**4.3 Sepsis, organ failure**

Compared to the other groups, increased early mortality in the liver group did not lead to a simultaneous reduction in late mortality. Patients with a liver injury showed - apart from the patients with head injuries – an average late mortality of 7.8%. One cause for the increased late mortality in comparison with patients with no liver injury is possibly the high sepsis rate (19.9%), if the first 24 hours were survived.

The increased sepsis rate in the liver group is also reflected in the frequency of organ failure (OF 48.6%) and multi-organ failure (MOF 33.3%). Compared to patients with abdominal injuries with no severe liver trauma, all three characteristics are significantly more fully developed (sepsis 11.0%, OF 33.2%, MOF 16.6%). Patients from the control group also showed a significantly decreased incidence for sepsis and multi-organ failure.

The frequency of a laparotomy is reduced from 71.6% (before 2001) to 60.4% (from 2001). Remarkably, mortality is reduced in the same period from 35.5 to 33.1%. The ISS is almost identical with 39.7 vs. 38.8.

#### **4.4 Severity adjustment**

308 Liver Transplantation – Basic Issues

Further analysis of these differences between abdominal trauma group and the control group showed that the higher mortality in the control group is explained by the high mortality of the accompanying head injuries. Thus, a subgroup analysis shows that of the 9,574 trauma patients in the control group, 2,160 patients had suffered a relevant head injury (AIS >3). In this subgroup, mortality even reached 32.8%. The investigation of early mortality showed that 27.3% of patients in the liver trauma group died within the first 24

Compared to patients with non-liver abdominal injuries, patients with severe liver trauma clearly had a greater need for blood transfusions (67.0% vs. 48.0%). The high blood loss in the liver group is correlated with the blood pressure pattern in both the preclinical and emergency room (ER) phases. Initial blood pressure was ≤90mmHg preclinically in 36.4% of the liver group and 30.0% of the abdomen group. Both groups are clearly above the rate in the control group (22.0%). Blood pressure in the liver group could not be raised in any definitive way during initial clinical care (ER phase in contrast to the abdomen group (RR <90mmHg, liver: 32.2% with delta RR 4.2 mmHG; abdomen: 18.2% with delta RR: 11.2mmHG). In the ER, an initial hemoglobin content of less than 8g/dl was much more frequent in the liver group with 38.1% than in the abdomen group with 16.9% and the control group with 13.9%. Analogous to this, the average amount of transfused erythrocyte concentrate (EC) until admission to the intensive care unit was much higher in the group of patients with liver injury (8.6 units) compared to the abdomen group (4.5 units) and the

Patients who fulfilled the criteria of a massive transfusion (number of transfused EC >10

Given that the average number of ECs and the average ISS in both groups of liver and abdominal trauma were almost the same (liver: 20.9 EC, ISS 39.2; abdomen 19.9 EC, ISS 38.5), the possible measured variable of an unequal EC quantity was leveled out. Thus, the high total mortality in the liver group (55.8%) compared to the abdomen group (36.5%) cannot be explained by the number of ECs. The same applies to the increased MOF (96.0%

Compared to the other groups, increased early mortality in the liver group did not lead to a simultaneous reduction in late mortality. Patients with a liver injury showed - apart from the patients with head injuries – an average late mortality of 7.8%. One cause for the increased late mortality in comparison with patients with no liver injury is possibly the high

The increased sepsis rate in the liver group is also reflected in the frequency of organ failure (OF 48.6%) and multi-organ failure (MOF 33.3%). Compared to patients with abdominal injuries with no severe liver trauma, all three characteristics are significantly more fully developed (sepsis 11.0%, OF 33.2%, MOF 16.6%). Patients from the control group also

showed a significantly decreased incidence for sepsis and multi-organ failure.

hours, while this rate was only 6.6% in the non-liver abdominal group.

**4.2 Blood transfusion** 

control group (2.1 units).

**4.3 Sepsis, organ failure** 

were filtered out of the liver and abdomen groups.

vs. 60.0%) and sepsis rate (72.0% vs. 36.0%) of the survivors.

sepsis rate (19.9%), if the first 24 hours were survived.

Adjusting for severity with the RISC Score shows that patients with liver trauma die significantly more frequently than expected. The 33.0% mortality observed (95.0% confidence interval 27.6 – 38.4) offsets a prognostic mortality rate of only 23.4%. In the other two groups of injuries, prognosticated mortality hardly deviates at all from the observed mortality. These results could imply that the resuscitation and/or operative management was suboptimal. However, this is not true. Liver trauma is rather underestimated regarding the expected prognostically impact and shows significantly worse mortality rates than in patients without liver injuries. Therefore, severe liver injury should be judged more critically with respect to mortality than the remaining abdominal injuries, with which the RISC prognosis illustrates actual mortality very well.

#### **4.5 Transplantation after blunt trauma to the liver**

Six LT were performed in 4 patients with acute liver injury (2 patients were re-transplanted). The demographics and the clinical presentation of these patients are reported individual. There were 3 men and 1 woman, ranging in age from 36 to 50 years (mean and median, 42 years and 41 years, respectively). All patients had uncontrollable liver injuries caused by motor vehicle accidents. After a median (range) follow-up of 32.95 months (10.3-55.6), 2 out of 4 patients are still alive. Half- and 4-year patient survival rates are 50% and 25% with a corresponding graft survival of 25%, respectively.
