ORIGINAL ARTICLE Magnetic Resonance Imaging for the Evaluation of ous Injury in the Pelvis: A Prospective Case-Controlled Study Joshua L. Gary, MD,* Michael Mulligan, MD, Kelley Banagan, MD,* Marcus F. Sciadini, MD,* Jason W. Nascone, MD,* and Robert V. O Toole, MD* 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: ous anatomy and injury about the pelvic ring appears to be easily evaluated with MRI, arguing that there may Accepted for publication May 1, 2013. From the *R Adams Cowley Shock Trauma Center, Department of Orthopaedics, University of Maryland School of Medicine, Baltimore, MD; and Department of Radiology and Nuclear Medicine, University of Maryland Medical Center, Baltimore, MD. The authors report no conflicts of interest. The medical center provided funding for the magnetic resonance imaging, and no patient was charged for these studies. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions this article on the journal s Web site (www.jorthotrauma.com). Reprints: Robert V. O Toole, MD, Department of Orthopaedics, R Adams Cowley Shock Trauma Center, University of Maryland School of Medicine, 22 S. Greene St, T3R62, Baltimore, MD 21201 (e-mail: rvo3@yahoo.com). Copyright 2013 by Lippincott Williams & Wilkins 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. Key Words: ligamentous injury, pelvis, pelvic ring, magnetic resonance imaging, sacrospinous ligament Level of Evidence: Diagnostic Level II. See Instructions for Authors for a complete description of levels of evidence. (J Orthop Trauma 2014;28:41 47) INTRODUCTION External rotation injuries of the pelvic ring are currently managed based on which pelvic ligaments are disrupted. The Young Burgess classification system classifies these external rotation injuries into anterior posterior compression (APC) type I, II, and III injuries. The distinction among the 3 types is based on which ligaments are injured, and the treatment varies from nonoperative to internal fixation of both the anterior and posterior portions of the ring based on the injury type. Although it is widely agreed that an injury to the ligaments drives the management, 1 5 the status of the ligaments is inferred from displacements of bony landmarks shown on plain radiographs where the ligaments cannot be directly observed. 4 Magnetic resonance imaging (MRI) provides a noninvasive method of evaluating ligamentous anatomy and has been in common clinical use for decades in determining injury in other areas of the body. 6 Despite this common use in other areas of orthopaedics, no studies to date have described the use of MRI to evaluate ligamentous integrity in patients with acute pelvic ring disruptions. We sought to know whether MRI can detect the presence and integrity of ligaments of the pelvic ring. Our primary hypothesis was that MRI would be able to visualize the ligamentous structures hypothesized to play a role in the stability of the pelvis. Our secondary hypothesis was that MRI would be able to differentiate between injured and uninjured pelves. PATIENTS AND METHODS Patients After receiving approval from our institutional review board, we obtained consent from and initially prospectively J Orthop Trauma Volume 28, Number 1, January 2014 www.jorthotrauma.com 41
Gary et al J Orthop Trauma Volume 28, Number 1, January 2014 enrolled a series of 3 patients as a pilot study. After expanded approval from our institutional review board, we added a consecutive series of 18 patients treated at our urban level I trauma center. Considering that ligamentous integrity plays a vital role in external rotation injuries, we included all patients with the Young Burgess 2 APC type I, II, or III injuries (OTA 61-A, B, or C injuries). 7 Additionally, we included patients with lateral compression (LC) type III injuries (OTA 61-B3.2 injuries) 7 because LC type III injuries are a combination of APC on one side and LC on the other. To qualify for inclusion in the study, patients had to be older than 18 years at the time of injury and able to undergo MRI within 14 days of injury. Exclusion criteria for the injury group were failure to obtain informed consent, presence of an acetabular fracture, pregnancy, and inability to undergo MRI based on clinical instability or other contraindication. The study group consisted of 21 patients with 25 acute external rotation injuries of the hemipelvis that met the inclusion and exclusion criteria. Patients in the pilot study group were prospectively enrolled from May 2008 to June 2008. The remaining 18 patients were enrolled from January 2010 to July 2011. The study included 16 male and 5 female patients, and the average patient age was 40 years (range, 19 64 years). Informed consent was obtained from each of the injured patients or from a legally authorized representative when the patient was not able to consent. The control group consisted of uninjured patients who underwent MRI of the pelvis at our institution between August 25, 2010, and September 24, 2010, for reasons unrelated to pelvic trauma. Inclusion criteria were age of at least 18 years and available magnetic resonance images that included the entire pelvis. The control group consisted of 26 consecutive patients (52 hemipelves) who met the inclusion criteria, 18 of whom were female patients. The average age of the patients in the control group was 48 years (range, 21 69 years). Pelvic Ring Injury Classification Plain radiography was used to classify the pelvic ring injuries, as is typically done in clinical practice. Anteroposterior view, inlet, and outlet radiographs of the pelvis were obtained of the patients in the injury group, as is the standard clinical practice. In addition, computed tomography with coronal and sagittal reconstructions was performed. The control group of uninjured patients had no conventional radiographs of the pelvis. Anteroposterior views of the pelvis without a binder were available for all except 2 patients. Eight patients did not have inlet and outlet radiographs without a binder obtained before surgical intervention. An orthopaedic trauma fellow and attending orthopaedic traumatologist independently classified each injury according to the criteria presented by Burgess et al 2 and the OTA classification 7 with a review of conventional films, computed tomography, and results of stress examination with the patient under anesthesia, if performed. If the fellow and traumatologist disagreed regarding a classification, a second attending traumatologist evaluated the films, scans, and stress examination results, if available, to determine the injury pattern; such disagreement occurred in only 3 cases. Figure 1 shows an anteroposterior view of a pelvis with an injury classified as a left APC type II injury. FIGURE 1. Anteroposterior view radiograph of a Young Burgess APC type II injury. An LC type III injury (OTA 61-B3.2 injury) is defined as a combination of an APC injury on one side and an LC injury on the other. For this study, the side that had an APC injury was classified and included in the analysis and the side that had an LC injury was not included. Injury patterns were classified as 3 APC type I (OTA 61-A), 19 APC type II (OTA 61-B), and 3 APC type III (OTA 61-C) patterns. Four patients had bilateral injuries; each injury was considered individually for the purposes of this study. Clinical Treatment and MRI Protocol For the injured patients, the attending trauma surgeon and orthopaedic traumatologist on call determined clinical treatment without information provided by MRI. Patients underwent MRI only once they were deemed hemodynamically stable by the general surgery trauma service. MRI was performed either preoperatively or postoperatively, depending on patient stability and machine availability. The average time from injury to MRI was 3 days (range, 0 11 days). Twelve of 21 patients underwent preoperative MRI, and 9 underwent MRI after surgical intervention. MRI was performed with a Siemens Magnetom Avanto 1.5-Tesla magnet (Siemens Medical Solutions, Erlangen, Germany). Two sequences were used to evaluate the patients with pelvic ring disruptions: an axial T1-weighted sequence and an axial 3D volume sequence with T2 weighting. Parameters for the T1-weighted sequence were repetition time of 500 to 700 milliseconds, time to echo of 10 to 15 milliseconds, echo train length of 2, and section thickness of 5 mm with a 1-mm section gap. For the T2-weighted sequence, parameters were repetition time of 1500 milliseconds, time to echo of 110 to 120 milliseconds, flip angle of 150 degrees, echo train length of 135, and section thickness of 1 mm. These sequences were selected by our musculoskeletal radiologists for the specific purposes of this study, and duration of the MRI process generally lasted 30 minutes. 42 www.jorthotrauma.com Ó 2013 Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 1, January 2014 MRI for Evaluation of ous Injury FIGURE 2. Axial T2-weighted magnetic resonance images of the sacrospinous (black arrows) and sacrotuberous (white arrows) ligaments, from proximal to distal. A, The sacrotuberous ligaments near their insertion into the ischium. B, The sacrospinous ligaments pass anterior to the sacrotuberous ligaments at the level of the ischial spines. C, The sacrotuberous ligaments near their origin at the sacrum. An attending musculoskeletal radiologist with.20 years of experience evaluated the images of the injury and control groups. For the injury group, both sides of the pelvis were evaluated for injuries to the above ligaments. For all included hemipelves, 5 structures were evaluated with MRI: the sacrospinous, sacrotuberous, anterior sacroiliac, and posterior sacroiliac ligaments and the pelvic floor muscles. Visualization on magnetic resonance images was defined as complete if the structure could be seen on all sections, partial if seen on only some of the sections, and poor if not seen. Figure 2 shows magnetic resonance images of the sacrospinous and sacrotuberous ligaments in a control group patient. integrity was classified as torn (complete disruption of all fibers), partially torn (some fibers torn, some intact), or normal (all fibers intact). Figures 3 through 5 show magnetic resonance images of ruptured anterior sacroiliac, sacrospinous, and sacrotuberous ligaments in injured patients. Statistical Analysis All continuous variables were compared using Student t tests. Categorical data were compared using Fisher exact test FIGURE 3. Axial T2-weighted magnetic resonance image of an intact left (white arrow) and a ruptured right (gray arrow) anterior sacroiliac ligament. FIGURE 4. Axial T1-weighted magnetic resonance image of a rupture of the left sacrospinous ligament (white arrow). Ó 2013 Lippincott Williams & Wilkins www.jorthotrauma.com 43
Gary et al J Orthop Trauma Volume 28, Number 1, January 2014 MRI (Table 1). The percent visualization for each of the 5 structures ranged from 85% to 100% in injured patients without (Tables 1 and 2). The rates of visualization of each ligament in the injured (Table 2) and control (Table 3) groups were independently calculated and are also reported and compared in Table 1. Of note, all structures still had good rates of visualization in the postoperative stage for the injured patients who were treated operatively, with the exception being the anterior sacroiliac ligament that was not visualized in 7 of 18 ligaments (9 patients with bilateral injuries) because of the artifacts from sacroiliac screws. Visualization of the posterior sacroiliac ligaments was limited by sacroiliac screws in only 1 of the 9 patients who underwent postoperative MRI. In the uninjured patients in the control group, at least 96% of all structures except the posterior sacroiliac ligaments were seen on the magnetic resonance images (Table 3). Visualization of the posterior sacroiliac ligaments was possible for only 85% of the uninjured patients. FIGURE 5. Axial T1-weighted magnetic resonance image of a rupture of the left sacrotuberous ligament (white arrow). (2-tailed). Statistical significance was set at P = 0.05. A previous post hoc power analysis (b =80%,a = 0.05) assessed that we would need at least 35 hemipelves to determine the true rate of ligament visualization within 10%. To differentiate the expected rate of ligament injury in the injured patients (100%) from the expected rate in the control group patients (0%) would require only 4 patients in each group to obtain adequate power. RESULTS Visualization of s with MRI Ninety-six percent (95% confidence interval, 94% 98%) of all ligamentous structures were visualized with Evaluation of Injury One hundred percent (25 hemipelves) of the injured pelves had at least 1 ligament injury diagnosed with MRI compared with 0% (n = 52) of the uninjured hemipelves (Table 4). The difference was statistically significant (P, 0.0001). The table shown in the appendix (see Appendix, Supplemental Digital Content 1, http://links.lww.com/bot/a87) details the MRI findings for each ligamentous structure in the injured hemipelves. The ligament injuries observed varied based on the fracture type. As shown in Table 5, none of the APC type I or II injuries had disruption of the posterior sacroiliac ligaments (n = 22), whereas both of the APC type III injuries had that posterior hinge disrupted (n = 2, P = 0.0004) along with the other 4 structures (sacrospinous, sacrotuberous, and anterior sacroiliac ligaments and pelvic floor musculature). Also in keeping with the model of pelvic ring instability, all patients with APC type II injuries had partial or full tears of the anterior sacroiliac ligament (15 ipsilateral, 1 contralateral). However, in contrast to the current theory, we did not observe that APC type II pelvic ring disruptions universally TABLE 1. Visualization of Pelvic Ring s and in Study and Control Groups All 5 Structures (%) Control group 50/52 52/52 50/52 44/52 52/52 248/260 (95) Study group excluding 34/40 40/40 35/35 40/40 40/40 189/195 (97) Total of control group plus 84/92 (91%) 92/92 (100%) 85/87 (98%) 84/92 (91%) 92/92 437/455 (96) study group excluding (100%) Study group including 34/4 40/42 35/42 40/42 40/42 189/210 (90) Total of control group plus study group including 84/94 (89%) 92/94 (98%) 85/94 (90%) 84/94 (89%) 92/94 (98%) 437/470 (93) 44 www.jorthotrauma.com Ó 2013 Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 1, January 2014 MRI for Evaluation of ous Injury TABLE 2. Visualization of Pelvic Ring s and in Injured Patients Without Metal Artifact in Postoperative Imaging No. studied 40 40 35 40 40 Normal 22 36 16 38 29 Injured 12 4 19 2 11 Incomplete 6 0 0 0 0 visualization Percent visualized 85 100 100 100 100 have disruption of the sacrospinous and sacrotuberous ligaments. Of the 16 hemipelves with APC type II injuries, only 8 had tearing of the sacrospinous ligament. Table 3 details the ligamentous injuries seen on magnetic resonance images, with injury patterns classified based on the Young Burgess classification system. DISCUSSION MRI seemed to be able to well visualize the ligaments about the pelvic ring and to differentiate injured from uninjured hemipelves. Gynecologic surgeons treating pelvic floor insufficiency have used MRI to evaluate the anatomy of the pelvic floor 8 11 ; however, to our knowledge, this is the first report of the use of MRI to evaluate ligamentous integrity of the pelvis, despite the widespread use of MRI in other areas of orthopaedics. Our findings in injured patients generally corresponded with the Young Burgess pattern diagnosed by using conventional radiography and computed tomography. With the Young Burgess classification system, APC type I injures are thought to include injury only to the symphyseal ligaments and not to the 5 structures evaluated in our study. In contrast, APC type II injuries are thought to have tearing of the sacrospinous, sacrotuberous, and anterior sacroiliac ligaments without tearing of the posterior sacroiliac ligament. APC type III injuries are theorized to have tearing of all ligaments, as was observed in our study. However, in contrast to this theory, we found that tearing of the sacrospinous ligament and other ligaments of the pelvic floor occurred in approximately half the APC type II injuries. Existing dogma states that both sacrospinous and sacrotuberous ligaments are torn when anterior sacroiliac ligament injury occurs. 2,12 This difference from the theory confirms a finding first reported in a previous cadaveric study of external rotation pelvic ring injuries. 13 For APC injuries of the pelvic ring, a symphyseal diastasis of.2.5 cm has historically been thought to differentiate APC type I injuries, which are universally treated nonoperatively, from APC type II injuries, which are typically treated operatively. This amount of displacement has been thought to imply disruption of the sacrospinous, sacrotuberous, and anterior sacroiliac ligaments based on a cadaveric sectioning video 14 that can no longer be located. 13 Disruption of the anterior sacroiliac ligament signifies an APC type II injury and is traditionally an indication for operative management. 1,4,15 Two recent studies showed that 27% to 50% of APC I injuries widened.2.5 cm during stress examination performed with the patients under anesthesia, indicating that static plain films were perhaps a poor representation of the amount of ligament disruption. 16,17 Additional data have called into question the presumption of anterior sacroiliac ligament injury based on symphyseal diastasis of.2.5 cm. 13 A biomechanical study of fresh-frozen cadaveric specimens demonstrated that failure of the anterior sacroiliac ligaments with a pure torsional moment varied from 1 to 4.5 cm of pubic diastasis and showed that the sacrospinous and sacrotuberous ligaments did not always rupture with displacement that caused anterior sacroiliac ligament rupture. 13 Our findings confirm the biomechanical findings suggested in that cadaveric study. Surgical management of APC type II injuries is not uniform. Although fixation of the symphyseal disruption is common, opinions vary regarding the need for supplemental sacroiliac screws. A recent study used stress examination under anesthesia for APC type II injuries and proposed a modification to the Young Burgess classification system based on the findings. Flexion or extension sagittal plane deformity of.1 cm with push pull examination was deemed a more severe APC type II injury and indicated the need for sacroiliac screw fixation in addition to anterior ring fixation. 17 TABLE 3. Visualization of Pelvic Ring s and in Uninjured Patients (n = 52, 26 patients with bilateral structures) Complete 16 50 17 24 52 Partial 34 2 33 20 0 Poor 2 0 2 8 0 Percent visualized (complete or partial) 96 100 96 85 100 Ó 2013 Lippincott Williams & Wilkins www.jorthotrauma.com 45
Gary et al J Orthop Trauma Volume 28, Number 1, January 2014 TABLE 4. Proportion of ous Injuries Shown by MRI of Patients With External Rotation Injuries of the Hemipelvis* Any Injury Injury Injury Anterior Sacroiliac Injury Posterior Sacroiliac Injury Injury Control group 0/52 0/50 0/52 0/50 0/44 0/52 Study group 20/20 11/20 4/24 19/21 2/24 11/24 P,0.0001,0.0001 0.01,0.0001 0.12,0.0001 *Fisher exact test, 2-tailed. Varying denominators indicate numbers of patients with ligamentous injuries visible on the images. Denominators vary because those whose ligaments were obscured on the images were not included. One case with an APC type I injury was shown by MRI to have an injury to the symphysis pubis. Rupture of the sacrospinous ligament, sacrotuberous ligament, and/or pelvic floor musculature can imply a more severe and unstable injury pattern than that indicated by rupture of the anterior sacroiliac ligament alone. Less severe injury patterns might require lesser interventions. Our data show that precise anatomic patterns of injury can be identified with MRI. Injury to the sacrospinous ligament can indicate a pattern of pelvic instability that is between Young Burgess APC type II and III patterns. It is possible that we have identified 2 variants of APC type II injuries: those with and those without sacrospinous ligament injury. It is possible that outcomes and the ideal treatment might vary based on this finding, and our radiographic findings might correlate with the different patterns recently described based on the examination under anesthesia. 17 These possibilities await further research. Our study had several important limitations. First, as with most orthopaedic classifications, the Young Burgess classification system has been shown to have only moderate interobserver reliability, 18,19 which could have affected our results. We also included the contralateral hemipelvis for LC type III injuries because they include an external rotation type vector. The ligamentous injury patterns in the windswept pelvis might be different from those seen in a pure APC type injury. Also, our sample size was relatively small; however, we showed very significant differences between study and control groups, indicating that the risk of type I error was very small for our primary and secondary outcome measures. All imaging interpretation was performed by an experienced musculoskeletal radiologist, so further work will be needed to show whether our results are generalizable to other centers and radiologists with different levels of experience. Finally, MRI studies in injured patients were accessed with the use of our electronic imaging system without formal blinding of demographic data. The study would have been strengthened by having multiple radiologists review the studies in a blinded fashion. Future studies with multiple radiologists can help to determine the true sensitivity and specificity of MRI for detecting ligamentous injury in pelvic ring disruptions. Strengths of the study include the prospective nature of patient enrollment and data collection. Additionally, the findings are strengthened by the standardization of MRI technique and interpretation. The study has significant statistical power to answer its primary study questions, as detailed above. In conclusion, visualization of the ligaments of the pelvic ring with MRI is possible for most of the patients. ous stability can also be easily determined, in keeping with other ligamentous injuries. Disruption of the anterior sacroiliac ligaments with an APC injury does not seem to always be associated with sacrospinous and sacrotuberous ligament rupture, as previously thought. The APC type II injury pattern might be divided into 2 groups: those with and those without sacrospinous ligament rupture. Based on these data, future studies are needed to determine the clinical usefulness of MRI for pelvic ring injuries. However, in the current series, ligaments of the injured pelvic ring were visualized with the use of MRI. MRI may play an important future role in better determining the structural injuries associated with pelvic ring disruptions and determining the optimal treatment of specific ligamentous injury patterns. TABLE 5. ous Injury to the Hemipelvis by Young Burgess Pattern as Shown by MRI* APC I (n = 3) 1/1 (2) 0/3 (0) 1/3 (0) 0/3 (0) 0/3 (0) APC II (n = 19) 8/16 (3) 2/19 (0) 15/16 (3) 0/19 (0) 3/19 (0) APC III (n = 3) 2/2 (1) 2/2 (1) 2/2 (1) 2/2 (1) 2/2 (1) *Values shown in parentheses are the number of structures not seen on magnetic resonance images. Note that contralateral injuries in LC type III patterns were classified as APC type I, II, or III injuries to the hemipelvis. For all 3 LC type III patterns, the contralateral injuries were classified as APC type II injuries. 46 www.jorthotrauma.com Ó 2013 Lippincott Williams & Wilkins
J Orthop Trauma Volume 28, Number 1, January 2014 MRI for Evaluation of ous Injury ACKNOWLEDGMENTS The authors thank Senior Editor and Writer Dori Kelly, MA, for invaluable assistance with the manuscript. REFERENCES 1. Bucholz RW, Heckman JD, Court-Brown CM. Rockwood and Green s Fractures in Adults. 6th ed. Philadephia, PA: Lippincott Williams & Wilkins; 2006. 2. Burgess AR, Eastridge BJ, Young JW, et al. Pelvic ring disruptions: effective classification system and treatment protocols. J Trauma. 1990;30:848 856. 3. Tile M. Acute pelvic fractures: II. Principles of management. J Am Acad Orthop Surg. 1996;4:152 161. 4. Tile M, Helfet DL, Kellum JF. Fractures of the Pelvis and Acetabulum. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2003. 5. Tornetta P, Matta JM. Internal fixation of unstable pelvic ring injuries. Orthop Trans. 1994;18:727 733. 6. Greenspan A. Orthopaedic Imaging: A Practical Approach. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2004. 7. Marsh JL, Slongo TF, Agel J, et al. Fracture and dislocation classification compendium 2007: Orthopaedic Trauma Association classification, database and outcomes committee. J Orthop Trauma. 2007;21(suppl 10):S1 S133. 8. Margulies RU, Hsu Y, Kearney R, et al. Appearance of the levator ani muscle subdivision in magnetic resonance images. Obstet Gynecol. 2006; 107:1064 1069. 9. Tunn R, DeLancey JO, Howard D, et al. Anatomic variations in the levator ani muscle, endopelvic fascia, and urethra in nulliparas evaluated by magnetic resonance imaging. Am J Obstet Gynecol. 2003;188: 116 121. 10. Tunn R, DeLancey JO, Quint EE. Visibility of pelvic organ support system structures in magnetic resonance images without an endovaginal coil. Am J Obstet Gynecol. 2001;184:1156 1163. 11. Umek WH, Morgan DM, Ashton-Miller JA, et al. Quantitative analysis of uterosacral ligament origin and insertion points by magnetic resonance imaging. Obstet Gynecol. 2004;103:447 451. 12. Tile M. Acute pelvic fractures: I. Causation and classification. J Am Acad Orthop Surg. 1996;4:143 151. 13. Doro CJ, Forward DP, Kim H, et al. Does 2.5 cm of symphyseal widening differentiate anteroposterior compression I from anteroposterior compression II pelvic ring injuries? J Orthop Trauma. 2010;24:610 615. 14. Pennal C, Suthheland G. Fractures of the Pelvis [Motion Picture].Park Ridge, IL: American Academy of Orthopaedic Surgeons Film Library; 1961. 15. Tile M. Pelvic ring fractures: should they be fixed? J Bone Joint Surg Br. 1988;70:1 12. 16. Suzuki T, Morgan SJ, Smith WR, et al. Stress radiograph to detect true extent of symphyseal disruption in presumed anteroposterior compression type I pelvic injuries. J Trauma. 2010;69:880 885. 17. Sagi HC, Coniglione FM, Stanford JH. Examination under anesthetic for occult pelvic ring instability. J Orthop Trauma. 2011;25:529 536. 18. Furey AJ, O Toole RV, Nascone JW, et al. Surgeon variability in the treatment of pelvic ring injuries. Orthopedics. 2010;33:714. 19. Koo H, Leveridge M, Thompson C, et al. Interobserver reliability of the Young Burgess and tile classification systems for fractures of the pelvic ring. J Orthop Trauma. 2008;22:379 384. Ó 2013 Lippincott Williams & Wilkins www.jorthotrauma.com 47