Diabetes and burns: retrospective cohort study.

Burn injuries are often associated with multisystemic complications, even in otherwise healthy individuals. It is therefore intuitive that for the diabetic patient, the underlying pathophysiologic alterations in vascular supply, peripheral neuropathy, and immune function could have a profoundly devastating impact on patient outcome. The effects of diabetes on morbidity and mortality of the burn-injured patient have not been examined in great detail. The purpose of this retrospective study was to compare clinical outcomes between diabetic and nondiabetic burn patients. We reviewed the charts of 181 diabetic (DM) and 190 nondiabetic (nDM) patients admitted with burns between January 1996 and May 2000, matched by sex and date of admission. Burn cause and size, time to presentation, clinical course, and outcomes were evaluated. Because age was a factor, the analysis was done by three age groups: younger than 18 years, 18 to 65 years, and older than 65 years. Of patients 18 to 65 years, 51% (98/191) were diabetic, whereas 84% (81/96) of those older than 65 and only 4% (3/85) of patients younger than 18 were diabetic. Because of the disproportion in numbers of diabetics compared with nondiabetics in the younger than 18 and older than 65 years-old groups, these patients will not be discussed. Diabetics were more likely to incur scald injury from tub or shower water rather than hot fluid spills (33% DM vs 15% nDM; P < or = 0.01), and have a delayed presentation (45 vs 23%; P = 0.00001). There was no difference in total burn size in all groups. Diabetics in the 18 to 65 years group had a higher rate of full-thickness burns (51 vs 31%; P = 0.025), skin grafts (50 vs 28%; P = 0.01) and burn-related procedures (57 vs 32%; P = 0.001), infections (65 vs 51%; P = 0.05), and longer lengths of stay (23 vs 12 days; P = 0.0001). Although there was no statistically significant difference in incidence of specific infections, the rates of cellulitis, wound infection, urinary tract infection, line infection, and osteomyelitis, were consistently higher in the diabetic population. Partial graft slough was 6% in diabetics 18 to 65 years with a 3% regraft rate, whereas nondiabetics had a 1% regraft rate. Comparing diabetics with controlled vs uncontrolled glucose levels, diabetics with uncontrolled glucose had higher rates of infection (72 vs 55%; P < or = 0.025), all burn-related procedures (68 vs 45%; P < or = 0.025), and longer ICU stays (24 vs 10 days; P = 0.048). Mortality rate was 2% for diabetics and for nondiabetics. In summary, presence of diabetes in the burn patient was associated with a worse outcome. A predilection for burn injuries in the diabetic was noted in the older adult population. Deeper burns, delayed presentation, higher rates of infection, graft failure and operations, and longer lengths of stay translate into an increased cost to society both economically and in lives. This data would suggest a need for better burn education for diabetics and health care professionals, recognizing the elderly population as a "high-risk" group. We believe that targeted prevention measures and treatment strategies, emphasizing earlier and more aggressive intervention for this population, may have a favorable effect on morbidity and mortality.