How good are we at predicting ischaemic chest pain?

Here’s a case I had recently:

62 year old male presents to the Emergency Department with what he calls a flare-up of his congestive cardiac failure. He states he is short of breath. He is a frequent presenter to the department with this complaint. He has not taken his frusemide for the past 48 hours.

He states that he feels very tired.

His examination is as follows:

He is a well looking man with normal vitals.

Heart sounds dual, no murmur. Normal JVP, bilateral pitting oedema to mid calves.

Chest clear.

Abdomen soft.

His ECG and CXR are normal.

Bloods including a troponin are normal.

I’m unsure of the diagnosis, but given he hasn’t taken his frusemide, I treat him with that.

I am going to discharge him, however it is late at night and he lives alone, so we decide to keep him in the department overnight.

In the morning, he looks well and feels better, but still lethargic, ECG is unchanged, but for some reason someone does a follow-up troponin and it is 4.

He is diagnosed with a NSTEMI and sent to cardiology.

So the question I posed was, “How good are we at diagnosing cardiac chest pain?” and “Are there some signs or symptoms such as shortness of breath that are more important than we may initially think they are?”

Pope et al (NEJM 2000;343:1167-1170) looked at the rate of missed diagnoses of cardiac ischaemia in the emergency department.

This was a study of some 10,689 patients and the conclusion was that there was a subgroup of patients more likely not to be admitted, who had ischaemia or infarction.

Those more likely to not be admitted were:

-Non-whites

-Females less than or equal to 55 years of age

-Those with shortness of breath as their only complaint

-Those patients with a normal ECG

Chadwick et al (Ann of Emerg Med 2004;44:565-574) looked at the question of our initial diagnostic impression as clinicians and if it was adequate in excluding cardiac disease. This was a prospective registry-based study, conducted at eight hospitals. They evaluated patients where the physician’s impression was that the cause of chest pain was not cardiac. Patients had a 30 day follow-up and adverse events included STEMI, NSTEMI, going to the cath lab or death were collated.

Of the nearly 3,000 patients in this study, a minimum of 2.8% and a potential maximum of 6.3% of those being clinically ruled out of having ischaemic chest pain, had a cardiac event. What was interesting was that 53.2% of patients who had been labelled as non-cardiac, still had a cardiac marker workup.

Factors that were found in this study, to be significant in predicting those patients with a cardiac event were:

–Past history of coronary artery disease

-Age

-Diabetes

-Hypercholesterolaemia

-History of congestive cardiac failure

-Feeling of ‘weakness’ correlated highly

So beware the patient who presents with shortness of breath or weakness as complaints. Lethargy or fatigue, is especially common in the elderly, and this is the group to be most aware of the diagnosis in.

The above list is similar to the American Heart Association list, that lists the following as being important factors in the initial diagnosis:

–The nature of the presenting angina – this is critical. If a patient gives me a great story of angina, I don’t care if they have risk factors or not, they need to be assessed by cardiology as the story means everything

–Past history of coronary artery disease

–Male

-Age

-Number of risk factors

There are also presentations that result in a lower likelihood of ischaemia and they include:

-Reproducible pain on palpation

-Positional pain

-Pleuritic pain

-Stabbing pain

-Pain radiating to the lower limbs

How good is the ECG at helping us?

We know that up to 50% of cases with acute myocardial infarction (AMI) have a normal ECG at presentation. Serials are important. What is more important is that the ECG is probably more significant than all history and examination put together and a normal ECG predicts a group of patients with a lower rate of complications.

Thomas et al (NEJM;2000:342;1187-1195) evaluated patients with acute chest pain. AMI was present in 80% of patients with ST elevation of greater than, or equal to 1mm and in 20% of patients with ST depression or T wave inversion. With a totally normal ECG, the risk of AMI was 4% if there was a history of coronary artery disease and the patient had chest pain and 2% if no past history.

How about other subtle presentations?

My patients will often describe the chest pain as indigestion, or they tell me that the chest pain makes them want to belch. Are these patients fine to discharge? Here are a few facts:

-About 20% of patients with an AMI describe their pain as that of ‘indigestion‘

-About 50% of patients having an AMI had belching associated with pain.

But surely, if I give them a GI cocktail(mylanta and xylocaine viscous) and the pain goes away, that should be enough? As it turns out , the GI cocktail has been found to have the same effect on ischaemia as it does on reflux (Ann of Emerg Med 1995).

So what does all this mean? Are we now going to admit everyone? No. We are still going to risk stratify. We know that despite our best efforts that we will all miss up to about 2% of cases. The purpose of this blog is to look at all those patients with more subtle presentations.

So in summary:

-Beware the patient with shortness of breath as their only complain, especially if elderly or a diabetic, as it could be ischaemia.

-Beware the patient presenting with lethargy or weakness, especially if they are elderly, they may have ischaemia.

-Beware the diagnosis of reflux, as symptoms of belching or pain even described as that of indigestion, can be cardiac.

-Pay attention to the story. If that is a good one for angina, nothing else matters.

Next time we’ll look at the same case from the viewpoint of it potentially being a pulmonary embolism.