Nonoperative Management of Hepatic Trauma
	Date: Fri, 26 Jul 1996 19:59:58 From: Karim Brohi [karim@trauma.org] I'd just like an idea of list members opinions on some aspects of non-operative management of liver trauma. Top of the list of criteria for non-operative management is haemodynamic stability. No paper I have read actually defines haemodynamic instability. So is it : A. Patient cardiovascularly normal on arrival B. Patient cardiovascularly normal after initial crystalloid bolus C. Patient cardiovascularly stable after bolus of crystalloid and some blood (how much?) D. A gut feeling that you'll be able to keep this person alive through CT. - ie requiring fluids constantly but not dramatically exsanguinating E. None of the above! Secondly, by virtue of our trauma system, patients with isolated intra-abdominal injuries are taken to the nearest teaching hospital. Patients with multiple injuries (especially neuro) are brought back to us by helicopter (excluding our catchment area of course). As such they often arrive anaesthetised & ventilated (head injuries). Can you non-operatively manage safely someone who is unconcious? How do you exclude the enteric injury (or can you?) If so what imaging/laboratory support would you require serially. We have many patients whose abdomens would meet non-operative criteria if they were awake. 'No more than 2 hepatic related transfusions' is also quoted. How do you know, in the multiple injured patient, which of your transfusions was hepatic???!!
	Date: Fri, 26 Jul 1996 16:37:39 From: Lisa S. Dresner [71211.2533@compuserve.com] We manage hemodynamically stable patients with blunt liver trauma (and splenic trauma) with non operative techniques. We define hemodynamically stable as ones who respond to initial resuscitation with 1-2 liters of crystalloid with normalized pulse rate, etc.While some of these patients do require blood transfusions during their hospital stay, as long as they do not appear to have evidence of ongoing blood loss (persistant metabolic acidosis, ongoing volume requirments, etc) we observe them. In the early days we did a lot of angiography on a those with greater than grade 2-3 on CT or with concommittent splenic injury but found that only those liver injuries with evidence of ongoing blood loss every required any intervention.
	Date: Sat, 27 Jul 1996 20:44:19 From: Aviel Roy-Shapira, M.D. [avir@bgumail.bgu.ac.il] Karim raises some interesting points. I too would like to see the results of this survey. >A. Patient cardiovascularly normal on arrival >B. Patient cardiovascularly normal after initial crystalloid bolus I consider either A or B as signs of cardiovascular stability. >C. Patient cardiovascularly stable after bolus of crystalloid > and some blood (how much?) >D. A gut feeling that you'll be able to keep this person alive > through CT. - ie requiring fluids constantly but > not dramatically exsanguinating >E. None of the above! > >Secondly, by virtue of our trauma system, patients with isolated >intra-abdominal injuries are taken to the nearest teaching hospital. >Patients with multiple injuries (especially neuro) are brought back >to us by helicopter (excluding our catchment area of course). >As such they often arrive anaesthetised & ventilated (head injuries). > >Can you non-operatively manage safely someone who is unconcious? >How do you exclude the enteric injury (or can you?) These two questions are clearly related. If the only concern were bleeding, head injury would not exclude non-op management, since the hemodynamic effects of head traua and bleeding are in different directions. The problem is that the signs of enteric injury on CT are subtle, and it is difficult to judge whether the fluid seen on CT is blood or enteric contents, or a mixture of the too. My tendency is therefore to explore these patients. However, it is a matter of judgement. If the CT shows no fluid, except around the liver, with grade I injury, and the patient is CV normal (not the same as stable, as Karim points out), I think that non-op management is justified.
	Date: Sat, 27 Jul 1996 10:57:40 From: Eric Frykberg M.D. [ERF.TRAUMAONE@mail.health.ufl.edu] We also watch hepatic injuries nonoperatively based solely on hemodynamic stability, meaning no evidence of ongoing blood loss regardless of how much crystalloid or blood is required to replete the initial loss, just as Dr. Dresner described. Since this management based solely on the patient's clinical condition clearly works, what the liver looks like on CT is irrelevant, a point much of the country (the world?) has yet to realize, much to the delight of our radiology departments. Beyond one initial CT to document that it is the liver that is injured, any further studies while the patient remains unchanged are superfluous. The liver injury grading system should not be used to make treatment decisions, since it is the patient's condition that alone determines the necessary course. If a stable patient has a Grade 4 or 5 injury, we will observe, and if a Grade 1 injury is unstable we will operate. Even the practice of ordering a 2nd CT to document "healing" before discharge of a stable patient has been shown unnecessary (not surprisingly) in a paper by David Ciraulo and Len Jacobs presented at EAST in Jan. 1996, abstract in Dec 1995 issue of Journal of Trauma. I am also unaware of any data in which decompensation requiring surgery in an initially stable patient being observed has any correlation with the grade of liver injury, even though that appears to be the presumption behind the practice of multiple CT's described by Dr. Smith. Even if such a correlation exists, however, the patient, and not the CT, will still dictate the treatment, and we should still treat patients, and not shadows on a piece of celluloid.
	Date: Mon, 29 Jul 1996 18:50:00 +1200 From: Dave Adams [dcrad@ihug.co.nz] In this context our idea of stable is haemodynamically stable after initial fluid challenge of 1,500 to 2,000ml of crystalloid or colloid and only requiring maintenance fluids thereafter. CT is not as sensitive for bleeding, for gut injury, and probably for pancreatic injury as DPL, so we do a DPL first as a screening measure. Then if we are going to be conservative we do CT to further document what it is that we are sitting on. ie DPL tells us whether there is an injury, but not what it is. We get about 6 significant liver injuries per year, and haven't had to operate on one for three years. Haven't had any complication of this approach.
	Date: Fri, 26 Jul 1996 19:26:23 From: Dr. Jay Smith [SSMITH4@surg.hmc.psu.edu] We use a pulse < 120, bp >90, base deficit <3 as criteria for "stability". Of course, stable implies stability over several measurements: 120--120 is stable but so is 60--60. The actual state of perfusion must be addressed as with the base deficit. We will also use 2 liters of replacement crystalloids to stabilize. We use lberal amounts of CT scans to determine the severity of injury and will even watch grade 4 liver injuries nonoperatively if stable. The grade 4 injury of the right lobe extending from the juxtahilar area to the costal margin often has a bleeding vessel off the right hepatic artery and coursing posteriorly and laterally that is embolizable via angiography. Embolization will leave an area of posttraumatic cyst, but these seem to be benign. Consciousness does not affect our decision because we follow these patients with blood counts and CT scans as needed.
	Date: Sat, 27 Jul 96 13:01:00 From: Stephen Streat [DCCM@ahsl.co.nz] There are many hepatic injuries which are able to be treated non-operatively (or operatively) with little consequence as to final outcome. This is probably the majority of injuries seen on CT. Before CT was used so liberally these were found at lap after positive DPL and were either ignored or had minor haemostatic procedures. Many of these patients have lifethreatening injuries in other body regions - usually the head. In my own view whether or when these patients have surgery is of little consequence - the worst that happens is that they end up with an impressive abdominal scar and a tiny increased risk of adhesive obstruction in the distant future. Conversely there are some patients who are clearly exsanguinating (eg "hypotensive and unchanged despite 2 litres rapid colloid infusion") and the luxury of selective non-operative management is not available. How long you wait before taking these patients to the OR will probably be of great importance in the overall number salvaged. If you wait till they have dilutional coagulopathy, hypothermia, acidosis and intra-abdominal hypertension you have waited too long and when you get in you are not sure which bit to put in the bucket and which bit to try and keep. It is important to bite the bullet early and take these patients to the OR fast. Many will die regardless. Finally and most importantly there are a group (a small group) who have what are often anatomically devastating injuries seen on CT who have a early moderate or even major bleed (>10 unit) and then have a period of stability and in whom operative disturbance would very likely be the difference between life and death. ("If we disturb that clot given that the disruption goes right down to the porta he will probably bleed out"). It is important to recognise this group and make vigourous effort to manage them non-operatively if possible - even if this means quite a substantial period of intensive care and repeated cautious transfusion. They are not the same as the patients who are bleeding to death in front of your eyes and should be distinguished from them by an experienced surgeon (if not an intensivist !). This message brought to you by an intensivist who (almost) never stays the hand of any surgeon prepared to cut anything. IMHO focussing discussion on "what set of numbers makes a patient considered stable or unstable" is a necessary but not sufficient part of the issue. We should also discuss the cognitive stuff too. Comments ?
	Date: Sat, 27 Jul 1996 22:04:06 From: Sue Taylor [sue@highway1.com.au] >IMHO focussing discussion on "what set of numbers makes >a patient considered stable or unstable" is a necessary >but not sufficient part of the issue. We should also discuss >the cognitive stuff too. Comments ? Yes, absolutely, and thankyou for your stimulating dissertation. But I haven't seen the discussion which gives us those "Magic Numbers" yet. Admittedly, the rules are made for breaking, and that is what the cognitive stuff is for, but some of us don't know the rules, yet...
	Date: Sat, 27 Jul 1996 21:24:20 From: Ken Mattox [KMATTOX@aol.com] A word of caution. A STRONG word of caution. Be careful. The data we generate in the large teaching hospitals with many house staff, monitors and other assistive systems is totally different from the lone surgeon in a 15,000 population community with a 25 bed hospital. If a single major complication occurs, including hepatitis C following blood transfusions, his gonads in that community are worthless. In that environment, the most "conservative" approach may be to operate and the aggressive approach is to watch.