Introduction It is uncertain whether treatment by recombinant human granulocyte macrophage colony-stimulating factor (rhGM-CSF) can promote healing of deep second-degree burns. This meta-analysis aimed to systematically review and assess randomised controlled trials (RCTs) that investigated the efficacy and safety of rhGM-CSF for treating deep second-degree burns.

Methods This meta-analysis conformed to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement. The PubMed, Cochrane Library, Medline and Embase databases and relevant references were systematically searched for RCTs (published up to November 2019). Main outcome measures included the wound healing rate, wound healing time and average optical densities of the vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF). We performed a meta-analysis using fixed or random effects models.

Results Seven RCTs comprising 982 patients with 1184 burns (652 patients received rhGM-CSF vs 532 controls) were included. Compared with standard wound care alone, the use of rhGM-CSF significantly reduced wound healing time by 4.77 days (weighted mean difference=−4.77; 95% CI −6.45 to −3.09; p<0.001) and significantly increased the wound healing rate on days 7, 10, 14 and 20 by 6.46%, 19.78%, 17.07% and 11.38%, respectively. There was no significant difference between the groups in the wound healing rate on day 28 and average optical densities of VEGF and FGF. No systematic adverse event occurred. Redder, more swollen and painful wounds were reported after using rhGM-CSF compared with the control.

Conclusions rhGM-CSF could be effective and safe for treating deep second-degree burns.