Your group sought to understand if the SCAI shock stages can classify mortality risk. Can you tell us about your research?
For this analysis, we reviewed the different papers that have been published since the original Society for Cardiovascular Angiography and Interventions (SCAI) shock proposed classification.1 In the SCAI shock classification, the severity of shock is graded on a five-stage scale from A to E. Stage A is people who don’t actually have shock, but who are at risk because they have acute cardiovascular disease. Patients in stage B have what I would call pre shock, meaning that they have some form of hemodynamic instability, but still have adequate organ perfusion. Under the SCAI shock classification, the definition of shock as a state of hypoperfusion is key, and Stage C is where you start to see the presence of shock, because there is hypoperfusion requiring some form of an intervention. Stage D shows evidence of deterioration, which is a little hard to define, but essentially means that the initial therapies have not worked effectively. Stage E would be the most severe form of shock, where patients have circulatory collapse, what we usually call extremis. The SCAI shock classification is designed to be very flexible, and as a result, each group came up with a different way to define the various SCAI stages within their own datasets. The different analyses we reviewed all used the same types of variables, but the way these variables were combined by researchers in terms of cutoffs was, in some cases, quite different. A few of the studies used physician chart review to assign the SCAI stages, which theoretically should be the gold standard, but most of the studies just used retrospective data and had to come up with some definitions as a result. In general, the data are combinations of vital signs, laboratory data, and the different therapies that were used. A patient with worse vital sign abnormalities, worse laboratory abnormalities, and/or who is requiring more vasopressor or mechanical circulatory support, would therefore be in a higher SCAI shock stage. Even though there were six or seven different definitions used to define the SCAI stages, all definitions showed the same ability to stratify mortality. Reflecting gradations in mortality is strong evidence that the SCAI shock classification is clinically useful, regardless of how the different stages are defined, and this is one of our key findings. Some of the different classification systems provided greater or lesser degrees of separation, but it is hard to know whether it is because the classification systems weren’t as accurate or if it is simply a reflection of the studies’ unique populations. None of the different SCAI shock stage definitions were compared within the same population, so it would be unfair to say that any one definition is better than another based on the data we have thus far. We did find several subgroups that were of interest, because some of these populations were fairly heterogenous. We sought to confirm that what we saw was not just an artifact of how the population is constructed, but a relevant variable. For example, in cardiac care unit patients who had acute coronary syndrome, we saw very substantial gradations in mortality in the same population. In patients that had heart failure, again, we saw mortality gradations, maybe not quite as impressive, but again, overall, quite similar. One of the key ideas put forward in the SCAI shock consensus statement was that patients with cardiac arrest are a unique and special subgroup at higher risk, which has been confirmed by studies looking at those patients specifically. Our study did look at the presence or absence of cardiac arrest; we found that at any level of SCAI shock stage severity, patients with cardiac arrest were more likely to die, which is a common and well-known observation. However, we looked at whether the mortality risk was due to shock or due to something else. We were able to show that cardiac arrest patients have worse outcomes at every level of shock severity, which suggests that there is more going on with these patients than just shock. While we may be able to successfully treat their shock, cardiac arrest patients may still have other problems like brain injury and organ failure that drive their outcomes. It doesn’t necessarily mean we shouldn’t treat these patients aggressively, but it does mean that we should have a clearer understanding of what to expect with their prognosis and understand that their problems include more than just hemodynamics.
What are the implications of being able to identify additional complicating factors?
It will allow us to create further gradations in risk. I mentioned cardiac arrest, but we identified several other subgroups as well. When we think about patients who have cardiogenic shock, it is with the idea that they have multi-systemic illness, not just shock. They have shock, they have organ failure, they have respiratory failure, and they have other issues, each of which is, on its own, potentially life-threatening. What is key is that we can define the severity of shock, which helps us to match appropriate therapies, and potentially triage and transfer when necessary. We can consider patient factors that either adjust the severity of shock, or more importantly, that are not related to shock at all, but will still determine the outcomes. The classic example would be older age. Some people do use the combination of SCAI stage and age to help triage when they are transferring patients in for extracorporeal membrane oxygenation (ECMO). Certainly age is a easily identifiable marker of higher risk. It is not the be-all and end-all, but age is independent of the shock severity as a determinant of mortality. All these studies showed that we can identify factors that worsen the shock severity, which is relevant and worth understanding, because these are factors can be addressed by hemodynamic therapies like circulatory support. There are other factors that are not directly related to the shock itself, but are complicating factors important for prognosis. We found that age, of course, as well as a number of hemodynamic variables and other factors on echocardiogram, all led to patients experiencing deterioration beyond the initial SCAI shock stage. All are things that made sense as predictors and now are borne out by the literature.
Does the presence of these complicating factors affect treatment?
In most cases, the presence of one of these variables means either simultaneous treatment of both problems or treatment of the cardiogenic shock, hoping to have a reversal of the shock treat the other problem. Trying to link mortality studies to treatment options is one of the most common issues we see in this type of observational outcomes research. The challenge is, how do you take a person who is at higher risk and tailor their treatment? We often simplify risk too much, because again, there are different components of risk. For a patient who has a very high severity of shock, where their high severity is what makes them high risk, an aggressive hemodynamic support platform could really help. But let’s say the patient is high risk because they have other factors such as brain injury, older age, or organ failure, and their shock stage is not that severe. Using a very aggressive hemodynamic support device may fix the shock component, but if the shock is not really what is making the patient sick, then maybe a hemodynamic support device won’t be as effective. We are on the first step of a more refined methodology of evaluating patients. We have identified multiple different components that must be considered as we evaluate shock severity within the SCAI classification and provided some insight about other hemodynamic-related risk factors that can be taken into account. We must think about the other organs and then the patient themselves, because obviously patients present with different levels of chronic illness that influence their outcomes. We have shown that there is a difference in mortality by SCAI shock stage, which is a first step. A difference in mortality doesn’t always mean that there needs to be difference in treatment, but it can help temper your expectations and it can help identify which factors help in decision-making.
Have there been any studies that have definitively confirmed optimal support or therapy for specific SCAI shock stages?
Unfortunately, none of the studies we reviewed specifically looked at the use of different support strategies as a the function of SCAI shock stage. That is the next necessary leap in the literature; i.e., we did this therapy at this SCAI shock stage, and this is the outcome, and we did the same therapy in a different SCAI stage, and the outcome was either the same or different. I hope to see people assigning the SCAI stage retroactively to randomized trials or ideally, prospective studies doing the same. My hope is that as we recruit for new randomized, controlled trials in cardiogenic shock, that the standard approach will be to assign the SCAI stage at randomization using predefined criteria. It will make it easier for clinical trialists to come up with consistent definitions that are supported by the literature and that other providers agree are reasonable.
Is assignment of a particular SCAI shock stage possible outside the cath lab?
This is an area of active research. In different care settings, we have access to different types of data. It can be very difficult, for example, to get a lactate level in the prehospital setting, such as in an ambulance. But at the same time, in the ambulance, you can get vital signs and examine the patient. There is information we can have in these settings that will allow us to make preliminary evaluation that is probably valid, but we just don’t have the data yet. I think we may determine that at each point along the care spectrum, you’re going to have a certain set of data available. It will allow you to do your best estimate of SCAI stage at that point in the care spectrum. The more data you have, the more accurate the estimate will be. I am working with colleagues in the emergency room (and I am probably not the only one doing so) to see if we can come up with a prehospital SCAI stage, an emergency department SCAI stage, a cath lab SCAI stage, and an intensive care unit SCAI stage that will all overlap, but will respect the limits of the data available in these different settings. Not every cath lab has a point-of-care lactate machine. If lactate is an important variable for defining SCAI stage, which it generally is, then that might allow a cath lab to justify getting one of those machines if they don’t already have one, and might further help with the triage and the severity definitions.
Any final thoughts?
This study, by reviewing all the relevant published literature, clearly validates the SCAI shock classification’s effectiveness for risk stratification of mortality specifically. This was a group effort; I have a long list of colleagues who were all part of this work. Many of the people on this writing group are the ones that are doing the research. It is a neat process of taking the research and feeding it back into the guidelines, and SCAI is on the cutting edge of this process. Unanswered questions remain, which we hope to highlight and address in an upcoming update to the SCAI shock classification. As this classification continues to evolve, it will become increasingly useful. Our hope is that the SCAI shock classification will become more than just a research tool and that we can help providers put it into practice.
Jacob Jentzer, MD, can be contacted at email@example.com
- Baran DA, Grines CL, Bailey S, et al. SCAI clinical expert consensus statement on the classification of cardiogenic shock: This document was endorsed by the American College of Cardiology (ACC), the American Heart Association (AHA), the Society of Critical Care Medicine (SCCM), and the Society of Thoracic Surgeons (STS) in April 2019. Catheter Cardiovasc Interv. 2019 Jul 1; 94(1): 29-37. doi: 10.1002/ccd.28329