Pelvic trauma: Are radiographs always necessary?

Author: Eric Legome, MD, Assistant Professor of Emergency Medicine; Director, New York University/Bellevue Emergency Medicine Residency, New York University School of Medicine, New York; and Moira Davenport, MD, Assistant Professor of Emergency Medicine, Assistant Professor of Orthopedic Surgery, New York University Medical Center/Bellevue Hospital Center/NYU Hospital for Joint Diseases, New York.

Peer Reviewer: Steven A. Santanello, DO, Medical Director, Grant Trauma Program, Grant Medical Center, Columbus, Ohio.

Introduction

Patients frequently present to the emergency department with the potential for pelvic trauma. Resultant morbidity and mortality varies depending on the mechanism studied (i.e., simple falls vs. high-energy, multi-trauma incidents). High-energy mechanisms (e.g., motor vehicle trauma, auto-pedestrian accidents, motorcycle collisions, falls, and direct crushes) commonly inflict pelvic injury. Pelvic fracture patterns can range from minor (e.g., non-displaced rami fracture) to life threatening (e.g., symphyseal pubis disruption with an unstable sacroiliac joint).

Early and accurate identification of pelvic fractures is critical for developing an effective management strategy that minimizes morbidity and mortality. In addition, recent literature has modified the paradigm of radiologic investigation for all patients with potential pelvic injury, focusing on risk stratification of patients to identify patients who do not need radiographs and the most effective radiographic imaging techniques. Appropriate utilization of radiographic imaging could save both time and expense for the patient.

Before specific injuries can be discussed, it is useful to review the relevant pelvic anatomy. The pelvis is composed of the sacrum, coccyx, and the acetabulum (formed by the fusion of the ilium, ischium, and the pubis). Several strong ligaments are essential to maintaining the integrity of the bony pelvis. The sacrospinous, sacrotuberous, and sacroiliac (SI) ligaments are particularly critical in maintaining the posterior pelvic anatomy.

Two classification systems are used to describe pelvic fractures. The Young-Burgess classification system is based on the mechanism of injury. Lateral compression (LC) fractures are classified by anterior rami fractures and sacral fracture; severity of injury is based on the extent of sacral fracture. This is the most common mechanism encountered in trauma. Anterior posterior compression (APC) fractures are characterized by symphyseal diastasis and anterior rami fractures. Again, severity of injury is based on the extent of sacral fracture and the amount of SI ligament disruption. Vertical shear injuries are typically seen after an axial load and often result in compete disruption of the affected hemipelvis. Several bony structures are affected, including the rami, iliac crest/wing, sacrum, and SI joint. A combination of mechanisms is common, resulting in each hemipelvis experiencing one of the constellations of injury.

The Tile classification system, much less commonly used, describes fractures based on the extent of disruption of the pelvic ring and the resultant stability of the pelvic ring. Type A fractures are either fractures that do not involve the pelvic ring or are minimally displaced fractures that do not affect stability. Type B fractures are rotationally unstable but vertically stable and may result from a variety of mechanisms. Type C fractures result in a pelvis that is unstable in both the vertical and rotational planes.

Minor trauma or no complaints: Are radiographs necessary?

Source: Duane TM, et al. Blunt trauma and the role of routine pelvic radiographs: A prospective analysis. J Trauma 2002;53:463-468.

Dr. Duane and colleagues prospectively evaluated the role of plain pelvic radiographs in blunt trauma patients. All blunt trauma patients presenting to their Level 1 trauma center from February 2000 to February 2001 were evaluated. Two groups were studied. The first group was protocol-driven and had a survey performed with radiologic studies based on survey findings. For the second, non-protocol group, x-rays were completed at the discretion of the treating physicians. There were 520 patients in the protocol group and 1441 in the non-protocol group. No significant differences existed between the groups for basic demographics and hemodynamic parameters. However, patients in the non-protocol group tended to be significantly younger with significantly higher trauma scores and lower Glasgow Coma Scale (GCS) score. In the protocol group, 172 patients were found to have positive physical examination findings; this correlated to 45 positive radiographs. Unfortunately, 75 patients with negative exam findings also received plain radiographs, breaking the study protocol; all exams were negative. No fractures were missed on physical examination alone. The physical examination at discharge was considered the gold standard. The authors conclude that clinical examination is a reliable means to detect pelvic fracture in awake trauma patients, given the 100% sensitivity in this study.

Commentary

Several methodological flaws question the clinical reliability of this study. The variable compliance of the trauma teams in completing the data form resulted in large enrollment differences and significant discrepancies in injury severity score (ISS) and GCS between protocol and non-protocol groups, with the non-protocol group having more significant injuries. This may have underestimated the utility of the physical exam in the severely injured/unstable patient. Of note, the authors did retrospectively evaluate the medical records of the non-protocol fracture patients and were able to identify physical examination findings that would have indicated the need for plain radiographs had the study data sheet been completed.

Source: Yugueros P, et al. Unnecessary use of pelvic x-ray in blunt trauma. J Trauma 1995;39:722-725.

All blunt trauma patients admitted to a tertiary care hospital over a 3-month period were considered for inclusion. The study prospectively enrolled 608 hemodynamically stable patients with a GSC greater than 10 and no clinical spinal cord injury. All subjects underwent physical examination and plain x-rays; patients were followed up in an unspecified amount of time by orthopedic surgeons. Fifty-seven patients had positive physical exam findings and all of their fractures were confirmed on x-ray. Two patients had positive x-rays and negative physical examinations. Both of these fractures were deemed clinically insignificant and did not result in any long-term complications. The authors conclude that physical examination has a 99% negative predictive value in evaluating the potential for pelvic fractures in blunt trauma. Hospital policy regarding mandatory pelvic x-rays in blunt trauma was changed as a result of this study. In the 2 years following the implementation of the new policy, the authors report no additional missed pelvic fractures in their blunt trauma population.

Commentary

The study by Dr. Yugueros and colleagues effectively enrolled 608 consecutive trauma patients over a 3-month period. Significant method-ological flaws exist in this study — most importantly, the plain radiograph is used as the gold standard, which has been proven to miss clinically significant fractures. Another weakness to the study was the relatively low number of patients with fractures. Although this study has several limitations, it does demonstrate that there is a subpopulation of trauma patients in whom it is sometimes safe to forego pelvic radiography. Patients who have the potential to be screened and not receive radiographs would need to be stable, and the clinician must be able to effectively evaluate the patient.

Limiting pelvic radiographs

Source: Gross EA, Niedens BA. Validation of a decision instrument to limit pelvic radiography in blunt trauma. J Emerg Med 2005;Apr;28(3):263-266.

This prospective observational study evaluated whether a set of criteria could predict, with a high degree of sensitivity and a high negative predictive value, which patients may safely forgo pelvic radiography in their initial trauma evaluation. Adult blunt trauma patients presenting from July 1, 2002 to June 30, 2003 who underwent pelvic radiography were eligible for the study. Whether to obtain radiography was at the physician's discretion. Physicians completed data points from 5 observations — altered level of consciousness, pelvic pain, pelvic tenderness on examination, distracting injury, and clinical intoxication — prior to interpreting the radiographs. Final radiograph results were then linked to the form. The correlation between initial (ED) and final (radiology) reads was not noted. Fractures were classified as clinically significant or insignificant based on the Tile classification. There were 973 patients enrolled in the study; 62 patients had pelvic fractures (prevalence 6.4%). The decision instrument predicted fracture in 60 patients (sensitivity 96.8%, 95% CI 92.4%-100%). Two fractures were considered clinically insignificant. If only clinically significant fractures were considered, the instrument had a sensitivity of 100%. The specificity was 47.5%. Using the decision rule would have decreased radiography approximately 45% without increasing morbidity.

Commentary

This paper strengthens the hypothesis that a blunt trauma patient who is awake and alert, denies pelvic pain or tenderness, and is without distracting injury can forego radiography.

While this paper's data support this claim, there are several methodologic concerns. First is that the pelvic radiograph is, by itself, not a perfect gold standard. Within this study there may have been several more misses, although probably not of overly significant fractures. Use of computed tomography (CT) diagnosis, when available, would have given a more accurate picture of the rule. Secondly, it was not noted if any of the patients were hypotensive or in shock. While distracting injury was an exclusion to forego radiograph, it was not clearly defined. While the NEXUS criteria used a similar criterion, most physicians would want a pelvic radiograph in the hypotensive blunt-trauma patient. There may have been a selection bias in patients studied, although probably consistent with how patients usually are selected for radiographs. Lastly, the criteria were not prospectively derived. This may have missed useful positive or negative criteria.

CT in Pelvic Fracture: A Substitute for Plain Film?

Source: Guillamondegui OD, et al. Pelvic radiography in blunt trauma resuscitation: A diminishing role. J Trauma 2002;Dec;53(6):1043-1047.

The authors, in an attempt to challenge ATLS dogma regarding the need for routine plain pelvic radiography, evaluated the clinical treatment of blunt trauma patients to determine if the plain pelvic film was necessary in all blunt trauma patients with a significant mechanism of injury. Dr. Guillamondegui and colleagues retrospectively reviewed radiology reports of all portable pelvic films and CT scans performed on the patients undergoing abdomen and pelvic CT scanning within 4 hours of arrival over a period of a year (2000–2001) at their urban Level I trauma center. CT scan findings were considered the gold standard in diagnosing pelvic fractures. Slightly fewer than half of the patients received a portable pelvic radiograph followed by a CT scan, while the rest received a CT scan only. The decision to proceed directly to CT was left to the treating physician; reasons were not documented. The patients who had both radiograph and CT scan were slightly more injured, based on ISS (12.3 +/- 0.7 vs. 8.0 +/- 0.5). A total of 686 patients with blunt trauma underwent CT scanning of the abdomen and pelvis. Group I consisted of 311 patients who underwent CT and plain pelvic film (PXR). Fifty-six (10%) patients were found to have at least 1 pelvic fracture on CT scan, 38 of which were also identified on the PXR. The sensitivity and specificity of the PXR in Group I was 68% and 98%, respectively. The false-negative rate for pelvic radiography was 32%. In all patients with a positive PXR, the majority (55%) had either additional fractures or an increase in the severity of fracture diagnosed on CT scan. Group II consisted of 375 patients, with 16 fractures noted in 13 (3%) patients, none of which required treatment.

The authors concluded that the CT scan is clearly better than a portable pelvic radiograph, both in diagnosing the presences of a fracture and the grading the severity of the fracture. They feel that if an abdominal pelvic CT is to be performed in the blunt-trauma patient, there is generally no indication for a portable film, as it provides no additional useful information. They do, however, recommend PXR in unstable patients, those with positive physical findings, or those who cannot undergo CT scanning because of other clinical priorities.

Commentary

This study helped introduce the concept of foregoing the pelvic radiograph in the trauma bay if CT is planned. The conclusion that the CT scan can replace the portable anteroposterior (AP) pelvis in the stable patient without evidence of acute pelvic injury is limited because the study did not really address specific criteria that could be used to stratify the characteristics of the patients. It is probably reasonable to find, however, that the CT scan (assuming a high specificity or the gold standard) is more accurate than the plain radiograph in the blunt trauma patient. They found significantly more fractures, including those that would change management, on CT scan than on plain film. While there initially appeared to be fractures on the radiographs not seen on the CT scan, clinical correlation showed them to be false positives. Given these findings, it seems reasonable to assume that it is appropriate to forego plain film in stable patients who will undergo CT. This is especially true today, as CT scanners allow for faster acquisition of data and better detail and resolution than when the study was performed. This should be predicated, however, on the assumption that a good physical examination has been performed and the reasonable certainty that the patient does not have an unstable injury or a hip dislocation that requires immediate attention.

Source: Vo NJ, et al. Pelvic imaging in

the stable trauma patient: Is the AP pelvic radiograph necessary when abdominopelvic CT shows no acute injury? Emerg Radiol 2004;Apr;10(5):246-249.

This retrospective descriptive study attempted to determine the utility of AP pelvic radiographs in stable trauma patients who underwent abdominopelvic radiography at the time of their initial evaluation. Attending radiologists retrospectively re-viewed (in a prospective manner without knowledge of the clinical course) radiology reports of all stable trauma patients who underwent both abdominopelvic CT and AP pelvic radiograph over a 4-month period in 2003 for findings of acute pelvic injuries. The series included 509 consecutive patients who were imaged using a multi-detector scanner with 5-mm or 7.5-mm slices through at least the level of the inferior pubic rami. No acute pelvic injury was found on abdominopelvic CT in 449 patients (88.2%). CT showed 163 acute injuries in 60 patients. AP radiographs showed 132 acute injuries in 52 patients. No patients with negative CT had an acute finding on the pelvic radiograph. There were 8 false-negative pelvic radiographs (negative predictive value 98.25%). In a single case, a transverse fracture of the left superior pubic ramus apparent on plain films was not described in the CT report, but this patient had multiple concomitant injuries demonstrated on both CT scans and plain films. With CT scan as the gold standard, missed injuries included 3 iliac fractures, 8 pubic bone fractures, 4 SI diastases, 8 sacral fractures, 8 acetabular fractures, and 3 loose bodies in the hip joint. AP radiograph showed several dislocations that were not on CT, but they were explained by intercurrent reductions. Overall, 8 AP radiographs were read as completely normal when fractures were found on the CT. The areas of greatest discrepancies were injuries to the sacrum, SI joints, and loose bodies in the hip joints. The sensitivity for plain radiograph was 77% for acute pelvic injury. However, the radiologists reading were not blinded to abdominopelvic CT findings in some cases, leading to a sensitivity that is probably higher than in actual practice.

Commentary

This study makes a compelling argument to forego the plain radiograph in the hemodynamically stable trauma patient who is to undergo CT scan. In their institution, they would have realized a $53,000 cost saving along with decreased patient radiation exposure without any loss of diagnostic accuracy. While the study may have been more robust by performing a prospective study where the radiologists were blinded to the abdominopelvic CT, this would probably only serve to decrease the sensitivity of the plain radiographs. It would also have been useful to know how many of the fractures were actual or potentially unstable fractures, or required operative intervention.

The CT is considered the gold standard, with an assumption that all positives are true positives. While a reasonable assumption, this has never been proven to be true. Foregoing the radiograph may not be a realistic practice in a center where there is a long wait time to perform CT, as the management and need for consultants may change or the patient may become unstable. However, in many EDs with adjacent CT and rapid availability, the authors make a compelling case for holding off on pelvic radiography in the stable trauma patient who will undergo CT as a part of the initial trauma evaluation. This will lead to a more accurate study with probably improved efficiency.

Does plain radiography miss hip and pelvic fractures?

Source: Dominguez S, et al. Prevalence of traumatic hip and pelvic fractures in patients with suspected hip fracture and negative initial standard radiographs — A study of emergency department patients. Acad Emerg Med 2005;12:366-369.

Dr. Dominguez and colleagues compared plain hip/pelvis radiography to magnetic resonance imaging (MRI) in an attempt to determine the number of fractures missed by plain x-rays. Their academic community hospital protocol is such that all patients with negative initial x-rays but persistent pain underwent MRI at the discretion of the treating emergency physician. This retrospective review examined all patients undergoing plain hip/pelvis radiography in a 6-month period. The study group consisted of 764 patients, 219 of whom had a positive plain radiograph. Of the remaining 545 patients, 62 went on to MRI. Twenty-nine fractures were identified in 24 patients, for a missed fracture rate of 9.9%. Of the missed fractures, 34.5% were injuries to the pubic rami. Followup of the remaining patients with negative radiographs identified did not reveal any subsequently discovered fractures.

Commentary

Dominguez and colleagues further the examination of plain film efficacy in suspected hip/pelvis fractures. The study has several strengths, notably the age of the patients (median >66) and the mechanism of injury (95.6% sustaining blunt trauma), both of which are consistent with demographics seen in most EDs. The ability to rapidly obtain MRI scans furthered their findings, as this modality is currently considered the gold diagnostic standard. It is interesting to note that of the fractures discovered on MRI, 34.5% were injuries to the pubic rami. As this injury is often treated conservatively, one must question the cost effectiveness of MRI in this case.

As with most retrospective studies, this work is weakened by the retrospective nature of enrollment and the selection bias inherent in this design. The lack of formal protocol may underestimate the results, since it is possible that several emergency physicians may not have ordered MRI scans immediately. This is particularly relevant in regards to pubic rami fractures, which are often missed on initial radiography. It is interesting to note that only 1 patient in 6 months underwent a CT scan to further evaluate hip pathology. This modality may be of particular benefit in the polytrauma patient or the hemodynamically unstable patient.

MDCT: Is it better?

Source: Herzog C, et al. Traumatic injuries of the pelvis and thoracic and lumbar spine: Does thin-slice multidetector-row CT increase diagnostic accuracy? Eur Radiol 2004; Oct;14(10):1751-1760.

CT scan is increasingly used to evaluate traumatic chest and abdominal/pelvic injuries. Multidetector CT (MDCT) scanners are also able to use collected and reformatted data for evaluation of bony trauma. The authors studied different MDCT strategies to best classify spinal and pelvic injuries and to evaluate if scans can replace conventional radiography in diagnosing spinal trauma. The authors also wanted to evaluate which reconstruction parameters allowed the best diagnostic accuracy for both pelvic and spinal injuries. Seventy intubated multi-trauma patients, with a variety of mechanisms, underwent conventional radiography and MDCT. Examinations included the pelvis, the lumbar spine, and the thoracic spine. Reviews were done by 2 independent radiologists. There was excellent correlation between their interpretations. Conventional radiographs, 3-mm and 5-mm scans, and 3-mm and 5-mm scans combined with multiplanar reformations were compared to surgery or autopsy (n = 33) and clinical course (n = 35). There were 22 thoracic spine fractures, 30 lumbar spine fractures, and 39 pelvic fractures, of which 30 pelvic fractures were considered unstable. Sensitivities ranged from 57% to 87% for all fractures for conventional radiographs and 57% to 80% for unstable fractures. For 5-mm transverse CT slices, sensitivities ranged from 83% to 90%, but 100% for all unstable fractures. For 3-mm slices, the sensitivities ranged from 85% to 100% for all fractures and 100% for all unstable fractures. Specificity was 100% for all CT findings. Two scans — 5-mm slices plus reconstructions and 3-mm slices plus reconstructions — had 100% sensitivity for all fractures except stable pelvic fractures, with close to 100% negative predictive values.

Commentary

As expected, this study showed a superiority of imaging by CT scan of fractures as compared to plain radiography, consistent with previous studies. The unique part of this study is the attempt to compare different types of multiplanar reconstructions using CT. While it may be somewhat confusing to the non-radiologist, the bottom line is that 3-mm or 5-mm slices along with reconstructions show high negative and positive predictive values. While 3-mm slices are slightly better, they do not seem to diagnose additional unstable fractures. Simple transverse scanning without reconstructions diagnose almost all unstable fractures, but still miss a significant minority of stable fractures. While the decision of slice thickness is usually made by radiologist, understanding the type of CT scanner used and type of CT performed can affect our use of data. When relying on CT to exclude fractures, it is important that reconstructions are performed.

Overall, this was a relatively well done study, although the use of clinical followup on some patients may have underestimated the number of fractures; however, it's unlikely that any of these missed fractures were unstable injuries.

How reliable is the clinical examination in detecting pelvic fractures?

Source: Sauerland S, et al. The reliability of clinical examination in detecting pelvic fractures in blunt trauma patients: A meta-analysis. Arch Orthop Trauma Surg 2004:124;123-128.

Given the recent controversy regarding the necessity of plain radiographs in apparently minor blunt pelvic trauma, Dr. Sauerland and colleagues conducted a meta-analysis to determine if the physical examination was reliable in detecting pelvic fractures. Twelve studies with 5454 patients were identified. Pooled sensitivity and specificity were both reported as 0.90. Both measures improved slightly when patients with altered mental status were excluded from review. Among the 441 patients with pelvic fracture (and normal mental status), only 3 fractures were missed on physical examination.

Commentary

As with all meta-analyses, one must consider the design of the included studies and the overall methodology of the analysis. Dr. Sauerland and colleagues mention that some selection bias is expected, based on the wide reported range of fractures; this may overestimate results. Furthermore, most studies used plain AP radiographs as the gold diagnostic standard. However, this lack of consistency may falsely elevate or lower the sensitivity/specificity based on alternate methods used. This study, along with the studies by Guillamondegui and Vo, may obviate the need for plain radiographs in the alert blunt trauma patient with clinical indications for immediate CT scanning. It must be noted however, that when the radiograph is used as the gold standard, the sensitivity of the physical exam may be falsely elevated; what this means for clinical practice is not fully clear.

Major trauma: What can the FAST tell us?

Source: Rucholtz S, et al. Free abdominal fluid on ultrasound in unstable pelvic ring fracture: Is laparotomy always necessary? J Trauma 2004;57:278-285.

Significant energy is required to produce an unstable pelvic fracture. Fractures involving the posterior ring of the pelvis predispose patients to significant retroperitoneal bleeds. Severe hemorrhages can often extend into the intraperitoneal cavity if the rate of bleeding is brisk and if the initial injury is severe or involves multiple vessels. In recent years, the FAST exam has been incorporated into ATLS protocol in an attempt to rapidly detect intraperitoneal fluid. Dr. Rucholtz and colleagues designed this study to determine the sensitivity and specificity of the FAST exam for detecting intraperitoneal free fluid in the setting of an unstable pelvic fracture.

All trauma patients presenting to the ED from May 1998 through December 2002 were prospectively enrolled in this study. A total of 1472 patients were admitted during this time period, but only those with Grade B or C fractures according to the AO/SICOT classification (i.e., most severe fractures), were included for a total of 80 patients. All patients underwent standard ATLS protocol, including FAST examination. Thirty-one patients had positive FAST examinations; 15 of these patients were hemodynamically unstable in the ED. Hemodynamically unstable patients with concurrent unstable pelvic ring fractures underwent immediate stabilization, usually via external fixator, and continued through the ATLS evaluation. The authors calculated a 97% positive predictive value for ultrasound detection of intraabdominal injury with concomitant pelvic ring fracture and a sensitivity of 75%.

Commentary

Dr. Rucholtz and colleagues determined that ultrasound detected free intraabdominal fluid in the setting of an unstable pelvic fracture. The presence of free fluid was highly indicative of significant intraabdominal pathology and usually required laparotomy. This study was conducted among a group of severely injured patients, and serves to reinforce the idea that ultrasound is a viable adjunct to the traditional ATLS evaluation. Patients with positive FAST exams had significantly higher ISS scores, higher rates of hemodynamic instability, and higher rates of mortality than those with a negative FAST. This study might have been expanded to include patients with type A pelvic fractures. Inclusion of less severe pelvic injuries would have challenged the diagnostic capability of ultrasound in less clinically distinct cases; it is in these cases that FAST findings may limit the number of laparotomies performed by favoring angiography followed by CT before a trip to the operating room.

Conclusions

Pelvic trauma, both major and minor, is common in the ED. The emergency physician has several diagnostic options when evaluating patients with suspected pelvic trauma. In the stable patient, the physician must decide whether to obtain any imaging study on the patient or rely on physical examination to exclude a fracture. While the evidence is not fully conclusive, there is a strong inference from ED-based studies that physical examination alone can be appropriate to exclude the vast majority of fractures and detect clinically significant fractures. If the need for a radiograph cannot be excluded based on physical examination, there are several options for imaging.

If the patient is stable and will undergo abdominal and pelvic CT scan, it appears safe, cost effective, and appropriate to proceed directly to CT without a screening radiograph. The CT is a better study and there is limited, if any, value in obtaining a plain radiograph in addition to the CT in this patient subset. However, if the patient is hemodynamically compromised or there is a high suspicion of an unstable fracture, the question becomes, "Should a portable radiograph be obtained first?" It does not add to the overall sensitivity, but may change methods or order of further diagnostic testing and treatment. Conventional CT is a better test than plain film for fracture, although it still is not 100% sensitive. MDCT appears to diagnose all unstable fractures and almost all stable fractures that are clinically important. The emergency physician should be familiar and play a role in formulating institutional policy for which diagnostic modality is the initial test of choice. This consideration may strongly influence the decision-making process.