A multicenter retrospective analysis of patients treated for tibial fracture was conducted to develop a score that correlates with fracture healing time and, ultimately, to identify the risk gradient of delayed healing. The clinical records of 93 patients treated for tibial fracture at three orthopaedic centers were evaluated. Patients were considered healed when full weight bearing was allowed and no further controls were scheduled. For the purpose of our analysis, we separated patients healed within or after 180 days. Patient's risk factors known to be associated to delay healing, as well as fracture morphology and orthopaedic treatment were recorded in an electronic Case Report Form (e-CRF). Information available in the literature was used to weight the relative risk (RR) associated to each risk factor; values were combined to calculate a score to be correlated to the fracture healing time: L-ARRCO (Literature-Algoritmo Rischio Ritardo Consolidazione Ossea). Among all information collected in e-CRFs, we identified other risk factors, associated to delayed healing, that were used to calculate a new score: ARRCO. Univariate logistic analysis was used to determine a correlation between the score and healing time. Analysis by receiver operating characteristic (ROC) and calculation of the area under the curve (AUC) were used for sensitivity and specificity. Complete information was available for 53 patients. The mean value of the L-ARRCO score among patients healed within 180 days was 5.78 ± 1.59 and 7.05 ± 2.46 among those healed afterwards, p=0.044. The mean value of the ARRCO score of patients healed within 180 days was 5.92 ± 1.78 and 9.03 ± 2.79 among those healed afterwards, p<0.0001. The ROC curve shows an AUC of 0.62±0.09 for L-ARRCO and an AUC of 0.82±0.07 for ARRCO, (p<0.0001). We have shown that the ARRCO score value is significantly correlated to fracture healing time. The score may be used to identify fractures at risk of delayed healing, thus allowing surgeon's early intervention to stimulate osteogenesis.
