Faculty Bibliography

OBJECTIVE:: To 1) determine the incidence and injury profile of open periarticular fractures about the knee joint in a cohort of patients presenting to the emergency department with a deep periarticular knee wound and to 2) determine the effectiveness of CT scan to detect and guide management of these open fractures compared to plain radiographs. DESIGN:: Retrospective Review SETTING:: Level I trauma center PATIENTS/PARTICIPANTS:: 78 patients (79 knees) with deep periarticular knee wounds of which 62 patients (63 knees) received both a plain radiograph and a CT scan of the knee. INTERVENTION:: Plain radiograph and CT scan of the injured knee. MAIN OUTCOME MEASUREMENTS:: Comparison of OTA fracture classification and surgeon produced management plan as determined by plain radiographs versus CT scans. CT scan was considered the gold-standard test to detect a fracture. RESULTS:: Twenty-one (27%, 21/79) knees had an open periarticular fracture of the knee and 95% (20/21) of these knees had intra-articular air indicative of an associated traumatic arthrotomy. Of 41 (52%, 41/79) knees with a traumatic arthrotomy, 51% (21/41) had an associated open periarticular fracture of the knee. Plain radiographs detected 18 fractures in 17 knees (of which 1 fracture was later determined to not be a true fracture) whereas CT scan detected 26 fractures in 21 knees. Overall, CT scan detected 9 additional fractures in 6 knees. The specificity and sensitivity and positive predictive and negative predictive values of plain radiographs to detect and rule-out a fracture was 98%/65% and 94%/82%, respectively. Compared to plain radiographs, CT scan altered the fracture classification in 48% of patients and altered the management plan in 43% of patients, respectively. Gunshot wounds to the knee had a 48% (12/25) incidence of an associated open periarticular fracture compared to an 17% (9/54) incidence for all other injury mechanisms combined (p<.01). CONCLUSION:: Patients with a periarticular knee wound have a high incidence of open periarticular fractures, and the incidence is even higher if the mechanism of injury is a GSW or there is associated traumatic arthrotomy. CT scan improves detection and management of open fractures of the knee compared to plain radiographs. Consideration should be given to routinely using CT scan to evaluate knees with deep periarticular wounds that present to the ED, especially if secondary to gunshot injuries, given the high incidence open periarticular fractures of the knee. LEVEL OF EVIDENCE:: Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.

OBJECTIVE:: To report our experience with computed tomography (CT) scans to detect traumatic arthrotomies of the knee joint (TAK) based upon the presence of intra-articular air. DESIGN:: Retrospective review SETTING:: Level I trauma center PATIENTS/PARTICIPANTS:: Sixty-two consecutive patients (63 knees) underwent a CT scan of the knee in the emergency department and had a minimum of 14 days follow-up. Cohort of 37 patients (37 knees) from the original 62 patients who underwent a Saline Load Test (SLT) INTERVENTION:: CT scan and SLT MAIN OUTCOME MEASUREMENTS:: Positive traumatic arthrotomy of the knee (+TAK) was defined as operating room (OR) confirmation of an arthrotomy or no intra-articular air on CT scan (-iaCT) (and -SLT if performed) with follow-up revealing a septic knee. Periarticular wound equivalent to no traumatic arthrotomy (pw=(-TAK)) was defined as OR evaluation revealing no arthrotomy or -iaCT (and -SLT if performed) with follow-up revealing no septic knee. RESULTS:: All 32 knees with intra-articular air on CT scan (+iaCT) had OR confirmation of a TAK and none of these patients had a knee infection at a mean follow-up of 140.0+/-279.6 days. None of the 31 patients with -iaCT had a knee infection at a mean follow-up of 291.0+/-548.1 days. Based on these results, the sensitivity and specificity of the CT scan to detect +TAK and pw=(-TAK) was 100%. In a subgroup of 37 patients that received both a CT scan and the conventional SLT, the sensitivity and specificity of the CT scan was 100% compared to 92% for the SLT (p<0.001). CONCLUSION:: CT scan performs better than the conventional SLT to detect traumatic knee arthrotomies and identify periarticular knee wounds that do not require surgical intervention and should be considered a valid diagnostic test in the appropriate clinical setting. LEVEL OF EVIDENCE:: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

OBJECTIVES:: To determine the lowest estimated radiation dose necessary to reproducibly detect intra-articular air in the knee joint of a cadaver model. METHODS:: 10 adult fresh-frozen cadaver knees with intact joint capsules provided by Science Care(R), Phoenix, AZ. were thawed and scanned at 5 decreasing radiation doses (decreasing by approximately half from 8.42 mGy to 0.74 mGy) after introducing increasing volumes (0 cc, 0.1 cc, 0.3 cc, 0.5 cc, 0.7 cc, 0.9cc) of intra-articular air. Scans were performed using 2.0 mm slice-thickness from the distal 1/3 of the femur to the proximal 1/3 of the tibia. Sagittal and coronal reconstructions of each scan using 1.0 mm slice-thickness were rendered. All scans were reviewed by 1) a single attending radiologist, 2) a single attending orthopedic surgeon, and 3) a single chief resident, for the presence of intra-articular air. RESULTS:: The sensitivity and specificity of CT scan to detect intra-articular air at each volume of intra-articular air (0.1 cc, 0.3 cc, 0.5 cc, 0.7 cc, 0.9cc) was 100% at 0.74 mGy - the radiation threshold dose (RadTH) (scan parameters: voltage 80kV, current: 33mA, and scan time: 12.17 sec). The effective radiation dose at 0.74 mGy for a CT scan of the knee is approximately 0.10 mSV CONCLUSIONS:: CT scan to detect traumatic knee arthrotomies can be successfully accomplished at a threshold radiation dose of 0.74 mGy and for an intra-articular volume of 0.1cc of air. This low radiation dose protocol and volume of intra-articular air should be taken into consideration with future studies evaluating the use of CT scan to detect traumatic arthrotomies.

OBJECTIVE:: To describe the use of the Saline Load Test (SLT) utilizing a new definition that more adequately characterizes its use in the emergency department (ED) setting. DESIGN:: Retrospective review SETTING:: Level I trauma center PATIENTS/PARTICIPANTS:: Fifty consecutive patients who underwent a SLT of the knee in the emergency department and had a minimum of 14 days follow-up. INTERVENTION:: Saline Load Test MAIN OUTCOME MEASUREMENTS:: Positive traumatic arthrotomy of the knee (+TAK) defined as OR confirmation of an arthrotomy (assumed to develop a septic knee) or -SLT with follow-up revealing a septic knee. Periarticular wound equivalent to no traumatic arthrotomy of the knee (pw=(-TAK)) defined as OR evaluation revealing no arthrotomy (assumed not to develop a septic knee) or -SLT whose follow-up revealed no septic knee. Development of a septic knee was considered the gold-standard for determining true positives/negatives and false positives/negatives. RESULTS:: The mean wound size was 3.9 +/- 4.3 cm and the mean saline load volume was 74.9 +/- 28.2 cm. There were 19 +SLTs of which there were 16 +TAK and 3 pw=(-TAK). The 3 pw=(-TAK) in the +SLT group were evaluated in the OR where inspection of the joint capsule revealed the absence of a traumatic arthrotomy. There were 31 -SLTs of which there were 1 +TAK and 30 pw=(-TAK). The SLT has a sensitivity of 94% and a specificity of 91% for detecting +TAKs and ruling-out periarticular wounds not requiring surgical intervention (pw=(-TAK)). The false-positive rate of the SLT to detect +TAK is 9%. CONCLUSION:: Using +TAK and pw=(-TAK) as the newly defined measures of the SLT, we report the sensitivity (94%) and specificity (91%) of the SLT in the ED setting while still maintaining the clinical relevancy of the test. Based on a small sample size, knees with small periarticular wounds and a -SLT and no other radiographic or clinical evidence of an arthrotomy appear to have an infection rate of 0% with non-operative management. LEVEL OF EVIDENCE:: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

INTRODUCTION: In the prevention of venous thrombo-embolic events (VTEs) in isolated low-energy fracture patients, management guidelines are conflicting and prior literature is lacking. We aimed to determine the incidence and factors associated with the development of symptomatic VTE in this patient cohort. MATERIALS AND METHODS: To identify patients with isolated, low-energy fractures, we studied billing records from all admissions to our tertiary care orthopaedic hospital from 2007 to 2009. We used International Classification of Diseases, 9th Revision codes to identify patients who developed deep vein thrombosis (DVT) and/or pulmonary embolism (PE) during their hospital admission or within 90 days of discharge. We also collected data on socio-demographics, type of injury, fracture treatment, co-morbidities and anticoagulation therapy at time of admission. This study was a retrospective review of a database. RESULTS: In total, 1701 admissions fit our criteria. Average patient age was 64.27 years and 64.4% were female. There were 479 (28.2%) upper extremity fractures and 1222 (71.8%) lower extremity fractures. Incidence of clinically significant VTE was 1.4%. Of the 24 patients with 25 documented VTE, there were 13 DVTs and 12 PEs, including 2 fatal PEs (0.012%). Nineteen VTEs occurred in association with lower extremity fractures and six with upper extremity fractures; 74% of patients were chemoprophylaxed. Patients with VTE had an average age of 69.5 years and an average body mass index (BMI) of 28kgm(-2). Logistical regression analysis found female sex (p=0.05) and elevated BMI (p=0.003) to be the only significant predictors of VTE. CONCLUSIONS: Clinically significant VTE among patients who sustained isolated, low-energy fractures was found to be low in the setting of standard VTE prophylaxis. Our incidence was consistent with that of patients undergoing total hip arthroplasty. Female sex and increased BMI were statistically significant predictors of VTE.

Atypical femoral fractures, deformities of the subtrochanteric region of the femur identified with plain anteroposterior or lateral lower extremity radiographs and characterized by a specific fracture pattern, are uncommon manifestations in osteoporotic patients. However, the high prevalence of these fractures in patients receiving long-term bisphosphonate therapy led to the many investigations of this association. The purpose of this article is to evaluate and address the link between this fracture type with long-term bisphosphonate therapy, outline the clinical scenario and better define treatment options for optimal care and recovery. In order to do this, a PubMed search was carried out for significant articles using the following keywords: 'alendronate', 'fracture', 'atypical' and 'femur'. 2013 Future Medicine Ltd

OBJECTIVE: : To describe the outcomes for patients treated at a single institution, who sustained incomplete bisphosphonate-induced femoral fractures. DESIGN: : Retrospective review. SETTING: : University-based academic medical center. PATIENTS: : Thirty-one patients with 43 incomplete fractures met the inclusion criteria. INTERVENTION: : Nonoperative management or surgical intervention for fractures with refractory symptoms or progression of fracture lucency on radiographs. MAIN OUTCOME MEASUREMENTS: : Radiographic assessments and the Short Musculoskeletal Functional Assessment to gauge functional status. RESULTS: : The cohort was all women with an average age of 69.2 (range: 46-92) years and had been treated with bisphosphonate therapy for an average of 9.1 (range: 5-20) years. The average healing time for all incomplete fractures was 9.4 (range: 1.5-36) months. Forty-nine percent of the fractures (21 of 43 fractures) were ultimately treated with surgery for impending complete fracture or failure of nonsurgical management. Of the incomplete fractures treated with surgery, 81% became pain free and 100% were radiographically healed at a mean of 7.1 (range: 1.5-12) months. In contrast, of the nonoperatively treated incomplete fractures, only 64% were pain free at latest follow-up, with only 18% of fractures demonstrating radiographic evidence of healing at an average of 11 (range: 6-24) months. Standardized dysfunction index from the Short Musculoskeletal Functional Assessment was better (19.7) in the surgical group than in the nonsurgical group (19.7 vs. 25.7, P = 0.0017). CONCLUSIONS: : A higher percentage of patients treated surgically became asymptomatic and demonstrated radiographic evidence of healing earlier than those treated nonsurgically. Surgical intervention is effective for relief of symptoms when treating incomplete bisphosphonate-related femur fractures, and patients should be counseled to the potential benefits of prophylactic surgery. LEVEL OF EVIDENCE: : Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

OBJECTIVES: : A unique complication of using full-length intramedullary nails for pertrochanteric and subtrochanteric femur fractures is nail penetration of the anterior cortex at the distal end of the femur because of a mismatch of the anatomic femoral bow with that of currently available cephalomedullary nails (CMNs). This study was performed to determine the rate of distal femoral cortical penetration after stabilization of the proximal femur in a consecutive series of patients using a long CMN with a curvature of 180 cm and to examine the final location of the nails within the femoral canal. DESIGN: : Retrospective chart review. SETTING: : Level 1 trauma center with tertiary care. PATIENTS/PARTICIPANTS: : Between June 2005 and September 2008, 271 consecutive proximal femoral fractures or impending fractures were stabilized using a specially designed CMN [pertrochanteric nail (PTN) system; Biomet, Warsaw, IN]. Fifty-seven fractures were excluded because of inadequate lateral radiographs, leaving 214 nails in 212 patients available for analysis. INTERVENTION: : The proximal femur fractures or impending fractures were stabilized with a full-length CMN with a shaft diameter of 11 mm and a radius of curvature of 180 cm. MAIN OUTCOME MEASUREMENTS: : Nail position was determined from the lateral radiograph, which included the minimal distance from the nail to the anterior cortex of the distal femur and the relative position of the nail within the medullary canal, at a level within 2 cm proximal to the superior pole of the patella. Based on the nail position within the distal femoral canal, the following were calculated: (1) the overall rate of distal femoral anterior cortical penetration, (2) the distribution of nail distances to the anterior cortex, and (3) the proportion of nails lying in each quarter of the space available for the nail. RESULTS: : Four surgeries (1.9%) were performed for pathologic fracture and 22 (10.3%) for impending pathologic fracture. The remaining 188 CMNs were used to stabilize OTA Type 31A1 fractures (52 nails, 24.3%), Type 31A2 fractures (62 nails, 29%), Type 31A3 fractures (15 nails, 6.5%), and Type 32 fractures (59 nails, 27.6%). Of the 214 cases available for radiographic analysis, there was 1 case (0.47%) of distal femoral anterior cortical penetration. Of the remaining 213 CMNs, 40% of nails ended up far anterior, 48% anterior, 10% posterior, and 2% far posterior. Sixteen percent (1/6) were within 3 mm of the anterior cortex and half were within 7 mm. The average distance from nail to the anterior cortex was 8.5 mm. CONCLUSION: : Use of a CMN with a radius of curvature of 180 cm to stabilize the proximal femur resulted in a very low rate of distal femoral anterior cortical penetration. However, most of the intramedullary nails ended up in the anterior half of the space available for the nail with 16% within 3 mm of the anterior cortex. These results highlight the importance of being aware of the complication of anterior cortical impingement and perforation during CMN insertion. LEVEL OF EVIDENCE: : Therapeutic Level IV. See Instructions for Authors for a complete description of levels of evidence.

A total of 96 displaced tibial plateau fractures in 94 patients (average age, 48 years) were treated with open reduction and internal fixation. At 12 months postoperatively, mean range of knee motion was 126 degrees and 10 (10%) of the patients had required a secondary surgery. Using a binary regression model, no demographic variable such as age, gender, smoking history, diabetes, BMI; or fracture characteristic such as mechanism of injury, initial plateau depression, Schatzker classification was identified that correlated with short-term functional outcome such as a complication, range of motion, healing time, or residual depression. We determined that radiographic fracture reduction was not superior following fractures with less initial displacement, and increased plateau collapse at 12 months postoperatively was not indicative of decreased function. However, an increased postoperative articular step-off was found to be associated with an increased risk of plateau collapse (p < 0.01). Furthermore, at 12 months, 76% of the patients had returned to their preinjury employment and the overall complication rate was 13%.

The posteromedial fragment in tibial plateau fractures is considered unstable and requires specific fixation. However, if not loaded by the femur, it may remain stable and not require additional fixation. Our purpose was to determine the size of the posteromedial fragment that would remain unloaded by the femoral-tibial contact area, as a function of fracture line orientation and knee flexion angle. Seven human cadaveric knees with intact capsule and ligaments were mounted in a mechanical rig and flexed from 0 to 30, 90, 105, and 120 degrees of flexion. The fiducial points and articular surfaces were digitized, and 3-dimensional software models of the knees at each flexion angle were created. The femoral-tibial contact areas were determined using the software under high- and low-load conditions. Posteromedial fragments of various sizes and fracture line orientations relative to the posterior femoral condylar axis (PFCA) were modeled, and their locations relative to contact areas were determined. The size of unloaded fragments decreased with increased flexion angle. Fragments occupying 60% of the medial plateau were loaded at all angles, but fragments with 30% of the plateau became loaded at 90 degrees under high load and 120 degrees under low load. Fracture line orientations of 0 to 20 degrees external rotation relative to PFCA allowed for the largest fragments to remain unloaded. The size of posteromedial tibial plateau fracture fragment that remains unloaded by the femur varies with knee flexion angle and fracture line orientation. This may have implications for the management of posteromedial tibial plateau fractures.