Filgrastim in the treatment of infected diabetic foot ulcers - Retrospective cast analysis of a phase II randomised clinical trial

Objective: A retrospective cost-minimisation study of a randomised clinical trial comparing the resource use and treatment cost of hospitalised diabet ic patients with infected foot ulcers with and without filgrastim (5 mu g/k g/day) support for 7 days consecutively. Design: Costs were estimated from the hospital perspective. Resource use da ta were collected from patient record files. Unit cost data were gathered f rom the hospital administration records. A decision tree model was built to estimate the mean cost far each treatment arm. Patients: Forty patients were randomised. At study entry, all had infected ulcers and were clinically investigated regarding their vascular condition. At inclusion no significant difference was observed between cases and cont rols. Results: Clinical results showed that the recovery of the filgrastim-treate d patient was quicker, resulting in an earlier hospital discharge. The inte nt-to-treat cost analysis demonstrated that the mean cost savings were poun d 2666 (36%) in favour of the filgrastim treatment arm. Sensitivity analysi s was performed on patient type, probability distribution, unit cost and ho spital duration. Subgroup analysis identified that the cost savings ranged from pound 3129 (39%) to pound 155 (4%) when patients with vascular problem s and/or tissue necrosis were excluded from the analysis. Conclusion: The overall clinical benefit observed with the use of filgrasti m could be translated into cost savings. These measured cost savings should , however, be interpreted cautiously as patient selection may have occurred that appeared during the in-hospital stay. More patients in the control gr oup experienced a bad vascular condition inducing more costly interventions . The results need therefore to be confirmed in a large phase III randomise d clinical trial.