Comparison of iloprost therapy versus non-iloprost therapy for severe frostbite: a retrospective cohort study of two Canadian cities

Objectives Frostbite is a common reason for emergency department (ED) presentations in Canada. Iloprost, a prostacyclin analogue, has been investigated to reduce the risk of amputation with its use expanding. Two Canadian cities implemented iloprost over different times leading to a practice variation that allowed for treatment comparison. Our objective is to evaluate the effectiveness of iloprost compared with non-iloprost treatment. Secondary objectives include assessing the impact of iloprost dosage and homelessness. Methods A retrospective cohort study was conducted on adult severe frostbite cases presenting to EDs in Calgary and Edmonton between November 2021 and April 2024. Data were abstracted from clinical databases and analysed for demographic and injury characteristics, treatment and amputation outcomes. Results Of 1812 total ED encounters for frostbite, 257 patients with grades 2–4 extremity frostbite were included for analysis. Logistic regression found that overall patients receiving iloprost were associated with reduced likelihood of any amputation (OR=0.49, 95% CI 0.25 to 0.96) and fewer digit amputations (p<0.001). In particular, iloprost use was associated with reduced amputation rates in grade 3 injuries compared with non-iloprost care (30% vs 52%, p=0.042). Although over 80% of patients with grade 4 frostbite required amputations regardless of iloprost treatment, its use was associated with improved digit salvage, with 10% fewer digit segments requiring amputation in treated patients in both hands (p=0.049) and feet (p=0.003). When assessing 65 patients who received iloprost for limb frostbite, higher iloprost doses were associated with a lower likelihood of amputation (OR=0.35; 95% CI 0.13 to 0.91; p=0.03). In the patient subgroup who underwent at least one amputation, total iloprost dosage was inversely related to outcome (p=0.033, βST=−0.26). Adverse events were reported in 61% of iloprost-treated patients, with headache being the most common. Homelessness was found to be a significant predictor of delays in arrival to the ED after injury (p<0.001); OR=2.90 (95% CI 1.73 to 4.91). Conclusions Iloprost infusion was associated with a reduction in amputation rates in grade 3 and 4 frostbite with the greatest association seen in grade 3 cases. Greater iloprost dosage was associated with improved digit salvage. Homelessness was associated with delayed ED presentation.