Dose-Response Association Between Cerclage Wire Number and Infection Risk

This dual-center study evaluates the biological impact of "implant burden" in femoral fracture fixation, specifically examining whether the risk of surgical site infection (SSI) is dose-dependent on the number of cerclage constructs used. While cerclage wiring is a recognized adjunct for achieving anatomical reduction in complex subtrochanteric and periprosthetic fractures, the cumulative surface area of multiple braided cables may provide a protected niche for bacterial attachment and biofilm formation.

Surgical Protocol and Intervention:

All procedures were performed or supervised by fellowship-trained orthopedic trauma surgeons using a "biological fixation" strategy to prioritize the preservation of periosteal blood supply. For femoral shaft and subtrochanteric fractures, metallic cerclage wires or cables were applied via small accessory incisions or through direct lateral approaches. Monofilament wires were secured with symmetric twist knots, while multifilament braided cables were tensioned to 40-50 kg and secured with swaged crimp mechanisms. The cohort was stratified into three groups based on the final construct count verified by postoperative radiographs: Group I (1 wire), Group II (2 wires), and Group III (≥3 wires).

Clinical and Laboratory Monitoring:

Patients were monitored preoperatively and postoperatively (Days 1, 3, 7, 14) for inflammatory markers, including C-Reactive Protein (CRP) and Erythrocyte Sedimentation Rate (ESR). Nutritional status was assessed via serum albumin levels on the first postoperative day. Wound complications were tracked, with persistent wound drainage (PWD) defined as drainage lasting more than 5 postoperative days.

Statistical Modeling:

To isolate the independent effect of cerclage count from potential confounders, a penalized (Firth) multivariable logistic regression model was utilized. This approach was selected to reduce small-sample bias and overfitting given the limited number of deep infection events. The core model adjusted for operative duration (continuous and >120 minutes), postoperative hypoalbuminemia (<3.0 g/dL), and fracture category (shaft, subtrochanteric, or periprosthetic). Infection-free survival across the three groups was estimated using the Kaplan-Meier method with log-rank (Mantel-Cox) comparisons.

Inclusion criteria

• Skeletally mature adult patients (≥18 years old).
• Diagnosis of femoral shaft, subtrochanteric, or periprosthetic femoral fractures (Vancouver Types B1, B2, and B3).
• Surgical stabilization using metallic cerclage wires or cables as an adjunct to intramedullary nails or locking plates.
• Minimum clinical and radiological follow-up of 12 months.

Exclusion criteria

• Open fractures (Gustilo-Anderson types I-III).
• Pathological fractures secondary to malignancy.
• History of active infection or septic arthritis in the ipsilateral limb prior to the index procedure.
• Treatment with non-metallic polymer cables.