It has been almost a week since I published a rant about my many complaints with regard to ICD-10 or the state of affairs concerning audits and “clinical validation.”
In a departure from my usual editorialized op ed, I wanted to shift gears and write about something from a more academic perspective. In this case, to address a question that was recently posed to me regarding what really constitutes septic shock. My research uncovered some surprising revelations, namely that the traditional definitions of shock we have all been given may not be 100 percent accurate, especially with respect to septic shock.
Insurance auditors have begun to deny septic shock based on the Sepsis 3 criteria. This is not a surprise to anyone who has been following issues with sepsis lately. The new definition of septic shock is hypotension sufficient enough to require vasopressor therapy to maintain a mean arterial blood pressure of 65 or greater. Viewed as an absolute cutoff, if a patient is not requiring vasopressors in order to maintain a mean arterial blood pressure of 65 or greater, the claim is denied in the process known as “clinical validation.”
Unfortunately, this is also wrong. This single criterion is not an absolute determining factor of the presence or absence of septic shock. It is true that any shock will generally feature hypotension as a symptom, but there are exceptions. It is also true that septic shock will usually present as refractory to simple fluid resuscitation due to a myriad of mechanisms that lead to systemic vasodilation and distributive shock. However, there are specific circumstances in which a patient may be able to compensate for the mechanisms intrinsic to shock and maintain their blood pressure within a relatively normotensive range. Those compensatory mechanisms often occur in the presence of the shock mechanisms that are still destructive at the cellular and organ level. Rather than proving that septic shock is not occurring, the patient is experiencing septic shock in spite of the fact that they are registering as relatively normotensive with regard to their blood pressure. This is in sharp contrast to other types of shock.
Changes occur at the microvascular and cellular level during septic shock that may not be manifested in the vital signs. I am speaking of such things as activation of inflammatory and coagulation cascades, dysfunctional utilization of oxygen and nutrients, vasodilation and vascular maldistribution, and capillary endothelial leakage. We will proceed to touch on many of these issues individually below.
Septic shock is a subtype of shock known as distributive shock. Excessive vasodilation and impaired distribution of blood flow via arteriovenous shunting is characterized by a decreased resistance or increased venous capacity from general vasomotor dilation and dysfunction. A patient like this can have a high cardiac output but end up with a low diastolic pressure and relative hypotension. Physical assessment, however, will not reveal much in the way of shock symptoms as they will have warm extremities with and good capillary refill. Here you have a patient who may only be mildly hypotensive and is not requiring vasopressors but is in septic shock. The problem is that standard criteria cannot capture the damage being caused by the endothelial dysfunction. Deciding when more aggressive hemodynamic support is required is very difficult in these patients.
While septic shock is poorly understood, we know that endothelial disruption, thrombosis of end organ capillaries, increased vascular permeability, and activation of inflammatory and coagulation cascades are common factors. These factors create additional stress on the host defense mechanisms, in effect creating a positive feedback loop and leading to even further endothelial and end organ damage. This process can certainly happen in the absence of frank hypotension. The hypotension occurs via the induction of nitric oxide in most cases.
Patients who are well hydrated and have an adequate cardiac output will compensate for these stresses and remain normotensive. Patients who are dehydrated or who do not possess the necessary cardiac output compensatory abilities will present as hypotensive. Standard medical care includes fluid administration at roughly 30ml/kg in a relative short time period. Since we artificially augment intravascular volume in this way, cardiac output is often elevated, sometimes referred to as the hyperdynamic phase of sepsis and shock. Although the cardiac output is elevated, the arterial-mixed venous oxygen is usually narrow and the blood lactate level is elevated. What is occurring is low global oxygen utilization limiting oxygen uptake in septic shock. This is termed maldistribution of blood flow. During such a fall in oxygen utilization, cardiac output adapts so that most vital organs remain perfused. Through what is known as the Frank-Starling mechanism, the cardiac output is often increased in response to the systemic vasodilation in patients with septic shock. If the cardiac output is increased or “compensated” via the Frank-Starling mechanism, then the septic shock will be masked or even missed completely if the decision is being made based solely on the basis of the mean arterial pressure.
As mentioned previously, shock is identified in most patients by hypotension and inadequate organ perfusion. Keep in mind however that this applies to most, not all. One of the problems with how we teach CDI and coding is that we utilize the “80/20 rule.” What I mean by that is we give the low-level clinicians and coders a guideline that works 80 percent of the time. In the interest of simplicity and brevity, we do not address what happens the other 20 percent of the time. When you have a rule that makes you right 80 percent of the time, you also risk being wrong 20 percent of the time.
The mechanisms related to a low cardiac output and/or low systemic vascular resistance are highly complex for a low-level clinician such as a record auditor. A patient can be relatively hypotensive, but if they are adequately perfusing and oxygenating their tissues, they are not in shock (at least, not yet). Conversely, the blood pressure can appear to be normal, but the patient is in shock if the perfusion is inadequate, oxygen utilization is reduced, or endothelial disruption is occurring at a cellular level. Due to the complex mechanisms involved in septic shock, this is more common than you might think.
Don’t take my word for it—talk to your hospitalists, intensivists, or ICU physicians about “cryptic shock” or “compensated shock” in patients who appear to be normotensive or who are only slightly hypotensive, but who exhibit signs of organ hypo-perfusion in the presence of a rising serum lactate level above greater than 3 or 4/with acidosis. Shock is not defined by vital signs only, but is more defined on a molecular level here, with evidence of a lack of perfusion and hypoxia in the tissues and vital organs.
In conclusion, Septic shock does not occur in a vacuum. You don’t go from not having shock at 10:04 and then suddenly have shock at 10:05 (at least not septic shock). If a patient’s blood pressure finally tanks on hospital day 2 or 3, you can bet that from a clinical standpoint, the process or continuum was already occurring the day prior. We often do see the drop in blood pressure, but we oversimplify the framework so that only hypotension is the definition. We tend to ignore the myriad of real clinical situations when the hypotension might not be apparent early on. The oversimplified definition of shock as vasopressor required to maintain a MAP of 65 doesn’t come anywhere near to explaining the processes which are already occurring and which are still defined as part of the shock syndrome which often happen before the observation of a decreasing blood pressure.