Purpose: Since 1996, our multidisciplinary medicosurgical team has decided to propose orthopaedic treatment for diabetic perforating ulcers with osteitis, cellulitis, or necrosis (“cooling down” the acute infected ulcers before programmed surgery) rather than conservative treatment with prolonged antibiotic therapy. We present here a prospective study of 44 cases of diabetic perforating ulcers.
Material and methods: Thirty-two diabetic patients underwent surgery: 77% males, mean age 65.2±8.6 year (range 43–86 years), 87% type 2 diabetes, 52% with a history of perforating ulcers, 45% with minor amputations, and 14% with history of vascular surgery. The lesions—perforating ulcer with osteitis (n=34), vascular necrosis of the toes (n=2), “acute feet” with cellulitis (n=8)—had progressed over 13.2±15.1 weeks. The preoperative work-up included: bacteriology samples 89%; standard x-rays of the foot 100% (osteitis 84%); duplex Doppler of the lower limb arteries 77% (tibial arteriopathy 87%); double bone scintigraphy 34% (osteitis 93%); TcPO2 (40±14mmHg); arteriography 27%; vascular surgery consultation 18%. Before surgery, 77% of the patients were hospitalised in an endocrinology unit (13±3 days) and 88% were on an antibiotic regimen for 26±18 days (50% i.v.).
Orthopaedic surgery (without tourniquet, anaesthesia block, mean duration 53±24 min) involved: partial resection of a toe 23%; amputation of a ray 36% (first ray one, second ray five, third ray one, fourth ray two, fifth ray six); transmetatarsal amputation 32%; resection of the metatarsal heads 4%; calcanectomy (n=1); below knee amputation (n=1); and systematic and multiple samples for bacteriology (deep soft tissue and bone tissue) and for pathology.
Results: Mean hospital stay in the surgery unit was 4±1 days, followed by 18±10 days in the endocrinology unit with antibiotics (oral for 88%) for 34±22 days, 91% of the lesions healed within 33±18 days; four required repeated procedures (two transmetatarsal amputations, one amputation of the first ray, one lower limb amputation); three lesions relapsed.
The peroperative bacteriology samples of the deep soft tissue and bone tissue demonstrated, in comparison with the preoperative samples, that antibiotics had sterilised only 14% of the lesions; with discordant comparison in 40%, partial concordance in 24%, and total concordance in 24%. For the diagnosis of osteitis (confirmed by histology of peroperative bone samples), the x-ray interpretations were largely confirmed (79% exact diagnosis, 87% sensitivity, false positives 12%), as were the bone scintigrams with labelled polymorphonuclears (exact diagnosis 93%, sensitivity 93%, false positives 7%).
Conclusion: This prospective study demonstrated the advantages of programmed surgery over emergency surgery, including for “acute feet”: limited resection, primary suture, rapid wound healing, short antibiotic treatment. It raises some questions concerning the validity of non-surgical bacteriological samples for perforating ulcers, even when performed under rigorous conditions (unique strain isolated from 76% of the samples) and on the possibility of antibiotic pressure on bacterial selection.
