasopressin secretion in the late phase of septic shock for predicting patient outcomes

Objective To evaluate the accuracy of vasopressin (VP) secretion in the late phase of septic shock for predicting patient outcomes and further investigate its relationship with the prognosis of septic shock. Methods Fifty-five patients presented at late phase of septic shock, who were admitted to the intensive care unit of our hospital, were enrolled.Their VP secretion was measured.The method for measurement was as follows: 3% sodium chloride solution 600 ml was infused over 2 h, serum concentrations of VP and sodium were measured before and after infusion, the difference in VP before and after infusion (△VP) and in Na before and after infusion (△Na) was calculated, and △VP/△Na was used to reflect VP secretion.The patients were divided into either abnormal secretion of VP group (△VP/△Na≤0.5 ng/mmol) or normal secretion of VP group (△VP/△Na>0.5 ng/mmol) according to △VP/△Na ratio.Immediately before testing VP secretion, venous blood samples were collected for determination of serum lactic acid and C-reactive protein concentrations.The consumption of vasoactive drugs at the moment of enrollment and 28-day fatality rate were recorded. Results There were 30 cases in abnormal group (54%) and 25 cases in normal group (46%). Compared with normal group, the serum lactic acid, C-reactive protein concentrations and consumption of dopamine or norepinephrine were significantly increased, and the 28-day fatality rate was increased (67% vs 40%) in abnormal group.ROC curve analysis showed that when ΔVP/ΔNa 0.5 ng/mmol was used as the criteria for determining prognosis, the sensitivity was 66.7%, specificity was 64.0%, and the area under the ROC curve was 0.828. Conclusion VP secretion in the late phase of septic shock may affect patient prognosis. Key words: Vasopressins; Secretory rate; Shock, septic; Prognosis