Chest pain and the Heart Foundation Risk Stratification Guidelines in the real world


Risk Stratification in the Real World

Anne-Maree Kelly from Western Health in Melbourne has examined over 700 patients with potentially cardiac chest pain in this paper published in the June issue of Emergency Medicine Australasia. Although the author's main conclusion appears to be that the NHF guidelines are somewhat of a failure in the real world, I think the study goes a long way to clear a lot of the fog around the issue of risk stratifying a chest pain patient.

NHF guidelines

In case you are unfamiliar with them, the NHF has a well publicised set of criteria for separating patients into high, intermediate and low risk when they present with chest pain. They are available here in a groovy wall poster (just to the right of middle).

High risk patients have one of the following characteristics: a really good story (repetitive or prolonged pain), a suspicious ECG (new T inv, ST depression or non-reperfusable ST elevation), positive Trop or CK-MB, sustained VT, haemodynamic compromise or new MR, syncope, LFEF<40%, PCI in last 6 months or CABG ever, DM with typical ACS symptoms or eGFR <60 with typical ACS symptoms.

Intermediate risk patients have a good story but a late presentation, are over 65, have known IHD, have 2 or more risk factors for IHD, have diabetes or renal impairment with atypical symptoms or are already on aspirin.

Low risk patients are everyone else (i.e. people who ought to have gone to their GP, lost some weight or stopped smoking so much).

The guidelines recommend putting all the high risk patients into CCU, on aggressive anticoagulation (subject to bleeding risk) and on a pathway to angiogram as an inpatient. The intermediates are supposed to get serial enzymes and a negative stress test before discharge and the low risk patients get discharged after a respectable period of observation and have urgent cardiology follow up.

Cough- splutter- choke!

If that is your response to the above then you have probably been working in emergency medicine in Australia for a while. Let's look at some of those recommedations in detail.

First the easy one: low risk patients. These are people who have lost some exercise tolerance on their stable angina or who come and tell you that they had some exertional chest pain 2 weeks ago. Picture these patients. They are actually a pretty broad group (which highlights the limitations of these attempt to break down into dot points the subtle art of medicine). some of these people you are going to want to send urgently to a cardiologist. The 50 year old who says they have had a month of new exertional chest pain that they have managed by curtailing their activity or the person who has been watched for a while with <50% stenoses who has gone from waling up hills to taking it slowly along the flat in the last month and has finally been dragged in by their partner certainly warrant a phone call to the cardiologists rooms to tee up an early review. The rest though, are going back to their GP in my town, so they can coordinate assessment of BP, lipids and lifestyle factors, stress ECG if appropriate and cardiologist referral.

Next, the high risk patients. Now I have no problem with putting people with dynamic ECG changes, positive biomarkers, ventricular tachycardia or cardiogenic shock into CCU or HDU. However, a history of DM or CKD and a good story might get you a telem bed in my hospital and there is every chance that we will monitor you in the ED SSOU for a 6 hour troponin and discharge you for outpatient provocative testing (besides, how dye happy do you want to be with people who have an eGFR<60 or Diabetes?). The fact that you have chronic heart failure or have, at some stage, had the plumbers in, doesn't impress me terribly much in the chest pain workup although it makes me more likely to send them to a cardiologist on discharge rather than back to their GP. Finally, syncope is important with chest pain but it makes me think of PE and dissection more than ACS.

As for the intermediate patients, I have never known what to do with these and most of them get a 6 hour troponin in SSOU and go back to the GP, perhaps with an outpatient provocative study arranged by me before they go.

And so for the good news

Well the great news in fact from from Kelly's paper in EMA is that for a start you can more or less forget about the intermediate risk patient. They just don't exist. Out of a total of 768 patients, 1/62 low risk patients went home with a discharge diagnosis of unstable angina and none had an MI or a MACE ("bad outcome" - see below), 1 out of 254 intermediate risk patients got revascularised during the follow-up period and all the rest of the MIs and MACEs were in the high risk patients. That means the high risk criteria were within a hair's breadth of 100% sensitive for the outcomes of interest. On the down side they were only about 50% specific so half the patients you "admit to CCU" have no need to be there (whether the other 50% actually needed to be there either is a whole other question, I think they probably don't for the most part we are not quite there yet). This certainly validates current Australian practice of making good use of SSOU and early outpatient provocative testing and reserving CCU for people who are actually requiring close watching.

What is a MACE?

MACE is a term that you come across a lot in chest pain and ischaemic cardiac disease literature. It stands for Major Adverse Cardiac Event and has become a standard composite end point for cardiology studies. It includes death, cardiac arrest, cardiogenic shock, arrhythmia, revascularisation and MI. It is a term to be very wary of. Remember that MI these days is not what it was 20 years ago. When a patient hears "heart attack" they figure something pretty awful happened to their heart. In fact, MI means a troponin became positive. A troponin of 0.5ng/mL from some inferior ischaemia is not actually in the same ballpark as death or cardiogenic shock. In fact we know that a lot of positive troponins have nothing to do with IHD, especially in a population that includes CKD patients. At most, small troponin leaks mean the patient is a bit more likely to have IHD than we thought and they probably warrant up-triaging toward an angiogram. Revascularisation is worth being a little suspicious of too. If it isn't corrected for insurance status I am not sure what it really means as a statistic. Basically it means a cardiologist thought that some plumbing might be beneficial. Once again, not in the ballpark with death or cardiogenic shock.

It is interesting to note in this study that of the 452 high risk patients, about half went home with a diagnosis of either MI or Unstable Angina, 85 had a troponin leak during the follow-up period, 7 were revscularised and 1 had an arrhythmia.

It is also worth noting that almost all the MACEs occured in the first 7 days after discharged (patients were followed up at 7 and 30 days) so early follow up means early.

Take Home

What I take from this is that what we do now is about right in terms of safety but it could be a lot more efficient. I am reassured that I can stop feeling uneasy about pretending that the intermediate risk group don't exist as a separate entitiy from the low risk group because for my purposes they pretty much don't (not to say that these factors won't weigh in to the decision making that occurs at follow-up, they just aren't going to make a major impact on ED decision making). I am pleased to see that even high risk only means 50% sensitive and that SSOU pathways are therefore probably valid for selected high risk patients.