At present, people still have different opinions on choosing which kind of crystalloid solution for patients with sepsis, and there is no unified standard yet. It is necessary to conduct systematic studies on comparison of different fluid resuscitation methods on the efficacy and safety of crystalloid solution for patients with sepsis. Therefore, this study focuses on the efficacy and safety of sodium bicarbonate Ringer's solution compared with normal saline.

Sepsis and septic shock are clinical emergencies requiring immediate treatment and fluid resuscitation. Early effective fluid resuscitation can improve the prognosis of patients. So far, there is no perfect crystalloid solution. Normal saline is the most readily available and economical crystalline fluid for clinical resuscitation. It is isoosmotic and can meet the basic needs of patients in the early rehydration. But it lacks acid-base buffer system and potassium, calcium and magnesium ions, and the level of chloride ions is also significantly higher than that of plasma.

Sodium Bicarbonate Ringer's solution contains physiological levels of Na+(130mmol/L), K+(4mmol/L), Ca2+(1.5mmol/L), Mg2+(1mmol/L), HCO3-(28mmol/L) and Cl-(109mmol/L). It also has citric acid/ sodium citrate buffer system (Citrate3-1.3mmol/L), pH7.3, and osmotic pressure 276mOsm/L. It is the most similar solution to extracellular fluid.

Theoretically，Compared with normal saline, Sodium Bicarbonate Ringer's solution can maintain acid-base balance faster and better when restoring the microcirculation without affecting the level of chloride ion. Therefore we hypothesize that Sodium Bicarbonate Ringer's solution can promote the body to restore to acid-base balance more quickly, and improve organ function more significantly than normal saline.

Inclusion Criteria:

• Diagnosed as sepsis or septic shock according to the definition of Sepsis-3.0 with fluid resuscitation requirement;
• aged between 18 and 80, male or female;
• Signed informed Consent (with delay within 24 hours).

Exclusion Criteria:

• Patients with hypermagnesemia;
• Patients with hypothyroidism;
• Patients predicted to die or discharged within 24 hours after admission;
• Patients with previous history of mental illness, severe hepatic and renal insufficiency, severe cardiac disease, primary severe central nervous system lesions;
• Pregnant or breast-feeding women;
• Patients who have received cardiopulmonary resuscitation；
• Patients who participated in other clinical trials within 30 days;
• Patients have other conditions that are not appropriate for inclusion according to the researcher's judgment.