Introduction:
Ballistic hand fractures (BHF) result from high-velocity, destructive impact, creating complex multi-structure injury patterns which can lead to high non-union rates. We sought to determine factors that may predict these higher complication and non-union rates and assess our management strategy of these fractures.
Methods:
Patients with BHF from 2016-2023 were identified from an institutional database. Charts and radiographs were reviewed for demographics, smoking status, treatment modality (including use of bone grafting), size of bone gap at the fracture, presence of comminution and, if comminuted, fractures were further categorized by presence and quality of remaining bone stock at the fracture site (ie: (1) complete loss of bone stock, with no bone existing between fracture fragments (NC); (2) partial bony contact (PC); or (3) full bony contact (FC), with comminuted fragments remaining within the gap and >~80% contact with a minimum of ninety-day follow up. Our primary endpoint of interest was union at ninety days. Union was determined based on radiographic and clinical findings at follow-up visits. Univariate and multivariate analysis was performed to determine factors associated with nonunion.
Results:
Seventy-seven fractures in 48 patients met inclusion criteria, of which 81% (N=62) were comminuted. Fractures were compared by severity of bone loss across the fracture site. A significant difference existed between groups, as measured by maximum bony gap (FC: 1.7 mm vs. PC: 11.3 mm vs. NC: 33.1 mm, p < 0.001). Non-comminuted fractures had an average gap of 0.63 mm.
The majority of fractures were treated with Kirschner-wire fixation or immobilization alone, followed by open reduction and external fixation (48.1% vs. 31.2% vs 13.0% vs 7.8%). When treatment modality was compared against severity of missing bone stock, there was no statistical difference, regardless of severity of bone loss, in how fractures were managed (p=0.367). No fractures underwent primary bone grafting.
Overall, 22% of fractures achieved union at ninety days (N=17). Of these, 94% (N=16) were either not comminuted or belonged to the FC group. Univariate analysis demonstrated loss of bone stock (NC and PC) was associated with significantly higher rates of non-union compared to FC fractures (90% and 95.2% vs 64.5%, p = 0.018). Additionally smoking and concurrent tendon injury were also associated with non-union (p=0.010, 0.043). On multivariate logistic regression, bone gap (in mm) was an independent predictor of non-union at 90 days (p=0.018).
Conclusions:
BHF lead to high rates of non-union, as they are frequently comminuted and have significant loss of bone stock. Due to many factors like soft tissue loss and inadequacy of bone stock, less invasive methods of treatment are frequently used, such as Kirscher-wires or immobilization alone, which ultimately lead to low union rates. We believe surgeons should consider a planned two-stage surgical reconstruction when presented with BHF, as it may alleviate the overall poor union results. Following initial stabilization and adequate soft tissue coverage, we recommend considering a planned second surgery for conversion to internal fixation, with a low threshold to perform secondary bone grafting, especially in the settings of large bony defects.
