Developing a Risk Score for Predicting Multiple Revision Surgeries in Patients With Fracture-Related Infections.

Introduction: Postoperative infections are a leading cause of morbidity following fracture repair. The purpose of this study is to develop a risk score predicting fracture-related infection (FRI) that will require one versus multiple revision surgeries related to infection eradication and bone healing.

Methods: This is a retrospective cohort study conducted at a single level I trauma center from 2013 to 2020. Adults with FRIs were identified through review of an institutional database on musculoskeletal infections maintained jointly by the infectious disease division and the orthopaedic surgery department. Inclusion criteria were surgically managed fracture of the humerus, olecranon, radius/ulna, clavicle, pelvis, femur, tibia/fibula, and calcaneus with an FRI and adequate documentation present in the electronic medical record. Exclusion criteria included infected chronic osteomyelitis from a non-fracture-related pathology and follow-up less than 6 months. Risk factors leading to multiple surgeries in FRIs, including demographics, comorbidities, injury characteristics, perioperative data, and microbiology, were recorded. Logistic regression was done to select variables predictive of multiple revision surgeries. Four prespecified methods of covariate selection were used.

Results: Eighty-eight patients underwent one FRI revision surgery, whereas 208 patients underwent two or more revision surgeries. From multivariable logistic regression, age older than 45 years (P < 0.001), purulent drainage at infection presentation (P < 0.001), and incomplete bone union at infection presentation (P = 0.013) were all markedly associated with multiple revision surgeries. The model of best fit was used to generate the risk score (area under ROC curve = 0.789). Variables included in the final risk score were age ≥ 45 years, purulent drainage, incomplete bony union, and wound dehiscence at infection presentation.

Conclusion: This study described a risk score for predicting multiple revision surgeries in patients with infection following fracture repair. Age older than 45 years, purulent drainage, and incomplete bony union at infection presentation were all markedly associated with multiple infection revision surgeries.

Level of evidence: III.