Predictive factors for the development of Gartland type IV supracondylar humerus fractures: a prospective clinical study.

Objective: This study aimed to identify the preoperative predictive factors for the development of Gartland type IV supracondylar humerus fracture based on the patient characteristic, fracture mechanism, and preoperative radiographic fracture characteristics.

Methods: This prospective study included the data of 120 patients with Gartland type III and IV supracondylar humerus fractures treated in a single center from 2020 to 2021. Patients' age, gender, height/weight percentile values, injury mechanisms, the proximity of fracture fragment to the skin (i.e., dimple sign), and time from trauma to surgical treatment were recorded. In the preoperative radiographs, the degree of extension or flexion deformity between fracture fragments in the sagittal plane, varus/valgus angulation between fracture fragments in the coronal plane, the amount of translation (medial or lateral) in the coronal plane, and the amount of osseous apposition between fracture fragments in the coronal plane were evaluated. With the authors' consensus, the patients were divided into 2 groups based on the presence of multidirectional instability during the intraoperative reduction: group 1 (Gartland type III; 99 patients) and group 2 (Gartland type IV; 21 patients). Fixation of the fractures was then completed.

Results: Significant differences were observed between groups in the valgus/varus angle and amount of osseous apposition (P < .001). Although no significant difference was found in terms of translation amount between the groups (P=.088), there was a significant correlation with medial translation in type IV fractures (P < .001). The correlation between the results and the groups was checked with Spearman's test. Medial translation (r=0.352), varus or valgus angulation (r=0.616), and osseous apposition (r=0.433) exhibited a positive correlation. The probability of type IV fracture was modeled for the preoperative parameters using binary logistic regression. The regression analysis showed that the diagnosis of type IV supracondylar fractures could be predicted, if varus or valgus angulation was more than 25.5° (81% sensitivity, 85% specificity, odds ratio=1.725; 95% CI=1.170-2.541, P=.001, r=0.616) and if the amount of osseous apposition was more than 9.5 mm (85% sensitivity, 81% specificity, odds ratio=1.471; 95% CI=0.714-3.029, P=.001, r=0.433) in the preoperative radiographs. There was also a significant correlation between medial translation (varus angulation) (P < .001, r=0.352), age (P=.019, r=0.255), and patients with more than 90 height/weight percentile values (P < .001, r=0.508) with the possibility to have Gartland type IV fractures.

Conclusion: This study has found some preoperative factors that may be relevant for type IV Gartland fractures. Height/weight values greater than the 90 percentile, varus or valgus angulation greater than 25.5°, bone apposition values greater than 9.5mm, medial translation values greater than 11mm, and older than eight years patients type IV fractures were more common in such patients. If surgeons can more accurately diagnose a Gartland type IV fracture preoperatively, the surgeon can more accurately inform the patient and plan better treatment.

Level of evidence: Level II, Diagnostic Study.