Focused assessment with sonography for trauma (FAST)

Transcription

1 Predicting the Need for Laparotomy in Pediatric Trauma Patients on the Basis of the Ultrasound Score Adrian W. Ong, MD, Mark G. McKenney, MD, Kimberley A. McKenney, MD, Margaret Brown, RN, MSN, Nicholas Namias, MD, Jana MaCloud, MD, MSc, FRCS(C), and Stephen M. Cohn, MD Background: It is possible to quantify the amount of hemoperitoneum seen on focused assessment with sonography for trauma (FAST) using a simple scoring system that had previously been shown to correlate with the need for subsequent laparotomy in adults. A score of 3 or greater was shown to be highly accurate in predicting the need for laparotomy. We hypothesized that this scoring system might also predict the need for laparotomy in pediatric trauma patients. Methods: We retrospectively reviewed all records for patients 15 years and younger who underwent FAST after blunt trauma. A positive ultrasound examination was defined as one containing free intraperitoneal fluid with or without solid organ injury. The ultrasound score (USS) was defined as the depth of the deepest pocket of fluid collection measured in centimeters plus the number of additional spaces where fluid was seen. Results: Thirty-eight (19.6%) of 193 patients who had FAST performed had positive ultrasound examinations. Thirtyseven patients with complete records were analyzed. There were no differences between patients with a USS < 3.0 and those with a USS > 3.0 in terms of admission pulse, Glasgow Coma Scale score, Injury Severity Score, or the proportion of patients who were initially hypotensive. One of 22 patients with a USS < 3.0 required therapeutic laparotomy versus 8 of 15 patients with a USS > 3.0 (p 0.002). For a USS > 3.0, sensitivity, specificity, and accuracy in predicting therapeutic laparotomy were 89%, 75%, and 78%, respectively. Conclusion: Ultrasound quantification of hemoperitoneum by a simple scoring system may serve as a useful adjunct to traditional clinical parameters in predicting the need for subsequent laparotomy in pediatric patients. Prospective validation with a larger study is required. Key Words: Hemoperitoneum, Focused assessment with sonography for trauma, Laparotomy, Pediatric trauma patients. J Trauma. 2003;54: Submitted for publication August 14, Accepted for publication November 26, Copyright 2003 by Lippincott Williams & Wilkins, Inc. From the Departments of Surgery (A.W.O., M.G.M., M.B., N.N., J.M., S.M.C.) and Radiology (K.A.M.), University of Miami School of Medicine, Miami, Florida. This work was scheduled for a poster presentation at the 61st Annual Meeting of the American Association for the Surgery of Trauma, which was canceled because of the terrorist attacks of September 11, Address for reprints: Adrian W. Ong, MD, Department of Surgery, Division of Trauma/Critical Care, Allegheny General Hospital, 320 East North Avenue, Pittsburgh, PA 15212; aong@wpahs.org. DOI: /01.TA D Focused assessment with sonography for trauma (FAST) has emerged as a rapid, noninvasive, and accurate method of evaluating blunt abdominal trauma that can be easily used by nonradiologist clinicians. 1 3 Several studies have demonstrated a relationship between the amount of intraperitoneal blood quantified by radiologic studies and the need for laparotomy in adults after blunt trauma In the pediatric population, despite a higher success rate for nonoperative management of solid organ injuries, a similar correlation has been described. 11,12 Ultrasound quantification of hemoperitoneum using a simple scoring system has been shown to correlate with subsequent laparotomy on the basis of a prospective study from our institution. 7 This scoring system, however, has not been previously studied in pediatric patients. We hypothesized that the amount of free fluid quantified by this scoring system might also predict the need for laparotomy in pediatric patients. PATIENTS AND METHODS Our institutional review board approved this study with waiver of consent. A retrospective chart review was undertaken on all patients 15 years old or younger who were treated at our institution from January 1, 1997, to June 30, 1999, after blunt trauma. Every trauma resuscitation was supervised by an in-house trauma attending physician. However, frequently, cases were comanaged by the trauma surgeon and pediatric surgeon during the ongoing process of evaluating the patient. At the start of this study period, FAST was performed on pediatric patients at the discretion of the trauma attending physician. During the study period, the indication was expanded to include all pediatric patients meeting the stateapproved pediatric trauma alert criteria who were brought to our trauma center (Table 1). Computed tomographic (CT) scan of the abdomen and pelvis was indicated when the ultrasound examination was positive and the patient was stable enough to go to the CT scanning suite, when the patient presented with abdominal pain or tenderness, when retroperitoneal injury was suspected, when there was a pelvic fracture, or at the discretion of the attending physician. Where possible, American Association for the Surgery of Trauma grades were recorded for solid organ injuries. Unstable patients who responded transiently or minimally to fluids did not undergo CT scanning. Volume 54 Number 3 503

2 Table 1 Pediatric Trauma Alert Criteria 1. Those injured persons with anatomic and physiologic characteristics of a person 15 years of age or younger that present with one or more of the following criteria: Category 1 (any 1 meets TTC) Category 2 (any 2 meets TTC) Size 11 kg or less, or Broselow tape red or purple Airway Assisted or intubated or manual airway opening Consciousness Altered mental status or paralysis or coma or suspected spinal cord injury or loss of sensation Amnesia or any reliable history of loss of consciousness Circulation Faint or not palpable radial or femoral pulse or SBP 50 mm Hg Carotid or femoral pulse palpable, but radial or pedal pulses not; or SBP mm Hg, or capillary refill 3s Fracture Any open long bone fracture or multiple fracture sites Single closed long bone fracture anywhere or pelvic fracture Cutaneous Major tissue disruption or amputation or second- or third-degree burns to 10% TBSA or any penetrating injury to head, face, neck, or torso 2. Ejection from a motor vehicle 3. EMT or paramedic judgment TTC, trauma triage criteria; SBP, systolic blood pressure; TBSA, total body surface area; EMT, emergency medical technician. A positive ultrasound examination was defined as one showing free fluid with or without abnormalities suggesting solid organ injury. Five areas (Morison pouch, perisplenic area, right and left subphrenic spaces, and pelvis) were scanned. To quantify the amount of free fluid on ultrasound, the scoring system developed by McKenney et al. 5 was used. This ultrasound score (USS) was defined as the depth of the deepest pocket of fluid measured in centimeters in any one area plus one point for each additional area where fluid was seen. The depth was measured along a line drawn from the most anterior aspect of the collection to the posterior extent using either the longitudinal or transverse images (Fig. 1). For example, if fluid was present in three areas and the deepest pocket measured 4 cm, the USS would be 4 2 6, where 4 was the depth of the deepest pocket in centimeters and 2 was the number Fig. 1. Measurement of hemoperitoneum, illustrating measurement of the depth of fluid in the hepatorenal space in a patient. of additional spaces that were positive. A senior radiology resident or trauma radiology fellow read all the ultrasound images using real-time images. These images were also stored and reviewed later by the radiologists. A trauma radiology attending physician was present to also read the images between 7:00 AM to 12:00 midnight every day. All positive ultrasound examinations were assigned a USS prospectively at the time of acquisition of images. Positive ultrasound examinations that did not have the USS reported concomitantly were reviewed by an attending radiologist (K.A.M.) blinded to all patient details and outcome, and the USS was assigned retrospectively. A therapeutic laparotomy was broadly defined as one requiring repair of intra-abdominal organ injury or surgical control of active bleeding. Hypotension in pediatric patients adjusted for age was defined as a systolic blood pressure of 50 mm Hg in children from birth to 6 months of age, 70 mm Hg in infants, 80 mm Hg in preschool children, and 90 mm Hg in adolescents using the Advanced Trauma Life Support guidelines. A USS of 3.0 or greater was found to be highly predictive of laparotomy in adults on the basis of a prospective study, 7 and this value was used to stratify the current study population into two groups. Across the two groups, the Student s t test was used to compare continuous variables and the 2 test or Fisher s exact test (where appropriate) was used to compare dichotomous data. Significance was assessed by a value of p 0.05, using two-tailed statistical tests. The statistical software Analyze-it for Microsoft Excel Version (Analyze-it Software, Ltd., Leeds, United Kingdom) was used to perform these calculations. The ability to predict the need for therapeutic laparotomy on the basis of a USS of 3.0 versus a USS 3.0, the presence or absence of hypotension, and the presence or absence of tachycardia (pulse 100 beats/min) were assessed in patients with pos- 504 March 2003

3 Ultrasound Evaluation of Pediatric Trauma itive ultrasound examinations by calculating sensitivity, specificity, and accuracy. RESULTS A total of 743 patients 15 years old and younger were treated during the study period at our Level I trauma center. Among this group, 568 (76%) patients sustained blunt trauma. One hundred ninety-three of 568 (34%) patients with blunt trauma had FAST performed (Fig. 2). Of the 193 ultrasound examinations, 38 (19.6%) were positive. Of the positive ultrasound examinations, CT scans were normal in seven patients. The CT scans showed free fluid with no solid organ injury in 4, splenic injuries in 10, liver injuries in 5, pancreatic contusion in 1, small bowel thickening suggesting injury in 2, and pelvic fractures in 4 patients. Six patients did not undergo CT scanning and went straight to laparotomy from the resuscitation area because of hemodynamic instability. Of the negative ultrasound examinations, CT scans were positive for intra-abdominal injury and/or free fluid in 6, normal in 32, and not obtained in 111. Repeat ultrasound examinations were performed in only two patients, and both remained negative. One diagnostic peritoneal lavage was performed, and it was positive for intra-abdominal bleeding in a patient whose FAST examination was negative. Ten ultrasound examinations of the 38 were each assigned a USS retrospectively. One patient with a positive ultrasound examination was excluded from analysis because he died from head injury, and autopsy results were incomplete because of organ procurement. The remaining 37 patients formed the basis for further analysis. Fig. 2. Outcomes of 193 pediatric patients who underwent ultrasound examination after trauma. U/S, ultrasound; TL, therapeutic laparotomy; NTL, nontherapeutic laparotomy; NO, no laparotomy required. Twenty-two patients had a USS 3 and 15 had a USS 3. There were no differences between patients with a USS 3.0 and those with a USS 3.0 in terms of admission pulse (mean, 115 vs. 117/min; p 0.86), Glasgow Coma Scale score (mean, 12 vs. 11; p 0.75), or Injury Severity Score (mean, 17 vs. 16; p 0.79). Adjusted for age, only 3 patients of the 38 were considered to be hypotensive on initial blood pressure assessment. One of the 22 patients with a USS 3.0 was hypotensive compared with two of 15 with a USS 3.0. Hypotension was not associated with a USS 3(p 0.71, two-tailed Fisher s exact test). Arterial blood gases were not obtained in seven patients (six with a USS 3.0 and one with a USS of 14). Of the remaining 30 patients, mean base deficit was 4.4 for those with a USS 3.0 versus a mean of 6.0 for those with a USS 3.0 (p 0.46). There were 9 therapeutic and 2 nontherapeutic laparotomies among the 37 patients with positive ultrasound examinations. Of the 22 patients with a USS 3.0, 1 (4.5%) patient required laparotomy for perforated jejunum. Of the 15 patients with a USS 3.0, 8 (53%) required therapeutic laparotomy. Mean time to laparotomy for the nine therapeutic laparotomies was (SD) minutes. There was a significant association between a USS of 3.0 and subsequent therapeutic laparotomy (p 0.002, two-tailed Fisher s exact test). CT scanning of the abdomen/pelvis was not performed in five of the nine patients who underwent therapeutic laparotomy because of persistent hemodynamic instability during trauma resuscitation (Table 2). The nontherapeutic laparotomy that occurred in a patient with a USS of 1.0 found a nonbleeding splenic laceration. CT scan in this same patient revealed a small spleen laceration and abnormal bowel wall thickening. The patient had an admission base deficit of 9, with a pulse of 136 beats/min. The other nontherapeutic laparotomy occurred in a patient with a USS of 5.8 with transient hypotension in the resuscitation area. CT scanning was not performed, and a grade 2 splenic laceration with minimal bleeding was found during laparotomy. Of the remaining patients with positive ultrasound examinations who did not undergo laparotomy, none required delayed operation for missed injuries. One patient died of severe head injury and had a USS of 1.0. The autopsy of this patient revealed no intra-abdominal injury (Table 3). There were 155 patients with negative ultrasound examinations. One was excluded because of incomplete records attributable to organ procurement. Two of 154 (1.3%) required therapeutic laparotomy: one had a diagnostic peritoneal lavage that was positive after a negative ultrasound examination. At exploration, she had bleeding from the spleen and ascending colon mesentery. The other had retroperitoneal fluid and air on CT scan and was found to have a duodenal perforation at the second and third portions. No intra-abdominal free fluid was found at laparotomy. One patient with a negative ultrasound examination died from head injury: no intra-abdominal injury was found at autopsy. Volume 54 Number 3 505

4 Table 2 Details of Injuries in Patients with Positive Ultrasound Examinations Who Required Therapeutic Laparotomies Patient SBP (mm Hg) Pulse (beats/min) BD CT Scan USS Procedure Outcome Grade 4 SI 8.5 Splenectomy Grade 3 SI 6 Became hypotensive, laparotomy revealed actively bleeding spleen, splenorrhaphy performed ND 8 Splenectomy Death ND 8.5 Transverse colectomy for devitalized colon, control of mesenteric bleeding Free fluid 2.7 Repair of jejunal perforation, control of mesenteric bleeding ND 4 Splenectomy Death Free fluid 4 Control of mesenteric bleeding ND 6.5 Ligation of transected splenic vessels,* repair of gastric rupture ND 7 Splenectomy SBP, systolic blood pressure; BD, base deficit; USS, ultrasound score; SI, splenic injury; ND, not done. * Reoperated on 24 h later because of abdominal compartment syndrome. At reexploration, the spleen was necrotic and splenectomy was performed. Because of severe head injury. Related to multisystem organ failure. Table 3 Details of Injuries in Patients with Positive Ultrasound Examinations Who Did Not Require Therapeutic Laparotomy Patient USS CT Scan Findings Outcome 1 1 FF, no SOI Grade 3 SI 3 3 Grade 3 LI 4 7 Grade 3 SI, perirenal hematoma FF, no SOI 6 1 Pelvic fracture 7 1 Normal 8 1 Normal Normal 10 1 Grade 3 LI, FF, pelvic fracture 11 1 FF, pelvic fracture Grade 4 SI 13 1 Grade 2 SI 14 2 Pelvic fracture 15 6 Grade 1 SI FF, no SOI Normal 18 1 Grade 2 SI, abnormal bowel wall thickening NTL, alive Grade 1 SI 20 1 Normal 21 1 Grade 1 SI 22 1 FF, no SOI Death from head injury 23 1 Normal 24 3 Grade 3 LI, adrenal hematoma 25 8 Grade 3 LI, grade 1 SI, perirenal hematoma 26 1 Grade 1 SI 27 1 Grade 2 LI Not done NTL (grade 2 SI), alive 29 1 Normal Death from head injury* USS, ultrasound score; LI, liver injury; SI, splenic injury; SOI, solid organ injury; NTL, nontherapeutic laparotomy; FF, free fluid. * Excluded from further analysis because of incomplete autopsy (organ donation). 506 March 2003

5 Ultrasound Evaluation of Pediatric Trauma None of the remaining patients with negative ultrasound examinations required delayed laparotomy for missed injuries. The results for sensitivity, specificity, and accuracy for a USS 3.0, hypotension, and admission pulse 100 beats/ min in predicting need for therapeutic laparotomy among patients with positive ultrasound examinations are listed in Table 4. Base deficit as a predictor was not analyzed because admission arterial blood gases were not measured in seven patients. In this small sample size, a USS 3.0 was as accurate as hypotension in predicting the need for laparotomy. When the USS was 3.0, 8 of 15 (53%) patients required laparotomy. When there was a small amount of fluid (USS 3.0) or no fluid, only 3 of 176 (1.7%) patients required laparotomy. DISCUSSION Over the last decade, ultrasound has emerged as an accurate tool in the initial assessment of both adult and pediatric patients suspected of having blunt abdominal trauma Ultrasound estimation of the amount of intraabdominal fluid is fairly accurate. Matsumoto and Ohshiro 18 described a correlation between the number of spaces with intraperitoneal blood seen on ultrasound and the actual volume of blood intraoperatively on the basis of observations of trauma patients and patients with ectopic pregnancies and further validated this prospectively. The relation between the amount of hemoperitoneum and subsequent need for laparotomy after blunt trauma has been well documented. McKenney et al. 5 and Huang et al. 9 have devised scoring systems in adult patients using measurement of actual fluid depth that correlate with the need for laparotomy. In the study by McKenney et al., there were 46 patients with a USS 2, of whom 35 (76%) required laparotomy. Of 10 patients with scores 2, only 1 (10%) required laparotomy. A subsequent prospective study found that a USS of 3 had a sensitivity of 83%, a specificity of 87%, and an accuracy of 85% in predicting subsequent therapeutic laparotomy. 7 More importantly, they found that 31 of 36 (86%) normotensive patients with a USS of 3 who were initially observed subsequently deteriorated, requiring laparotomy. Similarly, Huang et al. 9 prospectively studied adult blunt trauma patients using their scoring system: patients with an ultrasound score 3 had a 96% chance of Table 4 Sensitivity, Specificity, and Accuracy of Various Parameters in Relation to Subsequent Laparotomy in Patients with Positive Ultrasound Examinations Sensitivity (%) Specificity (%) Accuracy (%) USS 3.0 8/9 (89) 21/28 (75) 29/37 (78) Initial hypotension 2/9 (22) 27/28 (96) 29/37 (78) Pulse 100 beats/min 6/9 (67) 9/28 (32) 15/37 (41) USS, ultrasound score. undergoing therapeutic laparotomy compared with a 37% chance when the score was 3. In the pediatric population, two authors have described a similar correlation between the amount of intraperitoneal fluid estimated preoperatively and subsequent laparotomy. Akgur et al. 11 categorized the amount of free fluid seen on ultrasound as small, moderate, or large. They found that 0 of 8 patients with a small amount of fluid required laparotomy, as compared with 2 of 15 with a moderate amount of fluid and 5 of 7 with a large amount of fluid. Taylor and Sivit 12 reviewed 259 children with peritoneal fluid on abdominal CT scan. Using similar definitions, 10% of patients with a small amount of fluid required laparotomy versus 23% of patients with a large amount of fluid. Using our scoring system, we found a similar association between the amount of free fluid and the likelihood of requiring laparotomy. Although this is not intended to replace traditional clinical parameters, quantification of free fluid serves as a useful adjunct in the evaluation of the pediatric blunt trauma patient, especially where vital signs and clinical examination are sometimes difficult to interpret in this age group. A more meaningful comparison in the younger age groups would be to quantify the amount of free fluid per unit body weight 18 since, for example, a USS of 3.0 in a 1-yearold may have different implications for surgical decisionmaking than a similar score in a 14-year-old. Validation of this concept will require a larger prospective study. Several other limitations exist. The study was conducted retrospectively, and therefore the outcome assessment (whether a laparotomy was therapeutic) was not blinded to the results of the test in question. Also, ultrasound examinations were performed in this population inconsistently at a low rate (34%), which suggests selection bias. Because of this reason, the true incidence of positive ultrasound examinations cannot be determined from this study and hence calculations of positive and negative predictive values will not be valid. At present, as mentioned above, ultrasound examination is performed in almost all pediatric blunt trauma patients evaluated at our trauma center. It is well documented that ultrasound may miss approximately 10% of solid organ injuries in adults 19 and up to 45% of injuries in children 20,21 in addition to retroperitoneal injuries. 22 Nevertheless, several studies have documented the efficacy of using ultrasound in the initial evaluation of the pediatric trauma patient ,23 What is not clearly established is whether or not we can stratify pediatric patients with positive ultrasound examinations in terms of risk of significant intra-abdominal bleeding requiring laparotomy, so that valuable time would not be wasted in obtaining further radiologic imaging or observation. On the basis of our preliminary observations, there appears to be a correlation between the amount of hemoperitoneum quantified by this ultrasound scoring system and subsequent therapeutic laparotomy in pediatric patients. However, the low positive predictive value suggests that this Volume 54 Number 3 507

6 scoring system should be used in conjunction with assessment of other clinical parameters in surgical decision-making in the emergency room. A larger prospective study is needed to validate this scoring system in the pediatric age group. REFERENCES 1. Rozycki GS, Ochsner MG, Jaffin JH, Champion HR. Prospective evaluation of surgeons use of ultrasound in the evaluation of trauma patients. J Trauma. 1993;34: Thourani VH, Pettitt BJ, Schmidt JA, Cooper WA, Rozycki GS. Validation of surgeon-performed emergency abdominal ultrasonography in pediatric trauma patients. J Pediatr Surg. 1998; 33: Shackford SR, Rogers FB, Osler TM, Trabulsy ME, Clauss DW, Vane DW. Focused abdominal sonogram for trauma: the learning curve of nonradiologist clinicians in detecting hemoperitoneum. J Trauma. 1999;46: Porter RS, Nester BA, Dalsey WC, et al. Use of ultrasound to determine need for therapeutic laparotomy in trauma patients. Ann Emerg Med. 1997;29: McKenney KL, McKenney MG, Nunez DB, McDowell L, Martin L. Interpreting the trauma ultrasound: observations in 62 positive cases. Emerg Radiol. 1996;3: Federle MP, Jeffrey RB. Hemoperitoneum studied by computed tomography. Radiology. 1983;148: McKenney KL, McKenney MG, Cohn SM, et al. Hemoperitoneum score helps determine need for therapeutic laparotomy. J Trauma. 2001;50: Goletti O, Ghiselli G, Lippolis PV, et al. The role of ultrasonography in blunt abdominal trauma: results in 250 consecutive cases. J Trauma. 1994;36: Huang MS, Liu M, Wu JK, Shih HC, Ko TJ, Lee CH. Ultrasonography for the evaluation of hemoperitoneum during resuscitation: a simple scoring system. J Trauma. 1994;36: Huang MS, Shih HC, Wu JK, et al. Urgent laparotomy versus emergency craniotomy for multiple trauma with head injury patients. J Trauma. 1995;38: Akgur FM, Tanyel FC, Akhan O, Buyukpamukcu N, Hicsonmez A. The place of ultrasonographic examination in the initial evaluation of children sustaining blunt abdominal trauma. J Pediatr Surg. 1993; 28: Taylor GA, Sivit CJ. Posttraumatic peritoneal fluid: is it a reliable indicator of intraabdominal injury in children? J Pediatr Surg. 1995; 30: McKenney MG, Martin L, Lentz K, et al. 1,000 consecutive ultrasounds for blunt trauma. J Trauma. 1996;40: McKenney MG, Lentz K, Nunez D, et al. Can ultrasound replace diagnostic peritoneal lavage in the assessment of blunt trauma? J Trauma. 1994;37: Partrick DA, Bensard DD, Moore EE, Terry SJ, Karrer FM. Ultrasound is an effective triage tool to evaluate blunt trauma in the pediatric population. J Trauma. 1998;45: Akgur FM, Aktug T, Olguner M, Kovanlikaya A, Hakguder G. Prospective study investigating routine usage of ultrasonography as the initial diagnostic modality for the evaluation of children sustaining blunt abdominal trauma. J Trauma. 1997;42: Katz S, Lazar L, Rathaus V, Erez I. Can ultrasonography replace computed tomography in the initial assessment of children with blunt abdominal trauma? J Pediatr Surg. 1996;31: Matsumoto H, Ohshiro K. Ultrasound in abdominal trauma. In: Machi J, Sigel B, eds. Ultrasound for Surgeons. New York: Ikagu- Shoin; 1997: Ochsner M, Knudson M, Pachter H, et al. Significance of minimal or no intraperitoneal fluid visible on CT scan associated with blunt liver and splenic injuries: a multicenter analysis. J Trauma. 2000; 49: Krupnick A, Teitelbaum D, Geiger J, et al. Use of abdominal ultrasonography to assess pediatric splenic trauma. Ann Surg. 1997; 225: Coley B, Mutabagani K, Martin L, et al. Focused abdominal sonography for trauma (FAST) in children with blunt abdominal trauma. J Trauma. 2000;48: Yoshii H, Sato M, Yamamoto S, et al. Usefulness and limitations of ultrasonography in the initial evaluation of blunt abdominal trauma. J Trauma. 1998;45: Patel J, Tepas J. The efficacy of focused abdominal sonography for trauma (FAST) as a screening tool in the assessment of injured children. J Pediatr Surg. 1999;34: March 2003