Early prediction of tibial and femoral fracture healing: Are we reliable?

Abstract Introduction To evaluate the ability of orthopaedic trauma subspecialists to predict early bony union in femoral and tibia shaft fractures. Materials and methods Eight orthopaedic trauma subspecialists prospectively predicted the probability of bony union at 6 and 12 weeks post-operatively for an aggregate of 48 femoral and tibial shaft fractures treated at a Level 1 trauma centre. An additional orthopaedic trauma subspecialist was blinded to treating surgeon and adjudicated healing at 18 weeks. The Squared-Error Skill Score (SESS) determined the likelihood of accurate forecasting for bony union. Results Nine patients were lost follow-up, resulting in 39 fractures (81.25% retention) including 20 femoral and 19 tibial fractures. Fourteen fractures were open, 15 were not-yet united at final follow-up. SESS values were 0.25–0.77. The ability to predict union (sensitivity) was 1.000. The ability to predict nonunions (specificity) was 0.330–0.500. The probability of a correct predicted union was 0.727 and correct predicted nonunion at final follow-up was 1.000. AO/OTA type A fractures pattern predictions were highly accurate. As body mass index increased, predictions trended toward decreased accuracy (p = 0.06). Tobacco use, age, gender, associated injuries, open fractures, and surgeons’ years in clinical practice were not associated with accuracy of predictions. Conclusions At 12-weeks post-operatively orthopaedic trauma subspecialists can confidently predict the union state in this patient population. This data is most useful in the nonunion patient, directing early intervention, thereby decreasing patient disability and discomfort.