Abstract: Objective To compare and analyze the epidemiological characteristics of hospitalized elderly, young and middle-aged patients with severe burn in recent years, so as to provide reference for the prevention and treatment of elderly patients with severe burn. Methods Relying on the entry system of epidemiological case data and biological sample of severe burn from multicenter in clinic, medical records of patients with severe burn, aged above 18, hospitalized in 8 burn wards from January 2012 to December 2015 were collected. Six hundred and fifteen patients who were more than 18 years old and less than or equal to 65 years old were included in young and middle-aged group (YM). Eighty-two patients aged more than 65 years old were included in elderly group (E). Data of age, gender, residence, education level, cause of injury, location of injury, season of injury, total burn area, occurrence and area of full-thickness burn injury, wound site, inhalation injury incidence and severity, post burn admission time, proportion of delayed resuscitation, proportion of escharectomy or tangential excision and skin grafting, preinjury systemic disease, system complication during hospitalization, length of hospital stay, outcome of treatment, and reason of abandoning treatment of patients were analyzed. Data were processed with chi-square test and Mann-Whitney U test. The odds ratios of preinjury systemic disease, system complication during hospitalization, and adverse outcome of patients in group YM were compared with those in group E. Results (1) The majority of patients in the two groups were male, but the proportion of male patients in group YM was higher. There was statistically significant difference in gender distribution of patients between the two groups (χ²=18.727, P<0.001). The majority of patients in the two groups were from rural areas, but the proportion of rural patients in group E was higher. There was statistically significant difference in residence distribution of patients between the two groups (χ²=9.306, P=0.002). Patients in group YM mainly had secondary education, while patients in group E mainly had primary education. There was statistically significant difference in distribution of education level of patients between the two groups (χ²=146.797, P<0.001). (2) The most common causes of injury of patients in the two groups were both flame, but the proportion of patients with flame burn injury in group E was higher. There was statistically significant difference in distribution of cause of injury of patients between the two groups (χ²=25.063, P<0.001). The main locations of injury of patients in groups YM and E were respectively public place and private residence. There was statistically significant difference in location distribution of injury of patients between the two groups (χ²=46.313, P<0.001). The main seasons of injury of patients in groups YM and E were respectively summer and winter. There was statistically significant difference in season distribution of patients between the two groups (χ²=23.143, P<0.001). There was statistically significant difference in distribution of total burn area of patients between the two groups (χ²=25.799, P=0.002). The occurrences of full-thickness burn injury of patients in the two groups were similar (χ²=2.685, P=0.101), while there was statistically significant difference in area of full-thickness burn injury of patients between the two groups (χ²=26.702, P=0.002). There was no statistically significant difference in distribution of wound site of patients between the two groups (χ²=3.954, P=0.785). There were no statistically significant differences in incidence and severity distribution of inhalation injury of patients between the two groups (with χ² values respectively 0.425 and 0.672, P values above 0.05). (3) There was statistically significant difference in distribution of admission time of patients between the two groups (χ²=6.632, P=0.036), but there was no statistically significant difference in proportion of delayed resuscitation of patients between the two groups (χ²=1.261, P=0.261). The proportion of escharectomy or tangential excision and skin grafting of patients in group YM was 72.0% (443/615), which was significantly higher than 35.4% (29/82) of group E (χ²=44.498, P<0.001). The incidence of preinjury systemic disease of patients in group YM was 13.2% (81/615), which was significantly lower than 61.0% (50/82) of group E (χ²=108.337, P<0.001). The risk of preinjury systemic disease of patients in group E was 10.30 times of that of patients in group YM [with 95% confidence interval (CI) of 6.24-17.01, P<0.001]. During hospitalization, 59.8% (49/82) of patients in group E suffered from system complications, which was significantly higher than 36.6% (225/615) of group YM (χ²=16.282, P<0.001). The risk of system complication of patients in group E was 2.57 times of patients in group YM (with 95% CI of 1.61-4.12, P<0.001). The length of hospital stay of patients in group E was significantly shorter than that of group YM (U=36 735, P<0.001). There was statistically significant difference in treatment outcome of patients between the two groups (χ²=106.251, P<0.001). The risk of adverse outcome of patients in group E was 7.52 times of group YM (with 95% CI of 4.40-12.88, χ²=67.709, P<0.001). The proportion of abandoning treatment of patients in group E was significantly higher than that of group YM (χ²=150.670, P<0.001). The risk of abandoning treatment of patients in group E was 15.86 times of that of group YM (with 95% CI of 9.36-26.88, P<0.001). There was no statistically significant difference in distribution of reason of abandoning treatment of patients between the two groups (χ²=4.178, P=0.243). Conclusions There were significant differences in the epidemiological characteristics of patients in groups E and YM. In elderly burn patients, the proportion of rural population was higher and the education level was lower. Flame burn was common and burns mostly occurred in private residences and in winter. The total burn area was slightly lower but the area of full-thickness burn injury was larger. The length of hospital stay was shorter and the proportion of surgical treatment was lower. The incidences of preinjury systemic disease and system complication during hospitalization were higher, and therefore the risks of adverse outcome and abandoning treatment were higher.