Blunt
Hepatic Trauma
Nonoperative Management in Adults John A.
Federico, MD; William R. Horner, MD; David E. Clark, MD; Robert J. Isler, MD
\s=b\ Fifty-six adults were identified with blunt hepatic trauma. Sixteen patients (29%) were treated successfully with nonoperative management. There were no delayed laparotomies or deaths in the nonoperative group. All patients required close observation. Eight (50%) of 16 patients required transfusion of no more than 3 U of packed red blood cells. There were two significant late complications requiring readmission. Four (25%) of 16 patients had complex fractures, yet were treated successfully without laparotomy. The quantity of fluid in the abdomen, as estimated by computed tomography, did not predict failure of treatment. Nonoperative management of blunt hepatic injuries is a useful alternative in selected patients when the course of therapy is based on the hemodynamic stability of the patient and supported by
computed tomographic findings. (Arch Surg. 1990;125:905-909)
nonoperative management Successful adults, reported by others,
of blunt splenic inju¬ as our institution and several ries in has led to a consideration of similar management principles for adults with blunt hepatic injuries.'3 Historical¬ ly, the presence of a liver injury had been an indication for operative intervention.4 With the advent of high-quality com¬ puted tomographic imaging (CT) and interpretation, nonop¬ erative management in selected patients in stable condition has become a reasonable consideration.5"8 Concerns about nonoperative management of blunt liver injuries relate primarily to the incidence of delayed bleeding or infection. In addition, the ideal patient selection criteria are still undefined. This report is an analysis of our results of using nonoperative management of blunt hepatic injuries during the past 5 years. PATIENTS AND METHODS 1983 through September 1988, 56 consecutive adult with blunt hepatic trauma were examined at the Maine Medical Center, Portland (Fig 1). The age range was set at 16 to 69 years. Thirty-eight of these patients were in hemodynamically unsta¬ ble condition at the initial assessment and, on the basis of positive diagnostic peritoneal lavage, underwent immediate surgery. The remaining 18 patients were in stable condition and underwent CT From
patients
July
Accepted for publication April 16, 1990. From the Departments of Surgery (Drs Federico, Clark, and Homer) and Radiology (Dr Isler), Maine Medical Center, Portland. Read before the 70th Annual Meeting of the New England Surgical Society, Bretton Woods, NH, September 22,1990. Reprint requests to the Maine Medical Center, Department of Surgery,
22 Bramhall St, Portland, ME 04102 (Dr Clark).
evaluations. With two exceptions, these patients were treated nonoperatively and form the basis of the present study. A retrospective review of the charts of all patients with hepatic trauma was carried out to determine the mechanism of injury, associ¬ ated injuries, injury severity scores, transfusion requirements, hos¬ pital courses, and outcomes. In the subset of patients who were treated nonoperatively, particular attention was paid to early and late complications, hospital course, and CT findings. Computed tomo¬ graphic scans were performed on GE9800 Quick and Technicare 2060 scanners (General Electric, Milwaukee, Wis) following the adminis¬ tration of oral contrast medium and during infusion of intravenous contrast medium. Images were obtained at 2-second scan times with a 1-cm-thick slice from the liver dome to the iliac crest, and a 1.5-cmthick slice from the iliac crest to the pelvic floor. All scans were reviewed by a senior surgical resident, a staff radiologist, and a staff surgical attending physician. Each image was evaluated for the pres¬ ence of parenchymal lacerations, intrahepatic hematomas, anatomic location of injury, estimated volume of intraperitoneal fluid, and associated visceral injury. The selection of patients for nonoperative therapy was based on the patient's hemodynamic findings and physical examination results. A pulse of less than 110 beats per minute and a systolic blood pressure of greater than 90 mm Hg, after initial fluid resuscitation, qualified a patient's condition as hemodynamically stable. Roentgenographic criteria, such as type of injury, location of injury, or amount of intraabdominal fluid, did not determine operative vs nonoperative thera¬ py, with two exceptions. The first was a patient in stable condition whose scan showed more than 500 mL of intraperitoneal fluid without apparent solid visceral injury. Diagnostic peritoneal lavage con¬ firmed the presence of blood, and at laparotomy a bleeding injury of the lateral segment of the left hepatic lobe was found. The second patient had a laceration to the liver in the region of the right lobe with free fluid. Although the patient's condition remained hemodynami¬ cally stable, at the discretion of the attending surgeon the patient was taken to the operating room where a nonbleeding laceration of the liver was identified and 400 mL of free blood was evacuated. Other¬ wise, clinical variables were followed up closely and formed the prime determinants of treatment modality. Nonoperative treatment was employed only in patients who were awake, in hemodynamically stable condition, and without generalized peritoneal irritation. When a decision for nonoperative management was made, patients were transferred to a special care or critical care step-down unit for close observation. Pulse and blood pressure were monitored. Fre¬ quent abdominal examinations were performed and serial hematocrit determined. As their clinical course improved, patients were trans¬ ferred to a surgical ward but were restricted to minimal activity. Most patients received initial follow-up CT scans within 3 to 7 days. Subsequent outpatient scans were obtained from 1 to 11 months later. Follow-up clinical data were obtained on all nonoperative patients by a review of clinic charts and through telephone interviews. A statisti¬ cian performed statistical analysis using a two-tailed Student's t test.
Downloaded From: http://archsurg.jamanetwork.com/ by a Georgetown University Medical Center User on 05/23/2015
—
Patients 16 to 69
(Aged
To
.
y)
Score
No. Of
No. Of
Nonoperative Patients
(n 16)
Operative Patients (n 40)
0
17
0
1
=
=
60
Operating Room
2 to
16
Operating
Nonoperative
Room
34 Survived
Injury Severity
\
S
38 Unstable Conditions
+
Injury Severity Scores and Blood Transfusions*
Table 1
56
4
Died
2 Survived
The mean injury severity score was 23.9 for nonoperative patients and 36.7 for operative patients (P
Hepatic Trauma
Nonoperative Management in Adults John A.
Federico, MD; William R. Horner, MD; David E. Clark, MD; Robert J. Isler, MD
\s=b\ Fifty-six adults were identified with blunt hepatic trauma. Sixteen patients (29%) were treated successfully with nonoperative management. There were no delayed laparotomies or deaths in the nonoperative group. All patients required close observation. Eight (50%) of 16 patients required transfusion of no more than 3 U of packed red blood cells. There were two significant late complications requiring readmission. Four (25%) of 16 patients had complex fractures, yet were treated successfully without laparotomy. The quantity of fluid in the abdomen, as estimated by computed tomography, did not predict failure of treatment. Nonoperative management of blunt hepatic injuries is a useful alternative in selected patients when the course of therapy is based on the hemodynamic stability of the patient and supported by
computed tomographic findings. (Arch Surg. 1990;125:905-909)
nonoperative management Successful adults, reported by others,
of blunt splenic inju¬ as our institution and several ries in has led to a consideration of similar management principles for adults with blunt hepatic injuries.'3 Historical¬ ly, the presence of a liver injury had been an indication for operative intervention.4 With the advent of high-quality com¬ puted tomographic imaging (CT) and interpretation, nonop¬ erative management in selected patients in stable condition has become a reasonable consideration.5"8 Concerns about nonoperative management of blunt liver injuries relate primarily to the incidence of delayed bleeding or infection. In addition, the ideal patient selection criteria are still undefined. This report is an analysis of our results of using nonoperative management of blunt hepatic injuries during the past 5 years. PATIENTS AND METHODS 1983 through September 1988, 56 consecutive adult with blunt hepatic trauma were examined at the Maine Medical Center, Portland (Fig 1). The age range was set at 16 to 69 years. Thirty-eight of these patients were in hemodynamically unsta¬ ble condition at the initial assessment and, on the basis of positive diagnostic peritoneal lavage, underwent immediate surgery. The remaining 18 patients were in stable condition and underwent CT From
patients
July
Accepted for publication April 16, 1990. From the Departments of Surgery (Drs Federico, Clark, and Homer) and Radiology (Dr Isler), Maine Medical Center, Portland. Read before the 70th Annual Meeting of the New England Surgical Society, Bretton Woods, NH, September 22,1990. Reprint requests to the Maine Medical Center, Department of Surgery,
22 Bramhall St, Portland, ME 04102 (Dr Clark).
evaluations. With two exceptions, these patients were treated nonoperatively and form the basis of the present study. A retrospective review of the charts of all patients with hepatic trauma was carried out to determine the mechanism of injury, associ¬ ated injuries, injury severity scores, transfusion requirements, hos¬ pital courses, and outcomes. In the subset of patients who were treated nonoperatively, particular attention was paid to early and late complications, hospital course, and CT findings. Computed tomo¬ graphic scans were performed on GE9800 Quick and Technicare 2060 scanners (General Electric, Milwaukee, Wis) following the adminis¬ tration of oral contrast medium and during infusion of intravenous contrast medium. Images were obtained at 2-second scan times with a 1-cm-thick slice from the liver dome to the iliac crest, and a 1.5-cmthick slice from the iliac crest to the pelvic floor. All scans were reviewed by a senior surgical resident, a staff radiologist, and a staff surgical attending physician. Each image was evaluated for the pres¬ ence of parenchymal lacerations, intrahepatic hematomas, anatomic location of injury, estimated volume of intraperitoneal fluid, and associated visceral injury. The selection of patients for nonoperative therapy was based on the patient's hemodynamic findings and physical examination results. A pulse of less than 110 beats per minute and a systolic blood pressure of greater than 90 mm Hg, after initial fluid resuscitation, qualified a patient's condition as hemodynamically stable. Roentgenographic criteria, such as type of injury, location of injury, or amount of intraabdominal fluid, did not determine operative vs nonoperative thera¬ py, with two exceptions. The first was a patient in stable condition whose scan showed more than 500 mL of intraperitoneal fluid without apparent solid visceral injury. Diagnostic peritoneal lavage con¬ firmed the presence of blood, and at laparotomy a bleeding injury of the lateral segment of the left hepatic lobe was found. The second patient had a laceration to the liver in the region of the right lobe with free fluid. Although the patient's condition remained hemodynami¬ cally stable, at the discretion of the attending surgeon the patient was taken to the operating room where a nonbleeding laceration of the liver was identified and 400 mL of free blood was evacuated. Other¬ wise, clinical variables were followed up closely and formed the prime determinants of treatment modality. Nonoperative treatment was employed only in patients who were awake, in hemodynamically stable condition, and without generalized peritoneal irritation. When a decision for nonoperative management was made, patients were transferred to a special care or critical care step-down unit for close observation. Pulse and blood pressure were monitored. Fre¬ quent abdominal examinations were performed and serial hematocrit determined. As their clinical course improved, patients were trans¬ ferred to a surgical ward but were restricted to minimal activity. Most patients received initial follow-up CT scans within 3 to 7 days. Subsequent outpatient scans were obtained from 1 to 11 months later. Follow-up clinical data were obtained on all nonoperative patients by a review of clinic charts and through telephone interviews. A statisti¬ cian performed statistical analysis using a two-tailed Student's t test.
Downloaded From: http://archsurg.jamanetwork.com/ by a Georgetown University Medical Center User on 05/23/2015
—
Patients 16 to 69
(Aged
To
.
y)
Score
No. Of
No. Of
Nonoperative Patients
(n 16)
Operative Patients (n 40)
0
17
0
1
=
=
60
Operating Room
2 to
16
Operating
Nonoperative
Room
34 Survived
Injury Severity
\
S
38 Unstable Conditions
+
Injury Severity Scores and Blood Transfusions*
Table 1
56
4
Died
2 Survived
The mean injury severity score was 23.9 for nonoperative patients and 36.7 for operative patients (P
