Robert V. OʼToole

Geriatric Trauma: Demographics, Injuries, and Mortality

Objectives: To identify injuries that elderly sustain during high-energy trauma and determine which are associated with mortality. Design: Retrospective review of prospectively collected database. Setting: Academic trauma center. Patients: Patients selected from database of all trauma admissions from January 2004 through June 2009. Study population consisted of patients directly admitted from scene of injury who sustained high-energy trauma with at least one orthopaedic injury and were 65 years or older (n = 597). Intervention: Review of demographics, trauma markers, injuries, and disposition statuses. Main Outcome Measurements: Statistical analysis using &khgr;2 test, Student t test, and logistic regression analysis. Results: The most common fractures were of the rib, distal radius, pelvic ring, facial bones, proximal humerus, clavicle, ankle, and sacrum. The injuries associated with the highest mortality rates were fractures of the cervical spine with neurological deficit (47%), at the C2 level (44%), and of the proximal femur (25%), pelvic ring (25%), clavicle (24%), and distal humerus (24%). The fractures significantly associated with mortality were fractures of the clavicle (P = 0.001), foot joints (P = 0.001), proximal humerus or shaft and head of the humerus (P = 0.002), sacroiliac joint (P = 0.004), and distal ulna (P = 0.002). Conclusions: Elderly patients present with significantly worse injuries, remain in the hospital longer, require greater use of resources after discharge, and die at 3 times the rate of the younger population. Although the high mortality rates associated with cervical spine, hip, and pelvic ring fractures were not unexpected, the injuries that were statistically associated with mortality were unexpected. Injuries such as clavicle fracture were statistically associated with mortality. As our population ages and becomes more active, the demographic may gain in clinical importance. Level of Evidence: Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.

Are locking screws advantageous with plate fixation of humeral shaft fractures? A biomechanical analysis of synthetic and cadaveric bone.

Objectives: To investigate whether locking screws offer any advantage over nonlocking screws for plate fixation of humeral shaft fractures for weight-bearing applications. Design: Mechanical evaluation of stiffness in torsion, bending, and axial loading and failure in axial loading in synthetic and cadaveric bone. Setting: Biomechanical laboratory in an academic medical center. Methods: We modeled a comminuted midshaft humeral fracture in both synthetic and cadaveric bone. Humeri were plated posteriorly. Two study groups each used identical 10-hole, 3.5-mm locking compression plates that can accept either locking or nonlocking screws. The first group used only nonlocking screws and the second only locking screws. Stiffness testing and failure testing were performed for both the synthetic bones (n = 6) and the cadaveric matched pairs (n = 12). Fatigue testing was set at 90,000 cycles of 440 N of axial loading. Main Outcome Measures: Torsion, bending, and axial stiffness and axial failure force after cyclic loading. Results: With synthetic bones, no significant difference was observed in any of the 4 tested stiffness modes between the plates with locking screws and those with nonlocking screws (anteroposterior, P = 0.51; mediolateral, P = 0.50; axial, P = 0.15; torsional, P = 0.08). With initial failure testing of the constructs in axial loading, both plates failed above anticipated physiologic loads of 440 N (mean failure load for both constructs >4200 N), but no advantage to locking screws was shown. The cadaveric portion of the study also showed no biomechanical advantage of locking screws over nonlocking screws for stiffness of the construct in the 4 tested modes (P > 0.40). Fatigue and failure testing showed that both constructs were able to withstand strenuous fatigue and to fail above anticipated loads (mean failure >3400 N). No difference in failure force was shown between the 2 groups (P = 0.67). Conclusions: Synthetic and cadaveric bone testing showed that locking screws offer no obvious biomechanical benefit in this application.

A biomechanical comparison of a locking plate, a nail, and a 95° angled blade plate for fixation of subtrochanteric femoral fractures.

Objective: To compare the biomechanical performance of a cephalomedullary nail (CMN), a proximal femoral locking plate, and a 95° angled blade plate in a comminuted subtrochanteric fracture model. Methods: A comminuted subtrochanteric femoral fracture model was created with a 2-cm gap below the lesser trochanter in 15 pairs of human cadaveric femora confirmed to be nonosteoporotic. The femora were randomized to treatment with one of the previously mentioned 3 devices. Each was tested under incrementally increasing cyclic load up to 90,000 cycles from 50% to 250% of body weight to simulate progressive weight bearing during 3 months of an average 700-N (approximately, 70 kg or 150 lb) person. Force, number of cycles, and total load sustained to reach 10 mm of displacement were compared. Failure modes were also noted. Results: The CMN construct withstood significantly more cycles, failed at a significantly higher force, and withstood a significantly greater load than either of the plate constructs (P < 0.001). Varus collapse was significantly lower in the CMN construct (P < 0.0001). Modes of failure differed among implant–bone constructs with damage to the femoral head through implant cutout in 5 of 10 blade plate specimens and 2 of 10 CMN specimens, whereas no damage to the femoral head bone was observed in any of the locking plate constructs. Conclusions: The CMN construct was biomechanically superior to either the locking plate or 95° blade plate constructs. The locking plate construct was biomechanically equivalent to the blade plate construct.

Current bacterial speciation and antibiotic resistance in deep infections after operative fixation of fractures.

Objectives: Infection after fracture fixation is a major source of morbidity. Information regarding bacterial speciation and antibiotic resistance is lacking. We attempted to determine the speciation and drug resistance profiles associated with fracture fixation infections. Design: Retrospective study. Setting: Level I trauma center. Patients: Two hundred eleven patients with 214 infections underwent surgery for postoperative infection from December 2006 to December 2010. Deep postoperative infections within 12 months of fixation were included. Intervention: None. Main Outcome Measurements: Incidence of each bacterial species and rate of clinically relevant resistance in Staphylococcus aureus, gram-negative rod (GNR), and Enterococcus species. The effect of timing of infection presentation and location of fracture on bacterial speciation was also investigated. Results: Fifty-six percent of infections had S. aureus present, with 58% of those (32% of all infections) being methicillin-resistant S. aureus. Thirty-two percent of infections had at least one GNR present, with only 4% of those being multidrug resistant. We found a marked increase in the rate of GNR infections of the pelvis, acetabulum, and proximal femur (63%) compared with other locations (27%), which was statistically significant (P = 0.0002). Conclusions: At our center, S. aureus and GNR are most often found in deep postoperative infections after fixation. Methicillin-resistant S. aureus is common in this population. Our GNR rate is high, but resistance in this group was low. The proportion of GNR infections in the pelvis, acetabulum, and proximal femur was high even in closed fractures. These data provide a modern snapshot of orthopaedic infections after fracture fixation and might be useful in designing future studies and protocols for antibiotic prophylactic treatment. We are considering the use of aminoglycosides in the treatment of closed fractures of the pelvis, acetabulum, and proximal femur. Level of Evidence: Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.

Determination of Radiographic Healing: An Assessment of Consistency Using RUST and Modified RUST in Metadiaphyseal Fractures.

Objective: To determine the reliability of the Radiographic Union Scale for Tibia (RUST) score and a new modified RUST score in quantifying healing and to define a value for radiographic union in a large series of metadiaphyseal fractures treated with plates or intramedullary nails. Design: Healing was evaluated using 2 methods: (1) evaluation of interrater agreement in a series of radiographs and (2) analysis of prospectively gathered data from 2 previous large multicenter trials to define thresholds for radiographic union. Intervention: Part 1: 12 orthopedic trauma surgeons evaluated a series of radiographs of 27 distal femur fractures treated with either plate or retrograde nail fixation at various stages of healing in random order using a modified RUST score. For each radiographic set, the reviewer indicated if the fracture was radiographically healed. Part 2: The radiographic results of 2 multicenter randomized trials comparing plate versus nail fixation of 81 distal femur and 46 proximal tibia fractures were reviewed. Orthopaedic surgeons at 24 trauma centers scored radiographs at 3, 6, and 12 months postoperatively using the modified RUST score above. Additionally, investigators indicated if the fracture was healed or not healed. Main Outcome Measures: The intraclass correlation coefficient (ICC) with 95% confidence intervals was determined for each cortex, the standard and modified RUST score, and the assignment of union for part 1 data. The RUST and modified RUST that defined “union” were determined for both parts of the study. Results: ICC: The modified RUST score demonstrated slightly higher ICCs than the standard RUST (0.68 vs. 0.63). Nails had substantial agreement, whereas plates had moderate agreement using both modified and standard RUST (0.74 and 0.67 vs. 0.59 and 0.53). Union: The average standard and modified RUST at union among all fractures was 8.5 and 11.4. Nails had higher standard and modified RUST scores than plates at union. The ICC for union was 0.53 (nails: 0.58; plates: 0.51), which indicates moderate agreement. However, the majority of reviewers assigned union for a standard RUST of 9 and a modified RUST of 11, and >90% considered a score of 10 on the RUST and 13 on the modified RUST united. Conclusions: The ICC for the modified RUST is slightly higher than the standard RUST in metadiaphyseal fractures and had substantial agreement. The ICC for the assessment of union was moderate agreement; however, definite union would be 10 and 13 with over 90% of reviewers assigning union. These are the first data-driven estimates of radiographic union for these scores.

Magnetic resonance imaging for the evaluation of ligamentous injury in the pelvis: A prospective case-controlled study

Objectives: Management of external rotation pelvic ring disruptions is based on which ligaments are disrupted within the pelvis. We hypothesized that magnetic resonance imaging (MRI) can evaluate the ligaments of the pelvic ring and differentiate injured from uninjured pelves. Design: Prospective cohort study. Setting: Level I trauma center. Patients: Twenty-one patients with 25 acute external rotation injuries of the hemipelvis; control group of 26 patients without pelvic ring injury. Intervention: All patients underwent the same MRI protocol reviewed by 1 musculoskeletal radiologist. Main Outcome Measures: Integrity of 5 structures: sacrospinous, sacrotuberous, anterior sacroiliac, and posterior sacroiliac ligaments and pelvic floor musculature. Results: Visualization of sacrospinous, sacrotuberous, anterior sacroiliac, and posterior sacroiliac ligaments, and pelvic floor musculature was possible for 91%, 100%, 98%, 91%, and 100%, respectively, of all studied structures. No injuries were identified in control group patients in contrast to ligament injury observed with all injured pelves (0% versus 100%; P < 0.0001). Observed relationship of ligament injury to pelvic injury type generally agreed with the Young–Burgess classification system, with the important exception that patients with anterior–posterior compression type II injuries had damage to the sacrospinous ligament in only 50% of the cases. Conclusions: Ligamentous anatomy and injury about the pelvic ring appears to be easily evaluated with MRI, arguing that there may be a role for this imaging modality in managing these cases. Tearing of the sacrospinous ligament is variable among anterior–posterior compression type II injuries, arguing that the injury pattern can be subdivided into those with and without sacrospinous ligament tears. Level of Evidence: Diagnostic Level II. See Instructions for Authors for a complete description of levels of evidence.

Locking plates in osteoporosis: a biomechanical cadaveric study of diaphyseal humerus fractures.

Objective: To determine whether locking plates offer an advantage in fixation of fractures in osteoporotic humeral bone. Design: Biomechanical testing of 18 matched pairs of osteoporotic human cadaver humeri plated posteriorly with either all locked or all nonlocked screws. An established protocol was used to test the constructs with torque applied to a peak of ±10 Nm for 1000 cycles at 0.3 Hz or until failure. Eighteen pairs were tested for failure, 11 pairs were tested for cycles survived, and 10 pairs were tested for stiffness. Setting: University biomechanical laboratory. Main Outcome Measurements: Percentage surviving testing, mean cycles survived, and stiffness. Results: We observed catastrophic failure of the constructs in 47% of the samples. Humeri plated with nonlocking plates failed at a higher rate than those with locking plates (67% nonlocking vs 28% locking, n = 18 pairs, P = 0.008). Locking constructs also outperformed nonlocking constructs in mean cycles survived (707 cycles locking, 345 cycles nonlocking, n = 11 pairs, P < 0.05) and stiffness at 10 cycles (0.853 Nm/degree locking vs 0.416 Nm/degree nonlocking, n = 10 pairs, P < 0.001). Conclusions: Locking plates were shown to provide improved mechanical performance over nonlocking plates in torsional cyclic loading in a osteoporotic cadaveric fracture model. Our results confirm general conclusions of previous work that used a synthetic bone model of osteoporosis, but we found a high rate of catastrophic failure, questioning the validity of the previously published synthetic model of osteoporosis (overdrilling of synthetic bone) for this application.

Does Fracture Care Make Money for the Hospital? An Analysis of Hospital Revenues and Costs for Treatment of Common Fractures.

Objectives: To determine the relative profitability for a hospital of treatment of common fractures within a state-regulated reimbursement system. Design: Retrospective cohort. Setting: Regional trauma referral center with state-regulated hospital reimbursement system. Methods: We reviewed hospital medical and financial records of 1228 patients admitted from 2008 through 2012 with a principle diagnosis of acute traumatic fracture requiring surgical treatment. Patients whose principle diagnosis fit into 1 of 6 common anatomic categories were included. Sixty-five pelvic, 275 acetabular, 277 hip, 255 femoral shaft, 148 tibial shaft, and 208 ankle fractures were identified. Patients with a different principle diagnosis were excluded. Net revenue, total cost of inpatient care, and direct margin for each patients acute inpatient hospital course were recorded. Main Outcome Measurement: Direct margins, costs. Results: Per patient, the overall mean net revenue was

Embolization of pelvic arterial injury is a risk factor for deep infection after acetabular fracture surgery.

39,813, overall mean cost of inpatient care was

Radiographic Predictors of Compartment Syndrome Occurring After Tibial Fracture.

21,231, and overall mean direct margin (profitability) was