This chapter deals with the investigations and resuscitation of massive bleeding into the abdominal cavity.
There are really two main sources in the abdomen where one can bleed torrentially: a ruptured abdominal aneurysm or a retroperitoneal hematoma.
Ruptured abdominal aortic aneurysm
The AAA is most common in male smokers.
The question is, when will it rupture?
- Diameter is the best predictor of rupture
- Anything over 30mm in diameter is abnormal
- Anything over 50mm has a 1% annual risk of rupture
- Anything over 60mm has a 17% annual risk of rupture
In the presence of poorly controlled hypertension, these things tend to grow in diameter by about 3-5mm per year.
When it ruptures, one tends to have about 30% pre-hospital mortality, followed by a 40% perioperative mortality. And of course, it is almost uniformly fatal in patients who are not offered surgery. Which is a fair few - give the vascular disease profile which predisposes a person to AAA formation, one can expect that a fair few of these patients will have poor cardiorespiratory performance and poor quality of life.
But let us say you go ahead with the repair.
The key controversy is the target blood pressure. The patient is shocked and the lactate is rising; one may be tempted to react to this with fluids and vasopressors. However, this may not be the best strategy. After all, the aorta has a hole in it, and the systolic blood pressure is probably directly related to the rate of exsanguination though this defect. Authors have suggested a target systolic pressure of 70-80mmHg. It seems low, but it is a fair compromise - one simply has to accept that one's patient will progress equally in the direction of both bleeding to death and global hypoxic-ischaemic injury.
Concerns with anaesthesia
The key problem is induction of anaesthesia. The conscious ruptured AAA patient is in considerable pain, and their hemodynamic performance is largely reliant on their abdominal wall musculature contracting around their sore belly, tamponading the bleed. Consider now what might happen if you relax them with anaesthetic drugs and neuromuscular junction blockers. Now, nothing is putting pressure on the defect in the aorta. This is a recipe for a PEA arrest. The next phase of management should really be the cross-clamping of the aorta, which will result in a marked improvement of hemodynamic performance, but until the clamp is in place the anaesthetist is potentially stuck performing CPR on a bunch of empty ventricles.
The patient will ultimately return to ICU intubated.
Post-operative complications of emergency AAA repair
There are several issues one wought to look out for in the returning post-AAA repair patient:
- The abdominal viscera likely lost perfusion for some time; ischaemic phenomena should be expected, including
- Ischaemic renal failure (which is unlikely to recover)
- Ischaemic pancreatitis
- Ischaemic colitis
- Ischaemia-associated upper GI bleeding
- ischaemic paraplegia: spinal vessels may be compromised
- There may be a global ischaemic insult due to the shock state
- There may be lower limb ischaemia
- The patient will be hypothermic post-operatively
- Coagulopathy associated with massive transfusion seems almost mandatory
Key points for post-operative management of emergency AAA repair
- Extubate them early. There is usually no pulmonary insult, and their COPD does not fit well with prolonged intubation.
- A thoracic epidural will help
- Watch for lower limb infarction and multi-organ system failure, as these may be convenient signals to start palliative management.
Why would you bleed into the retroperitoneal potential space? Well, you may have had some sort of procedure performed (eg. a renal biopsy). Or, trauma (eg. axe attack). Alternatively, you might have bled spontaneously, and this may be completely attributable to your anticoagulation therapy.
Aetiology of retroperitoneal haemorrhage
- Procedural complication
- Antiplatelet therapy
- Malignancy of the kidney or adrenal gland
- Vascular disease of the kidney
- Spontaneous rupture of retroperitoneal veins
Renal biopsy has actually come up in an SAQ, in the sense that the bleeding complications following renal biopsy took up one entire 10-mark CICM exam question (Question 4 from the first paper of 2022). One can imagine how a questions like this could have been stimulated by somebody's really negative experience.
What could have caused such bleeding following a renal biopsy? Or to borrow from Question 4, what five factors could have contributed? Don't all immediately say "poor technique". To be sure, most people who work in ICU will notice a distinct concentration of these events at the beginning of the academic year, when new nephrology trainees are cavorting excitedly around the department, and we have all seen our share of renal biopsy results return as "normal arterial intima". However, there are also some modifiable and non-modifiable patient risk factors. Borrowing from The Boston kidney biopsy cohort (Palsson et al, 2020) and Whittier (2012), the following
|Modifiable risk factors
||Non-modifiable risk factors
- Anticoagulation or antiplatelet agents
- Hypertension (SBP > 140 during the procedure)
- Agitation, tachypnoea
(i.e. moving target)
- Female sex
- Poor baseline renal function
(i.e. smaller target)
- Aetiology of renal failure (autoimmune disease, amyloid, acute tubular necrosis)
- Use of high dose steroids
Steps to reduce the risk of bleeding include:
- Assess coagulation: ideally, with TEG or ROTEM
- Withhold anticoagulation and antiplatelet agents
- Administer DDAVP to control the uraemia-associated platelet dysfunction
- Control BP during the procedure with a mixture of short-acting antihypertensives and adequate sedation
- Use general anaesthetic or sedation to reduce patient movement during the procedure
- Use ultrasound guidance
But let's say they bled anyway. Apart from controlling the coagulopathy and replacing the lost blood, one has several treatment options.
Endovascular repair or angio-embolisation
- Minimally invasive: can be tolerated by a reasonably awake patient, which means no need for a general anaesthetic
- Stents and suchlike can be deployed via an endovascular approach, i.e. embolisation is not the only possible solution
- Less blood loss than with an open approach
- This bleeding needs to be arterial, or you will get nowhere.
- There needs to be a significant rate of bleeding for it to be detectable as an extravasation of contrast
- You do often end up embolising various useful structures (eg. kidney).
- The bleeding may continue without your knowledge, i.e. it may not be immediately apparent from imaging
- Interventional radiology services may not be available everywhere
- Radiation and contrast exposure are necessary
- Haemostasis is easier to confirm by direct vision
- Repair of damaged vessels is possible, preserving the function of the kidney
- Contrast is not required
- Venous bleeding can be controlled in this manner
- Allows the abdomen to be left open, preventing abdominal compartment syndrome
- Highly invasive procedure, with nontrivial associated pain and potentially long recovery time
- Requires a general anaesthetic
- Requires trained surgical staff, i.e. also not available everywhere
- More blood loss than IR procedure
- No invasive procedures, no radiation or contrast, no complications of anaesthetic
- Haematoma may tamponade itself and the bleeding will eventually stop
- If the haematoma is caused by coagulopathy and antiplatelet effects, the bleeding should stop once these are corrected
- No embolisation or ligation means no loss of organ perfusion and function
- This is the option with the highest transfusion and blood product requirement
- Transfusion is itself not without risk
- Tense haematoma may put pressure on surrounding structures, and the kidney may lose blood supply anyway
- Abdominal compartment syndrome may develop
- In a proportion of cases, this management strategy will fail, and open surgery or IR procedures will become necessary.