utility of Plain Film Pelvic Radiographs in Blunt Trauma Patients in the Emergency Department AMAL KAMIL OBAID, M.D., ANDREW BARLEBEN, M.D., DIANA PORRAL, B.S., STEPHANIE LUSH, M.S.N., MARIANNE CINAT, M.D. From the University of California Irvine Medical Center, Orange, California The objective of this study was to evaluate the utility and sensitivity of routine pelvic radiographs (PXR) in the initial evaluation of blunt trauma patients. A retrospective review was performed. One hundred seventy-four patients with a pelvic fracture who had computed tomography (CT) and PXR were included (average age, 36.; average Injury Severity Score, 6.3). Nine (5%) patients died. Five hundred twenty-one fractures were identified on CT. One hundred sixteen (22%) of these fractures were missed by PXR. Eighty-eight (5%) patients were underdiagnosed by PXR alone. The most common fractures missed by PXR were sacral and iliac fractures. Eight patients required angiograms, with four undergoing therapeutic pelvic embolization. Forty-seven (27%) patients were hypotensive or required a transfusion in the emergency department. These patients were more likely to require an angiogram (7% vs 0%, P < 0.000) and were more likely to require embolization (9% vs 0%, P < 0.00). This study demonstrates that CT scan is highly sensitive in identifying and classifying pelvic fractures. PXR has a sensitivity of only 78 per cent for identification of pelvic fractures in the acute trauma patient. In hemodynamically stable patients who are going to undergo diagnostic CT scan, PXR is of little value. The greatest use of PXR may be as a screening tool in hemodynamically unstable patients and/or those that require transfusion to allow for early notification of the interventional radiology team. T HE ADVANCED TRAUMA LIFE SUPPORT (ATLS) course was compiled by the American College of Surgeons and provides guidelines for the appropriate management of trauma patients. Current ATLS protocol recommends that an anteriorposterior pelvic radiograph (PXR) be performed in the initial evaluation of all patients with a blunt mechanism of injury.' With the advent and widespread use of computed tomographic (CT) scanning, the utility of the plain film PXR during the acute resuscitation of the stable trauma patient has come into question. The purpose of this study was to evaluate the utility and sensitivity of PXR in the initial evaluation of trauma patients. Our hypothesis was that plain films of the pelvis in hemodynamically stable patients who will undergo a CT scan of the pelvis do not add significant Presented at the 7th Annual Conference of the Southem Califomia Chapter of the American College of Surgeons, Bacara Resort and Spa, Santa Barbara, CA, January 20-22, 2006. Address correspondence and reprint requests to Marianne Cinat, M.D., Department of Surgery, University of Irvine Medical Center, 0 The City Drive, Building 53, Route 8, Orange, CA 92868. information to the initial evaluation and will miss many pelvic fractures identified by CT. We also sought to identify a subset of patients who may benefit from PXR during the immediate resuscitation by altering or facilitating the treatment course. Materials and Methods This was a retrospective cohort study conducted at an urban Level I trauma center. From January 2004 to June 2005, all patients with pelvic fractures were identified who had PXR and CT of the pelvis. The trauma registry (Trauma One database; Lancet Technology Inc., Cambridge, MS) was reviewed. Information was recorded regarding patient demographics, injury severity score (ISS), transfusion therapy, abbreviated injury score (AIS), type of pelvic fracture, angiographic evaluation, operative intervention, and death. Final radiographic diagnoses of pelvic fracture as dictated by attending radiologists were recorded for plain film radiographs and CT scans of the pelvis. The ISS and AIS for each patient were calculated and recorded using standard methodology.^- ^ Overall mortality was determined. Statistical analysis was performed using Student's t tests to compare groups. Data was reported as mean ± SEM. 95
952 THE AMERICAN SURGEON October 2006 Vol. 72 Results One hundred seventy-four patients were included (Table ). The average age ofthe patients was 36. ± 5.5 years (range, 3-82). There were 07 men and 67 women. The mechanism of injury was blunt trauma in all cases. The average ISS was 6.3 ± 8.8. Seventyfive (43%) patients had significant injury as defined by ISS s 6. Thirty-four (9%) patients had critical injury as defined by ISS > 25. Nine patients died (mortality, 5.2%). Thirty-four (9%) patients required blood transfusions in the ED. The average number of units transfused was 3.8. Twenty-nine (6.9%) patients were hypotensive in the ED as defined by a SBP <90. Seventy patients required operative intervention during their hospital stay. Forty-six (40%) patients required an operative procedure within the initial 24 hours of admission. 'Ten patients required an urgent pelvic procedure within 24 hours of admission; 2 of these were external fixators. Ten patients required a nonurgent pelvic procedure during their hospitalization. Eifteen patients required an urgent laparotomy. A total of 52 fractures were diagnosed in 74 patients (Table 2). The most common fractures identified were pubic rami fractures (n = 26, 50%), sacral fractures (n = 95, 8%), and acetabular fractures (n = 60, %). PXR missed 6 fractures (22% of fractures) in 88 patients (5% of patients), giving PXR a sensitivity of only 78 per cent. The most common fractures missed on PXR included iliac and sacral fractures (45% missed injury rate for both; Table 2). Eifteen (8.6%) patients had a completely negative PXR and fractures were subsequently identified on CT scan. None of these patients required angiogram; three of these patients required emergent laparotomy. A total of eight patients required angiograms. All eight patients were hypotensive or required transfusion while in the ED. Eour patients had therapeutic pelvic embolization. PXR identified at least one fracture in all eight of these patients. A total of 24 fractures were identified in these eight patients by CT; 7 fractures were missed by PXR. TABLE I. Patient Demographics Average age 36 ± 9 years Gender: Male 07(6%) Female 67 (39%) Average ISS I6±ll ISS > 6 (patients) 75 (43%) ISS > 25 (patients) 34 (20%) Patients requiring pelvic operation 20 (%) Transfusion in emergency department (ED) (patients) 33(9%) Average no. of units PRBCs transfused 3.8 units Systolic blood pressure (SBP) < 90 in ED (patients) 29(7%) TABLE 2. Classification of Peivic Fractures Identified hy CT Scan and Fractures Missed hy PXR Juxta acetabular Iliac Sacral Parasymphyseal Acetabular Pubic rami Sacroiliac joint diastasis Ischial Pubic symphysis diastasis Total Fractures Identified by CT 40 95 4 60 26 27 0 3 52 Fractures Missed by PXR 8 43 4 6 29 3 6 Percentage of Fractures Missed 00% 45% 45% 29% 27% % % 0% 8% 22% Eorty-seven (27%) patients were hypotensive or required a transfusion in the ED. These patients were more likely to require a pelvic angiogram (7% vs 0%, P < 0.000) and were more likely to require embolization (9% vs 0%, P < 0.000; Table 3). There were three patients that had fracture reported on PXR but not on CT scan. In reviewing the medical records of these patients, no pelvic pain was reported on physical examination and the patients were able to ambulate without difficulty. Given this clinical history and physical finding, it was concluded that the PXR was likely incorrect. Discussion ATLS guidelines recommend that a plain PXR be performed as an adjunct in the evaluation of all blunt trauma patients. Our study suggests that this practice may not be necessary. Routine PXR in hemodynamically stable patients who will undergo CT scan evaluation of the pelvis as part of their initial work-up appears to be of little value. Our study demonstrates that PXR had a high missed injury rate. PXR was read as negative in 9 per cent of patients who subsequently had a fracture identified on CT. PXR missed 22 per cent of fractures identified on CT and underdiagnosed of 5 per cent of patients with pelvic fracture on CT. However, some patients may benefit from PXR in the ED. In hypotensive patients and in those that require transfusion, PXR can serve as a screening tool to identify patients who may benefit from early pelvic TABLE 3. Impact of Hypotension or Transfusion in the FD ED Characteristics n Angiogram Embolization SBP < 90 29 Transfusion 34 SBP < 90 and/or transfusion 47 None of the above 27 2 3 4 0
No. 0 UTILITY OF PLMN PXR IN BLUNT TRAUMA PATIENTS IN THE ER Obaid et al. 953 angiography. This patient population has a 7 per cent chance of requiring angiogram, and a 9 per cent chance of requiring embolization. The presence of a pelvic fracture on PXR in hypotensive patients or patients that require transfusion would allow for early notification and mobilization of the interventional radiology team. Previous studies have identified patients that may benefit from PXR based on physical examination findings in the ED. Yugueros et al."* performed a prospective analysis and found that a negative physical examination had a negative predictive value of 99 per cent in patients that are adults with a Glasgow Coma Score of greater than 0, are hemodynamically stable, have no evidence of blood loss, and have no spinal cord injury. Similarly, Gross et al.^ established a decision instrument to determine which patients should undergo PXR in the ED. The decision tool included five criteria: altered level of consciousness, complaint of pelvic pain, pelvic tenderness, distracting injury, and the presence of clinical intoxication. The presence of any of these factors initiated a PXR in the ED. Nine hundred seventy-three patients were included. This decision tool identified fracture in 60 of 62 patients who had pelvic fractures (sensitivity, 96.8%). The two fractures that were missed were considered clinically insignificant. Using this tool, approximately 44 per cent of the population could have foregone PXR in the ED. Clinical findings on physical examination were not included in this study. Only objective data (i.e., blood pressure, transfusion requirements, patient demographics, etc.) were included. Physical examination findings and reporting can be subjective and inconsistent if obtained retrospectively from dictated or written reports. To eliminate this source of error, physical examination fmdings were not analyzed. CT scan has become the gold standard for the diagnosis of solid organ,^''' retroperitoneal,^ and orthopedic injury to the pelvis and spinal column.^-' Most patients who sustain significant blunt force trauma undergo CT scan of the abdomen and pelvis.^ When compared with CT scanning, PXR had a high missed injury rate (22%), and did not identify all fractures in 5 per cent of patients with fractures diagnosed on CT. The most common injuries missed were sacral and iliac fractures, each with a 45 per cent missed injury rate. This data is corroborated by earlier reports. Guillamondegui et al.^ retrospectively evaluated 30 pediatric blunt trauma patients with pelvic fracture who had CT scan and PXR. PXR missed 54 per cent of fractures identified by CT scan. An increase in the grade of the fracture was also reported by CT scan. Berg et al.'^ similarly showed that plain films of the pelvis missed 34 per cent of fractures identified by CT scan. CT scan provided additional information such as injury force pattern, pelvic ring stability, and posterior ring fractures that could not be identified on plain film.'^ Finally, CT scan can provide information about soft tissue involvement, active extravasation, and the presence and size of a pelvic hematoma, all of which may guide the need for further evaluation such as angiogram and embolization.'^ None of this information can be obtained accurately from a plain film PXR. A selective approach to PXR in blunt trauma patients can result in significant cost savings. In this study, if PXR was limited to those only those patients with hypotension or a transfusion requirement, 27 patients would have foregone PXR. Assuming a cost of $8 per PXR, this would have resulted in a cost savings of $22,987. Similar findings have been reported by other authors. Duane et al.''^ prospeetively showed a cost savings of $5,50 by eliminating 273 films in patients who were awake and alert and lacked signs or symptoms of pelvic and/or back injury. Salvino et al.'^ similarly demonstrated that routine PXR resulted in an increase cost of $88,028. One caveat to this study is that all patients underwent CT scan and PXR. During evaluation in the ED, if a patient is not going to undergo CT scan, PXR should be obtained to rule out pelvic pathology. This may occur in patients who are to undergo an urgent procedure or in patients who do not require further diagnostic evaluation. In conclusion, we have established that PXR in the acute trauma patient has a high missed injury rate (22%) compared with CT scan and can be avoided in hemodynamically stable patients who are to undergo CT scan. However, any patient who is hemodynamically unstable or requires transfusion in the ED should undergo PXR to identify pelvic fracture as a potential source of blood loss. Early diagnosis is imperative in this patient population so that early stabilization maneuvers such as external fixation, pelvic binders, and angiogram can be instituted. Early mobilization of the interventional radiology team can be a significant factor in reducing morbidity and potentially mortality in this patient population. Patients who are not going to undergo CT scanning or have significant physical findings should also receive a PXR. However, in the hemodynamically stable patient without significant physical examination findings who is undergoing CT scan, PXR can be safely avoided. REFERENCES. American College of Surgeons. Committee on Trauma, Abdominal Trauma, Advanced Trauma Life Support Program. Chicago, IL: American College of Surgeons, 993, pp 4-55. 2. Baker SP, O'Neill B. The Injury Severity Score: a method for
954 THE AMERICAN SURGEON October 2006 Vol. 72 describing patients with multiple injuries and evaluating emergency care. J Trauma 974; 4:87. 3. Baker SP, O'Neill B. The injury severity score: an update. J Trauma 976;6:882-5. 4. Yugueros P, Sarmiento JM, Garcia AF, Ferrada R. Unnecessary use of pelvic X-ray in blunt trauma. J Trauma 995;39: 722-5. 5. Gross EA, Niedens BA. Validation of a decision instrument to limit pelvic radiography in blunt trauma. J Emerg Med 2005; 28:263-6. 6. Hawkins ML, Wynn JJ, Schmach DC, et al. Nonoperative management of liver and/or splenic injuries: effect of resident surgical experience. Am Surg 998;64:552-7. 7. Freshman SP, Wisner DH, Battistella FD, et al. Secondary survey following blunt trauma: a new role for abdominal CT scan. J Trauma 993;34:337^. 8. Pollack HM, Wein AJ. Imaging of renal trauma. Radiology 989;72:297-308. 9. Dunn EL, Berry PH, Connally JD. Computed tomography of the pelvis in patients with multiple injuries. J Trauma 983;23: 378-83. 0. Pereira SJ, O'Brien DP, Luchete FA, et al. Dynamic helical computed tomography scan accurately detects hemorrhage in patients with pelvic fractures. Surgery 2000; 28:678-85.. Guillamondegui OD, Pryor JP, Gracias VH, et al. Pelvic radiography in blunt trauma resuscitation: a diminishing role. J Trauma 2002;53:043-7. 2. Berg EE, Cheebuhar C, Bell RM. Pelvic trauma imaging: a blinded comparison of computed tomography and roentgenograms. J Trauma 996;4:994-8. 3. Miller PR, Moore PS, Mansell E, et al. External fixation or arteriogram in bleeding pelvic fracture: initial therapy guided by markers of arterial hemorrhage. J Trauma 2003;54:437^3. 4. Duane TM, Tan BB, Golay D, et al. Blunt trauma and the role of routine pelvic radiographs: a prospective analysis. J Trauma 2002;53:463-8. 5. Salvino CK, Esposito TJ, Smith D, et al. Routine pelvic x-ray in awake blunt trauma patients: a sensible policy? J Trauma 992;33:43-6.