Original Article

Non-operative management of blunt hepatic injury in multiply injured adult patients

A.S. Al-Mulhim H.A.H. Mohammad*
Department of Surgery, King Fahad Hospital, Hofuf, P.O. Box 1164, Hofuf, Al-Hassa 31982 * Department of Surgery, Zagazig University, Egypt

Correspondence to: A.S. Al-Mulhim, Department of Surgery, King Fahad Hospital, Hofuf, P.O. Box 1164, Hofuf, Al-Hassa 31982 Email: abdu3939@yahoo.com

Introduction Clinical material

Results Discussion References

Keywords: Hepatic injury, non-operative
Surg J R Coll Surg Edinb Irel., 1 April 2003, 81-85

Background: Non-operative management of blunt liver trauma has now evolved into a common practice especially since abdominal CT has enabled a more precise evaluation of these patients. Clinical Material: Sixty-three patients, haemodynamically stable, were eligible for the study and enrolled into the protocol of non-operative management of blunt hepatic injury. Fifty-two (82.5%) patients were successfully managed non-operatively (non-operative group). The remaining 11 (17.5%) patients failed the non-operative management and underwent exploratory laparotomy (laparotomy group). Results: Patients managed non-operatively tended to be younger than patients managed operatively (p < 0.05). The mean values of ISS were 16.2 ± 6.1, 26.1 ± 8.5, p < 0.001, in the nonoperative and laparotomy groups, respectively. Stay in the ICU was significantly decreased in the nonoperative patients (p < 0.001). Patients who had a laparotomy significantly increased requirement for blood transfusion (p < 0.001). Six (9.5%) patients managed non-operatively developed complications; perihepatic collections were observed in two patients, an urinoma in one patient and chest infection in three patients. Perihepatic collections and urinoma were successfully drained percutaneously by CT guidance and no further treatment was required. The mortality rate of the entire series of patients was 4.8% (three patients); one death could be related to hepatic injury itself and the other two deaths were attributed to non-hepatic causes. No deaths occurred in the non-operative group. Conclusion: Non-operative management should be the initial approach to all patients with blunt liver injuries if haemodynamic stability can be ensured. When continued bleeding can be safely ruled out, a period of close monitoring in the ICU is warranted

INTRODUCTION
Until recently, operative management was the standard of care in treating liver injuries. Non-operative management of blunt liver trauma, however, has now evolved into a common practice, especially since abdominal computed tomography (CT) has enabled more precise evaluation of these patients.^1,2 The strategy for this arose from the observations that many liver injuries had stopped bleeding at the time of laparotomy and needed little or no intervention.³

Guidelines for non-operative management of blunt liver injuries have been produced and published.^4,5,6 Improvement in resuscitation, careful monitoring in the intensive care unit (ICU), coupled with advances in diagnostic radiology has helped to make non-operative policy possible and acceptable amongst surgeons.^7-9

The aim of our study was to determine the place of non-operative management in haemodynamically stable patients with blunt liver injuries, and appropriate CT information.

CLINICAL MATERIAL
Of the 91 patients, 28 (30.8%) were haemodynamically unstable and judged to be unsuitable for non-operative management, and underwent immediate exploratory laparotomy. The remaining 63 patients were enrolled

into the protocol of non-operative management and constituted the basic group of this study (Figure 1).

The decision to implement immediate surgical intervention was based on clinical findings. Exploratory laparotomy was performed on patients who had evidence of massive bleeding on presentation, persistent hypotension despite active resuscitation, transfusion requirements of more than half their blood volume (40ml/kg bodyweight) or associated hollow viscus injury.⁹

The criteria for non-operative management of blunt hepatic trauma, in our study, included the following: patients with liver injuries documented by CT who were haemodynamically stable, regardless of the extent of liver injury; hepatic-related transfusion limited to four units of blood; and absence of other abdominal injuries that would require exploratory laparotomy.¹⁰ Haemodynamic stability was defined as serial blood pressure and heart rate measurements within the appropriate age range and with adequate end-organ perfusion, either initially or after limited fluid resuscitation.^1,2,11

Initial resuscitation with Ringer’s lactate solution was carried out for all patients. Blood samples were taken and sent for routine haematology and biochemistry (including hepatic enzymes and serum amylase levels), blood typing, etc.

After resuscitation, all patients underwent CT scanning with 1cm intervals from the lower chest to the pubic symphysis. Double contrast CT was performed with intravenous and orally administered contrast to improve visualisation of intraabdominal structures. Additional investigations were obtained based on suspicion of other injuries.

According to the Organ Injury Scale (1994 revision) adopted by the American Association for the Surgery of Trauma, all hepatic and associated other abdominal organ injuries were graded based on the appearance of the CT scan.¹² The amount of haemoperitoneum was quantitated as follows: minimalperihepatic blood in the subphrenic or subhepatic space (approximately 500ml); moderate-perihepatic and blood along the paracolic gutter (less than 1L); and large-perihepatic and blood along the paracolic gutter and accumulating in the pelvic cavity (more than 1L).¹³

Once the diagnosis of hepatic injury was established by CT, all patients were admitted to the intensive care unit (ICU) for close observation and monitoring for a variable length of stay. The patients were monitored in the ICU with serial haemoglobin assessments (every four hours). When four serial assessments were stable, patients were transferred to a general surgical unit. On post-admission day three, patients were allowed to start quiet activity. On postadmission day seven, if haematology had not altered significantly, patients were discharged home after repeating the CT examination. Patients were instructed to undertake restricted activities at home for three weeks from the time of injury. A repeat of the abdominal CT was also considered in patients who had a decrease in haematocrit that may be related to liver injury to see if the injury had progressed. The patients were regularly evaluated on an outpatient’s basis; weekly for four weeks and then monthly for two months.

At three months post-injury, all the patients returned to the clinic for a final clinical assessment and CT examination. If they were clinically well at that time, they were allowed to return to full activity without restriction.

Failure of non-operative management was determined by clinical and laboratory evidence of ongoing haemorrhage or by the development of peritonitis. Continued haemorrhage was defined clinically by progressive abdominal distension and/or by the presence of tachycardia or hypotension refractory to fluid resuscitation. A fall in the haematocrit or persistent acidosis following resuscitation, were used as laboratory evidence of ongoing haemorrhage.⁷

The injury severity score (ISS), the Glasgow coma scale (GCS), transfusion requirements, liver transaminase levels, associated injuries, length of stay in the ICU, complications both in hospital and within three months of discharge and total mortality were analysed.^14,15 All the data were reported as the mean and standard deviation, and were compared using the student’s t test.

RESULTS
The present study included 63 patients with blunt hepatic injuries. Men constituted 65.1% of the population. Table 1 shows the mechanisms of injury: motor vehicle accident (49.2%) and fall from a height (25.4%) were the most common causes of blunt hepatic injuries in our study. Fifty-two (82.5%) of 63 patients, initially selected for nonoperative management, were managed successfully (non-operative group). The remaining 11 (17.5%) patients failed non-operative management and underwent exploratory laparotomy (laparotomy group). Table 2 compares the non-operative management and laparotomy groups. The mean values of age were 32.8 ± 14.6, 46.2 ± 15.9, in the non-operative and laparotomy groups, respectively. Patients managed nonoperatively tended to be younger than those treated operatively (p < 0.05).

A significant (p < 0.001) increase in the mean values of ISS, requirements for blood transfusion and ICU stay, was observed in the laparotomy compared with the non-operative patients.

The initial liver transaminase (SGPT) value was elevated in 56 (87.5%) patients of the entire population. The mean values of SGPT were 274.3 ± 111.5, 254.1 ± 139.3, in the non-operative and laparotomy groups, respectively; were not significantly different between the two groups (Table 2). Despite evidence of hepatic injury on admission by CT scanning, the initial serum SGPT values were normal in seven (11.1%) patients.

Eleven (17.5%) of our studied patients failed non-operative management, 1-7 days after admission. Seven of the 63 (11.1%) patients were explored for clinical evidence of ongoing haemorrhage with a falling haematocrit or a worsening appearance on CT scanning. The remaining four patients underwent exploratory laparotomy for the following indications, intestinal injury (two cases), pancreatic injury (one case) and splenic infarction (one case).

Associated and multiple injuries were observed in 58 (92.1%) patients. Associated intraabdominal injuries most commonly involved: spleen (14.3%); intestine (4.8%); kidney (7.9%) and pancreas (1.6%).

Extra-abdominal trauma most frequently resulted in thoracic injury (36.5%), head injury (26.98%), orthopaedic injuries (26.98%), soft tissue damage (19.05%), maxillofacial (7.9%) and spinal injury (4.76%).

In the current study, there were six complications in the non-operative group. Two patients developed perihepatic collections. The collection was infected in one patient. Percutaneous CT-guided drainage successfully dealt with these two collections, at seven and nine days after injury. No further treatment was required. A urinoma developed in one patient with renal injury and was successfully drained percutaneously. The remaining three patients developed chest infections and responded to antibiotic therapy.

The mortality rate of the entire series of patients was 4.76% (three patients). There was no mortality in patients who underwent successful non-operative management. All of the three deaths occurred in the laparotomy group: two deaths were attributed to non-hepatic causes (severe head and thoracic injury). Only one death could be related to hepatic injury itself. Analysis of this case revealed that the patient had cirrhosis and became hypotensive 30 hours after the injury. Despite active effort to stop the bleeding, this could not be controlled.

Figure 1: Protocol used for the enrolement of patients into the study

DISCUSSION
Non-operative management of blunt hepatic trauma has become an accepted treatment in recent years.⁹ The primary requirement of this policy is haemodynamic stability.^16,17 Any physiological instability after initial resuscitation mandates laparotomy and current guidelines recommend surgery if the patient requires replacement of more than half of his/her blood volume.^1-2,5,6,9

Until continued bleeding can be ruled out, a period of intensive monitoring is necessary and an experienced surgical team must follow the patient closely.

In our series, non-operative management of blunt liver injuries was successful in 82.5% of patients. This result is comparable to findings reported by Croce et al (1993).¹⁷ The latter reported a success rate of up to 90% in their patients. Other studies have documented successful non-operative management in 85% to 100% of their patients.^1,18-20

The hepatic injury grading scale proposed by the American Association for the Surgery of Trauma can serve as a yardstick so that data from different centres can be standardised and compared.²¹ However, its use as a decision-making tool has been called into question by many authors.^9,22,23 Moon and Federle (1983) found that the need for surgery for patients with blunt hepatic injury was more closely correlated with the amount of haemoperitoneum than the size of the hepatic laceration.¹⁸ In our study, neither the grade of hepatic injury nor the amount of haemoperitoneum could predict the need for laparotomy. This result was in accordance with the previous results of Hiatt et al (1990).²⁴ They suggested that the decision for laparotomy should be determined by the patient’s overall clinical picture, not by the “exact nature of liver injury”. In our series there were limited numbers of high-grade liver injuries. Patients with grade V and VI liver injuries were lacking in our study. However, a grade VI injury has a very high mortality rate and patients with this type of injury are unlikely to be in a haemodynamic stable condition.

The initial serum SGPT values were elevated in 56 (87.5%) patients studied. The mean values of SGPT were 274.3 ± 111.5, 254 ± 139.3, p > 0.05, in the non-operative and laparotomy groups, respectively. Despite evidence of hepatic injury on the admission CT, the initial serum SGPT values were normal in seven (11%) patients. The initial serum SGPT value neither excluded the presence of hepatic injury nor predicted the need for laparotomy, in our patients.

The severity of injury, in our study, was measured by the ISS.^14,15 A highly significant (p < 0.001) increase in the mean value of ISS was observed in the laparotomy group, when compared with the non-operative group. Despite this result, ISS was of limited clinical value when considering any individual case. Judgement to operate or not, in our patients, was dependent mainly on the clinical parameters rather than numerical values of the ISS.

Critics of non-operative management cite the potential risk for missing other serious intraabdominal injuries and increasing utilisation of blood transfusion, with all its attendant risks.9 In the current study, there were no missed injuries. The mean values for blood transfusion requirements were 37.3 ± 4.1, 19.1 ± 6.9, p < 0.001, in the laparotomy and non-operative groups, respectively. The significant increase in requirement for blood transfusion in the laparotomy patients could not be attributed to the severity of the hepatic injury per se, as there was no significant difference in the mean values of grading of hepatic injuries between the two study groups. The significant (p < 0.001) increase in the mean value of ISS, in the laparotomy group of patients, may explain the increased requirements for blood transfusion in this group. Patients are more likely to receive more blood transfusions if more than one organ is affected.

Advancing age has been implicated to influence the success rate of non-operative management of blunt abdominal trauma.^9,22,23 Our data show that patients managed non-operatively tended to be younger than patients who underwent surgery (p < 0.05). The majority of our patients who were managed non-operatively were less than 30 years, while 45.5% of the laparotomy group of patients were more than 50 years of age. Five of eight patients who were older than 50 years, failed non-operative management; two of them were responsible for 66.6% of the total mortality in this series. Intolerance to severe haemodynamic instability may explain the increased incidence of failure in these patients.

The non-operative management of blunt hepatic injury, when accompanied by head injury, remains problematic and controversial. Head injury may complicate the management of these patients in two ways; firstly, because of an altered level of consciousness, physical examination may become unreliable, secondly, the potential release of tissue thromboplastin from head trauma (although experimentally unproven) may lower the success rate with non-operative management by impairing the coagulation profile.^25-27 Head injuries were observed in 17 (26.9%) patients of our studied population. Analysis of this subgroup revealed that five patients (29.4%) failed non-operative management and their mean GCS value was 10.6 ± 2.96. As the mean value of GCS in nonoperative patients insignificantly differed from that of the laparotomy group of patients (p > 0.05), the failure of non-operative management may be related to factors other than head injury. Patients who failed non-operative management were older with significantly high ISS (p < 0.001). The existence of these two factors may explain the incidence of failure in these patients.

In conclusion, this series has reaffirmed the validity of a selective policy of non-operative management of blunt liver injury. The most important decision facing the surgeon is to distinguish between patients who need an operation and those in whom an operation may be avoided. Haemodynamically unstable patients need an operation whilst those patients who are stable on arrival or those who respond quickly to fluid resuscitation can be managed non-operatively with a high success rate (82.5%), thereby, reducing the number of unnecessary laparatomies.

REFERENCES
1. Meredith JW, Young JS, Bowling J and Roboussin D. Non-operative management of blunt hepatic trauma: the exception or the rule. J Trauma 1994; 36: 529-35.
2. Pachter HL, Hofstetter SR. The current status of non-operative management of adult hepatic injuries. Am J Surg 1995; 169: 442-54.
3. Fabian TC, Croce MA, Stanford GG, Payne LW, Mangiante EC, Voeller GR, Kudsk KA. Factors affecting morbidity following hepatic trauma: a prospective analysis of 482 injuries. Ann Surg 1991; 213: 540-44.
4. Riad C, David C, Keith G and Gene AG. Evolution in the management of the complex liver injury at a level I trauma center. J Trauma 1998; 45: 1079-82.
5. Holland MJ, and Little JM. Nonoperative management of blunt liver injuries. Br J Surg 1992; 78: 968-74.
6. Fang JF, Chen RJ, Lin BC, Hsu YB, Kao JL and Chen MF. Blunt hepatic injury: minimal intervention is the policy of treatment. J Trauma 2000; 49: 722-28.
7. Lauren PA, Frederick BR and Steven RS. Selective nonoperative management of liver and spleen injuries in neurologically impaired adult patients. Arch Surg 1996; 131: 309-14.
8. David AP, Denis DB, Ernest EE and Frederick MK. Non-operative Management of Solid Organ Injuries in Children Results in Decreased Blood Utilization. J Pediatr 1999; 34: 1695-99.
9. Carrillo EH, Spain DA, Wohltmann CD, Schmieg RE, Boaz PW, Miller FB, Richardson JD. Interventional techniques are useful adjuncts in nonoperative management of hepatic injuries. J Trauma 1999; 46: 619-24.
10. Carrillo EH, Platz A, Miller FB, Richardson JD and Polk HC. Non-operative management of blunt hepatic trauma. Br J Surg 1998; 85: 461-68.
11. Durham RM, Buckley J, Keegan M, Fravell S, Shapiro MJ, Mazuski J. Management of blunt hepatic injuries. Am J Surg 1992; 164: 477-51.
12. Moore EE, Cogbill TH, Malangeni MA, Jurkovi GJ, Shackford SR, Champion HR, McAninch JW. Organ injury scaling. Surg Clin North Am 1995; 75: 293-97.
13. Yih-Gang C, Mu-Shun H and Jer-Ming L Non-operative management for extensive hepatic and splenic injuries with significant hemoperitoneum in adults. J Trauma 1998; 45: 360-65.
14. Baker SP and Neil B. The injury severity score. J Trauma 1972; 16: 882-85.
15. Champion HR, Sacco WJ, Copes WS, Gann DS, Gennarelli TA and Flanagan ME. a revision of the trauma score. J Trauma 1989; 29: 623-26.
16. Boone DC, Federle M, Billiar TR et al. Evolution of management of major hepatic trauma: identification of patterns of injury. J Trauma 1995; 39: 344-50.
17. Croce MA, Fabian TC and Menke PG. Non-operative management of blunt hepatic trauma is the treatment of choice for haemodynamicallystable patients. Results of a prospective trial. Ann Surg 1998; 221: 744-55.
18. Moon KL and Federle MP. Computed tomography in hepatic trauma. Am J Radiol 1983; 141: 309-17.
19. Meyer AA, Crass RA, Lim RC, Jeffrey RB, Federle MP and Trunkey DD. Selective nonoperative management of blunt liver injury using Computed Tomography. Arch Surg 1985; 120: 550-54.
20. Knudson MM, Lim RC, Oakes DD and Jeffrery RB. Nonoperative management of blunt liver injuries in adults: the need for continued surveillance. J Trauma 1990; 30: 1494-500.
21. Harold S, Beth AS, Larry MJ, Roger RB. Barbara AL, Jorge RV, Clyde EM, Robert TJ and Andrew HM. Non-operative management of blunt hepatic injuries: safe at any grade. J Trauma 1994; 37: 616-21. 
22. Smith JB, Wengrovitz MA, and Delong BS. Prospective validation of criteria, including age, for safe, non-surgical management of the ruptured spleen. J Trauma 1992; 33: 385-89.
23. Charles DG, Ralph LW, Robert LS and Charles JM. Non-operative management of blunt splenic injury in adults: age over 55 years as a powerful indicator for failure. Am J Surg 1996; 183: 133-39.
24. Hiatt JR, Harrier HD, Keonig BV et al. Non-operative management of major blunt hepatic injury with hemoperitoneum. Arch Surg 1990; 125: 101-07.
25. Sande JJ, Veltkamp JJ, Mussert RJ, Vielvoye GJ. Hemostasis and Computerized tomography in head injury. J Neurosurg 1981; 55: 718-26.
26. Gando S, Tedo I and Kubota M. Post-trauma coagulation and fibrinolysis. Crit Care Med 1992; 20: 594-603.
27. Pietropaoli JA, Rogers FB, Shackford SR, Wald WL, Schmoker JD and Zhuang J. The deleterious effects of intraoperative hypotension: an outcome in patients with severe head injuries. J trauma 1992; 33: 403-07.

Copyright: 12 March 2003