Abstract: Objective To observe and primarily evaluate the feasibility and validity of continuous blood purification (CBP) during the early stage of severe burn. Methods Forty-one patients with severe burn admitted to our ward from January 2013 to July 2015, conforming to the study criteria, were divided into conventional treatment group (CT, n=21) and blood purification group (BP, n=20) according to the random number table and patient's personal consent. Patients in group CT received CT conforming to the traditional resuscitation principle for severe burn, while patients in group BP received CT and blood purification treatment in the mode of continuous venous-venous hemodiafiltration in addition up to post injury hour (PIH) 72. On post injury day (PID) 1, 2, 3, the vital signs, volume of fluid input, and volume of the urine output were observed and recorded; femoral artery blood was drawn to determine lactate, bicarbonate radical, and base excess, and oxygen index was calculated. At PIH 12, 24, 48, 72, femoral vein blood was drawn to determine white cell count, platelet count, neutrophils, creatine kinase-MB, creatine kinase, lactic dehydrogenase, aspartate transaminase (AST), alanine aminotransferase (ALT), creatinine, urea nitrogen, and blood glucose (the ratio of AST to ALT was calculated). The incidence of infection, sepsis, and multiple organ dysfunction syndrome (MODS) and the mortality of patients were recorded during 2 months after injury. Data were processed with chi-square test, analysis of variance for repeated measurement, t test and Wilcoxon test, and the values of P were adjusted by Bonferroni. Results The observation was completed in the 41 patients without exclusion. (1) There were no statistically significant differences in vital signs, volume of fluid input, and volume of the urine output of patients between two groups on PID 1, 2, 3 (with t values from -1.64 to 1.48, P values above 0.05). (2) Compared with that in group CT, the level of lactate of patients in group BP declined significantly on PID 2 and 3 (with Z values respectively -2.37 and -2.46, P values below 0.05). Compared with those in group CT, the levels of bicarbonate radical and base excess of patients in group BP declined significantly on PID 3 (with t values both as -2.51, P values below 0.05). The oxygen index of patients in group BP on PID 3 was (370±98) mmHg (1 mmHg=0.133 kPa), which was significantly higher than that in group CT [(305±81) mmHg, t=2.27, P<0.05]. (3) There were no statistically significant differences in white cell count, platelet count, neutrophils, creatine kinase, lactic dehydrogenase, AST, ALT, and AST to ALT ratio of patients between two groups at PIH 12, 24, 48, 72 (with t values from -1.47 to 1.19, Z values from -1.58 to -0.03, P values above 0.05). At PIH 24, 48, 72, the levels of creatine kinase-MB and blood glucose of patients in group BP were respectively (81±43), (55±34), (58±40) U/L and (7.9±2.0), (6.7±0.9), (6.9±1.8) mmol/L, which were significantly lower than those in group CT [(179±184), (124±71), (103±57) U/L and (10.1±3.8), (9.1±2.4), (8.8±4.1) mmol/L, with Z values from -3.73 to -2.02, P<0.05 or P<0.01]. Compared with those of patients in group CT, creatinine at PIH 48 and urea nitrogen at PIH 24, 48, 72 were obviously lower in group BP (with t values from -4.23 to -2.44, P<0.05 or P<0.01). (4) During the two months after injury, the infection rate of patients in group BP was 60.0% (12/20), which was significantly lower than that in group CT [95.2% (20/21), χ²=5.51, P<0.05]. The incidence of sepsis and MODS and the mortality of patients in group BP were all lower than those in group CT, but there were no statistically significant differences (with χ² values from 0.22 to 2.93, P values above 0.05). Conclusions Conducting CBP in the early stage of severe burn is safe and feasible, which does not obviously affect the vital signs, volumes of fluid input and urine output, or platelet count of patients, additionally, it could help protect the function of vital organs, eliminate stress hyperglycemia, and reduce infection rate. Clinical trial registration Chinese Clinical Trial Registry, ChiCTR-TRC-12002616.