Waist-Hip Ratio Surrogate Is More Predictive Than Body Mass Index of Wound Complications After Pelvic and Acetabulum Surgery.

OBJECTIVES

To determine whether a novel surrogate of waist-hip ratio (WHR) is more predictive of wound complications after pelvis or acetabulum stabilization than body mass index (BMI) and describe the method of measuring a WHR proxy (WHRp).

DESIGN

Retrospective review.

SETTING

One Level 1 Trauma Center.

PATIENTS: One hundred sixty-one patients after operative repair of pelvis and acetabulum fractures.

INTERVENTION: Operative stabilization of a pelvic ring injury or acetabular fracture.

MAIN OUTCOME MEASUREMENTS: Infection (pin, superficial, and deep) and wound healing complication.

METHODS

We retrospectively reviewed 161 subjects after operative repair of pelvic and acetabular fractures. Primary outcome was any wound complication. BMI was acquired from medical records. WHRp was derived from anteroposterior and lateral computed tomography scout images. BMI and WHRp results were analyzed as continuous and categorical variables. BMI was grouped into high-risk categories of ≥30 and ≥40. WHRp data were grouped utilizing the WHO's high-risk profile for females (>0.85) and males (>0.90). An alternative optimal WHR was also assessed. Covariate analysis included demographic data, Injury Severity Score, mechanism, tobacco use, presence and types of open approach, injury type, associated injuries and comorbidities, failure of fixation, and thromboembolism.

RESULTS

The mean follow-up was 15.9 months. Twenty-four (15%) patients developed wound complications. Increasing BMI (P < 0.007) and WHRp (P < 0.001) as continuous variables and female sex (P < 0.009) were associated with wound complications. Applying unadjusted continuous data to a receiver operating characteristic curve revealed a greater area under the curve for WHRp than for BMI (P < 0.001). The optimal predictive WHRp was ≥1.0 (P < 0.001, odds ratio 43.11). The receiver operating characteristic curve from adjusted data demonstrated a greater area under the curve for WHRp ≥1.0 (0.93) compared with BMI ≥30 (0.78) or ≥40 (0.75) and WHO WHRp (0.82). Computed tomography generated WHRp demonstrated excellent interrater reliability (0.99).

CONCLUSION

The WHRp of ≥1.0 was more predictive than BMI of wound complications after operative treatment of pelvis and acetabulum fractures. In our series, WHRp calculated using scout images performed sufficiently well to predict wound complications.

LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.