Contribution of thermal imaging in determining the depth of pediatric acute burns.

Abstract

Early burn wound assessment is a major problem in pediatrics. It involves regular clinical evaluations, and may be helped by objective tools, such as thermal imagers. Because of the skin temperature difference between partial and full thickness burns, due to injured vascular perfusion of the dermis in the latter, it could precisely identify areas with superficial burns which could heal from controlled wound healing, and those with deep burns which would need skin graft, at an early stage. The objective of this study was to assess the validity of the Flir one thermal imager® for this purpose. Thermal images were obtained by a single observer three times after burn: day 1-3 (T1); day 4-7 (T2); day 8-10 (T3). ΔT (temperature difference between burned area and healthy skin) was calculated on a tablet with Flir One application. Validity was assessed by comparing ΔT obtained at each time between two groups at day 15 depending on the healing time of the burn wounds : before or after day 15. Primary outcome was the AUC (area under curve) of the ROC (Receiver operating characteristic) curve for ΔT at T1, T2 and T3, representing the ability of the Flir One Thermal Imager® to discriminate between a healing time <15 days and >15 days. 40 patients were included (13 months to 13 years old). AUC at T1 and T2 are 0.700 (95% CI: 0.649; 0.925) and 0.787 (95% CI: 0.523; 0.877), respectively. AUC at T3 is 0.968 (95% CI: 0.918; 1.00). A ΔT threshold of -1.2°C at T3 is associated with a specificity of 100% allowing to diagnose as of day 8 all the patients having deep burns which heal after 15 days. The Flir One Thermal Imager® is an inexpensive, non-invasive, and easy-to-use device. It measures indirectly but with an excellent specificity the dermis perfusion reflecting the burns depth in children. However its low sensitivity calls for additional research to allow skin grafts earlier in common practice and decrease the length of stay.