Amputation-Free Survival and Time-to-Healing in Diabetic Foot Ulcers: A Multistate Competing-Risks Survival Analysis

Asmat Burhan (1School of Nursing, College of Nursing, Taipei Medical University, Taipei)
Maria Angelica Dela Cruz (College of Nursing, Visayan Highlands University, Cebu, Philippines)
Grace Tan Wei Ling (School of Health Sciences, Lakeside University, Jurong, Singapore)
Indah Susanti (School of Nursing, Faculty of Health, Universitas Harapan Bangsa, Indonesia)
Napat Kittisak (Department of Community Health Nursing, Eastern Valley University, Si Racha, Thailand)
Eza Kemal Firdaus (School of Nursing, Faculty of Health, Universitas Tanjungpura, Indonesia)
Septian Mixrova Sebayang (School of Nursing, Faculty of Health, Universitas Harapan Bangsa, Indonesia)

Journal of Wound Research and Technology
Vol. 2 No. 2 (2025): June - October 2025

Background: Diabetes-related foot ulcers (DFUs) remain a major burden in low-resource services, with wide variation in healing and limb outcomes. Evidence using competing-risks or multistate methods to accurately estimate amputation-free survival (AFS) and time-to-healing, and to identify modifiable risks within routine care remains limited. Purposes: This study aimed to estimate AFS and time-to-healing and to test associations of peripheral arterial disease (PAD), kidney function, infection, glycemia, and off-loading with these endpoints among adults with DFUs. Methods: In a multicenter prospective cohort in Indonesia (October 1, 2022-September 30, 2023), we enrolled 620 adults with active DFUs meeting predefined criteria; exclusions included prior major amputation, non-diabetic ulcers, malignancy, incomplete dates, or day-0 loss to follow-up. Outcomes were AFS (first amputation or death) and time-to-healing (complete epithelialization), each under competing risks (amputation/death or healing). We estimated cumulative incidence with Aalen-Johansen and modeled transitions using cause-specific Cox and Fine-Gray sub-distribution hazards, adjusting for prespecified covariates; multiple imputation addressed missingness, proportional hazards diagnostics were performed, and subgroup/sensitivity analyses were prespecified. Results: Among 620 participants (PAD 41.9%, neuropathy 47.1%, infection 35.6%), the 12-month cumulative incidence was: healed 71%, minor amputation 14%, major amputation 4%, and death 8%. PAD, reduced eGFR (<60), infection, and higher HbA1c worsened AFS (sHR 1.78, 1.55, 1.42, and 1.09 per 1% respectively), while higher HbA1c and infection slowed healing (sHR 0.90 and 0.71); PAD reduced healing (0.79), and off-loading improved healing (1.22). Findings were consistent across sensitivity and subgroup analyses, indicating robustness. Conclusion: Most patients healed within 12 months, yet one in four experienced amputation or death. Results support routine vascular triage or revascularization, infection bundles, glycemic or renal optimization, and mandated off-loading within DFU pathways, and motivate further prospective evaluations of durability, safety, and cost-effectiveness.