Permissive hypotension in haemorrhagic shock

[toc]

Question 20 from the first paper of 2011 asked about all 3 major aspects of damage control resuscitation:

• Permissive hypotension
• Haemostatic resuscitation
• Damage control surgery

The basic principle of permissive hypotension is the pursuit of haemostasis rather than blood pressure. With this in mind, one aims to perfuse the organs with some sort of minimum blood pressure (an SBP of 70-80mmHg or a MAP of 50mmHg appears to be safe), thereby limiting the volume of resuscitation fluid required. A smaller volume of resuscitation fluid means less haemodilution; a lower blood pressure means the sources of arterial blood loss will be slower to lose the blood, and more easily controlled by the pressure of internal self-tamponade.

The LITFL entry on this topic is an excellent starting point, and it is also where the time-poor exam candidate should end their reading. If however you've got a few hours to kill, trauma.org has an excellent bibliography of all the current research. This sort of reading should start with Ken Mattox's brilliant editorial article. " I have lived long enough to recall a long list of the techniques "de jour" to elevate the blood pressure" he starts, presumably in a Texan drawl. He also earns massive kudos by quoting Machiavelli at the end. Excellent stuff from one of the original authors of the 1994 Bickell paper, but perhaps best left for after the exam.

Key concepts in permissive hypotension

Definition of permissive hypotension in trauma

• Allowing a subnormal SBP or MAP in a trauma patient;
• " The strategic decision to delay the initiation of fluid resuscitation and limit the volume of resuscitation fluids/blood products administered to the bleeding trauma patient by targeting a lower than normal blood pressure, usually a systolic blood pressure of 80–90 mmHg or a mean arterial pressure (MAP) of 50 mmHg" - Kaafarani et al, 2014

Rationale for permissive hypotension:

• SBP and MAP are poor surrogates for tissue perfusion (which really should be considered in terms of oxygen delivery per minute per gram)
• Tissue perfusion equates with blood loss when the tissue itself has been injured and bleeding.
• The pursuit of a normal blood pressure is physiologically incorrect because hypotension may be one of the normal defenses against haemorrhagic shock.
• With lower blood pressure, blood flow though injured tissues and organs may slow enough for haemostatic thrombi to form.
• The goal of maintaining organ perfusion is laudable, but wasteful to life if haemostasis is not achieved.
• When pursuing organ perfusion, one must consider what one is perfusing those organs with. Crystalloid contributes nothing to oxygen carrying capacity, but rather degrades it - to perfuse organs with a uselessly hypoxic fluid seems pointless, regardless of what pressure it is under.
• Vast quantities of resuscitation crystalloid have the added disadvantage of degrading clotting function.
• Ergo, one ought to delay aggressive volume resuscitation until such time as one has controlled the bleeding in the operating room.

Key principles: of permissive hypotension in traumatic haemorrhage

• Volume resuscitation should be minimal (and ideally consist of a balanced blood transfusion, whatever you think that is).
• The subnormal blood pressure itself is not the target; rather haemostasis is the target, and the subnormal MAP is what we tolerate in order to achieve haemostasis.
• Management should be directed to earliest possible surgical control of the life-threatening injuries and the reversal of the "lethal triad";

Arguments for and against permissive hypotension in trauma

Theoretical arguments in support of permissive hypotension

Argument from biology:

MAP of 50 appears to be associated with decreased transfusion requirements but not increased adverse events. This is consistent with the idea that MAP only vaguely correlates with organ perfusion, and that in pathological states aiming for normal organ perfusion may in any case be counterproductive. Evolution has guided the survival of those primates who are able to tolerate haemorrhagic shock long enough to recover and breed; this has been going on for some time in the absence of effective prehospital retrieval services. Specific defensive strategies developed by the mammalian autonomic nervous system have largely involved the intentional decrease in the perfusion of nonvital organs. Ergo, it is probably inappropriate to aim for a normal healthy perfusion of those systems.

Argument from analogy:

The tolerance (indeed, the pursuit) of systolic pressures around 50-100mmHg has already become accepted practice in the management of rupture abdominal aortic aneurysm (van der Vliet, 2007), though anything below 70mmHg may still be too low (Powell et al, 2014). The intra-abdominal exsanguination which occurs there is functionally similar to that experienced by blunt torso trauma patients. It seems unusual to use a seperate set of standards for two  sorts of analogous intra-abdominal haemorrhage. What's good for the ruptured AAA should also be good for the ruptured spleen.

Argument from zoology:

Animal models of permissive hypotension uniformly support this practice. Mapstone et al (2003) reviewed the then-existing sum of animal trials and came to the conclusion that within the animal model setting, permissive hypotension is beneficial.

Argument from unaided logic

The systolic blood pressure is the variable which is least related to organ system perfusion. It makes no sense at all to have a resuscitation strategy in shock (defined as inadequate tissue oxygenation) guided by the systolic blood pressure. It would perhaps be wisest to ignore it completely.

Theoretical arguments against permissive hypotension

Argument from technicality

If we decide that the systolic blood pressure is a worthless parameter and decide we are going to ignore it, which variables do we use to guide resuscitation? Without endpoints, there is nothing to guide management. MAP is well validated as a therapeutic target and is known to correlate with organ bloodflow (if not oxygen delivery and utilisation per se). It is hard to abandon traditional variables in the absence of agreement on which better variable to measure (or, indeed, how low one can go with the traditional ones). 

Argument from sloth

Permissive hypotension may be interpreted as the permission to ignore hypotension of all causes; the result may be an intolerable increase in the rate of missed cardiac tamponade and tension pneumothorax.

Argument from competing interests

If the patient has a traumatic brain injury, the practice of permissive hypotension conflicts with the imperative of maintaining cerebral perfusion. This is explored to a satisfactory depth in the discussion of a recent case report (Kohli et al, 2014) as well as the reader's response to it (Senthilkumaran et al, 2015). I will summarise by saying that the easy answer is to always protect the brain above all other priorities. Permissive hypotension may become possible later, if cerebral oximetry is available (i.e. one decides to target something other than ICP/CPP)

Argument from gerontology

If the patient is elderly, theoretically they may require a higher MAP target, and the pursuit of a "young" MAP might result in tragic underperfusion of elderly organs. However, this theory has not been supported by a recent audit of trauma database (Bridges et al, 2015) which could not find any association between mortality and the combination of old age and hypotension.

Evidence for the use of permissive hypotension

Bickell et al (1994): prospective trial; 598 adults with penetrating torso injuries. Some received standard pre-hospital fluid therapy; in others the fluids were delayed until they got to the operating theatre. There was a small mortality benefit in favour of delayed resuscitation (30% vs 38%), though it must be pointed out that both the groups had similar blood pressure on arrival to the operating theatre.

Turner et al (1999): randomised controlled trial, 1309 trauma patients. In the treatment group, resuscitation with fluid was delayed until they reached the operating theatre, much like the Bickell study. No difference in mortality and morbidity was found. The trial was crippled by protocol violations (of those who were supposed to get fluids only 31% actually got fluids; of those not supposed to get fluids, 20% received them in violatio of protocol). Overall, it was difficult to draw any conclusions from this.

Dutton et al (2002): randomised controlled trial, 110 patients with haemorrhagic shock. The systolic pressures between groups were 100 mmHg and 114 mmHg. Survival was identical: 92.7%, with four deaths  in each group. The proponents of permissive hypotension were gladdened by the absence of any increased mortality with the lower blood pressure. Opponents of the practice pointed at the lack of benefit, and the relatively normal blood pressure even among the "hypotensive" group.

Scheriber et al (2015): multicentre RCT; 192 patients randomised. The hypotensive group was really hypotensive (SBP >70). Surprisingly, they did very well: the 24 hour mortality was 5%, vs 15% in the normotensive group. The difference was even greater on sub-group analysis of the blunt injury patients (3% vs 18%)

The HypoResus Trial has completed, and its publication is awaited; the groups received either a 2000ml or a 250ml bolus of saline. According to the (already available) study results, 24 hour mortality was better among the small-volume group (5% vs 15%). 

Support for permissive hypotension among the guideline-makers

• No professional society-supported evidence-based recommendations exist.
• The ATLS manual (9^th edition) has acknowledged this practice by changing the recommended starting point for resuscitation from 2L of crystalloid to 1L of crystalloid. The cautious statement made was that "in penetrating trauma with haemorrhage, delaying aggressive fluid resusicitation until definitive control may prevent additional bleeding".
• LITFL reports that permissive hypotension is "not widely accepted in Australia". There certainly seems to be regional variation in practice, even when comparing two adjacent resuscitation bays of the same emergency department.