Chest pain/protocols and algorithms. Not everyone needs an EST!

May 30, 2012 by · 3 Comments

[I know the video isn't immediately relevant but it's still awesome...]

From the blogs and podcasts, it seems that everyone in the US with a chest and some pain in it will get a rule out in the ED followed by some provocative testing. Amal Mattu (who I’m stalking, seeing at an advanced ECG course pre-ICEM2012) falls into the category of Very Smart People, and I’ve heard him say that we need provocative testing. It has not been my (limited) experience that EDs in the UK and Ireland do provocative testing on everyone.

This study seeks to show that a protocol of not stress-testing everyone is a good and fine thing to do.

METHODS

  • observational data on the protocol used in St.Paul BC, Canada
  • got in on the basis of triage criteria
  • pts got the usual ECG and assessment (and this “assessment” is probably the most important part and the most difficult to quantify and reproduce.)
  • trops at 2 and 6 hrs
  • if considered low risk, pt could be discharged
  • if considered higher they could get EST (or other appropriate stress) within 48 hrs
  • fairly low threshold for referral to cardiology it seems
  • telephone follow up and they checked death registry if needed
  • primary outcome was AMI or confirmed unstable angina within 30 days (the second part is a bit of a problem as admission and tests formed part fo the diagnosis and it’s always hard to know if everything the cardiologists do is gold)

RESULTS

  • 1255 pts of whom they removed 55 from the protocol (remember this is a nurse initiated one so this is pretty damn good)
  • 50% discharged with no planned provocative testing
  • didn’t miss a single ACS by 30 days (oevrall rule in rate was 10% and most picked up in ED, only a few by provocative testing after the initial assessment)
  • 2% lost to FU who didn’t attend ED or die in the region
  • of note of the 10% rule ins; 10% of these had a TIMI of 0 (but hopefully you’re all clear by now that the TIMI isn’t a clinical decision instrument that we can use as EPs)

MY THOUGHTS

  • this is all about patient selection – if we can work out who goes into this protocol then we can rest safe
  • reassures us that low risk patients probably are just that
  • fairly compelling that we can keep doing what we’re doing – in my experience in Norn Iron we only did provocative testing on a select bunch of our chest pains. There were some who got rule out trops who we did no further testing on.
  • It’s useful to know that a 6 hr trop is a useful rule out. Where I’ve previously worked we’ve based things around a 12 hr troponin as our rule out (though I confess it may have changed since I left a couple of years ago.)

Scheuermeyer, F X, G Innes, E Grafstein, M Kiess, B Boychuk, E Yu, D Kalla, and J Christenson. “Safety and Efficiency of a Chest Pain Diagnostic Algorithm with Selective Outpatient Stress Testing for Emergency Department Patients with Potential Ischemic Chest Pain.” Annals of Emergency Medicine 59, no. 4 (April 1, 2012): 256–264.e3. PMID 10.1016/j.annemergmed.2011.10.016 PMID 22221842

Filed Under: ACS, critical appraisal ·

The SMART EM stress test episode

February 24, 2012 by · 4 Comments

[ WARNING - These are a bunch of scribbled, fairly rough unedited notes on the SMART EM podcast on stress testing. I was on a rather crowded bus in north wales at the time and I was sharing the seat with a mid-sized dog (seriously you can read about it here...) so it is likely full of inaccuracies and comedy autocorrects so buyer beware...]

2010 AHA guidelines are a philosophy of chest pain

AHA recommend EST first line. Other stuff is for people with a reason (ie DSE for poor mobility)

Bausfield 1918 one of the first reports of ECG use to diagnose ischemia.

1950 was first common use of EST for diagnosis

(?) Cuhn 1970 was first to use cath as gold standard. Looking for obstructive cardiac disease.

CAS study showed that in all comer chest pain referred for CABG only 20% had stenosis. Given that CABG is only done for a proportion (left main, triple vessel).

They actuallly used this info to skip the EST and go straight to cath in classic stories.

Goldman 1982 thought EST didn’t add enough. History and enzymes to direct you to angio +/- CABG

Universal agreement that low risk shouldn’t get anything.

Duke treadmill score publihsed 1981. A more nuanced way of Reading the EST and applying it. Adds 5% accuracy. Used it to prognose death not CAD. This is a change. Goldman and everyone else talked about prediction of CAD. But they used prediction of death to prod toward angio.

1998 Frolicker (?)  published best EST study. Everyone who got EST got angio here (no WU bias). Sens 45% spec 85%. Need to know the rule in rate overall though.

Now people start using it on in patient following Q wave mi. EST here able to predict left main and triple vessel; it predcted CABG with some use.

What the hell is unstable angina?
- it has morphed and drifted over time.
- unstable angina are often now nstemi.
- non q wave mi would be a term from then.
- they overlap is the key point.

1977 one guy stated that you should not do EST on unstable angina/nstemi.

Butman and Wilcox (… snigger…) EST 72 hrs after unstable angina. Showed that 0.5% had an MI after EST. Therefore (in their eyes) it’s safe. Now it’s ok to do EST in everyone. (both q wave and unstable angina).

Larssen published data set of mi and unstable angina. 20% stemi and the others would be mi by today’s satndards.of the unstable 20% death/MI in a year.

The lasrssen paper is the one that repeatedly gets cited to justify stressing unstable angina pts before the leave the hospital.

1994 AHCPR guideline suggests EST wihtin 72 hrs is safe and should be done for unstable angina.

1997 first AHA guide on EST in unstable angina. Say that unstable angina either progress to mi/death or settle to be stable. They presume these are high risk pts and suggest EST is to point toward CABG. They are not trying to diagnose CAD – they already presume this.

So what about low risk ED chest pain?
- Late 80s “accelerated diagnostic protocols” in ED to pick up mi in not classic acute mi.
- suggests this was done to reduce admission at the front door.
- also the litigation issue.

Pope 2000 is the quoted one for missed mi rate and quoted to say that missed mis have double mortality rate. Of the 11000 pts they brought ALL of them back for a 2nd visit. They found 11 mis on second visit( this was the 2%) Overall 1 in 550 was a missed MI. There is a risk adjustment done to find the mortality difference. The raw numbers are equal.

Now EST does not predict revascularisation, it causes it.

Back to the beginning. The reason the aha says do this is
- convenience
- avoid admission
- predict short term mortality/mi

It does not do that.

Now it is trying to find people whom ECG and enzymes missed.

Gibbler chest pain unit study.
- some people got EST while the ckmb was being done
- some mis were not classed as mi and they probably were.
- very few mi overall.

It did not give us the Test that we need – to exclude cad in the 2%

When you look at the outcomes remember that they’re composite and most of the composite is made up of EST causing angio.

There is a question as to whether the angios that are done actually save lives and prevent mi needs addressed

Reinforces the old message that the low risk ed chest pain is truly very low risk.

New candian study in annals shows that if you don’t do the EST it’s ok. No one dies, no one has an mi.

Using the diagnostic properties you convert 1% risk to 0.7% risk. And this isn’t even a risk of anything we’re worried about.

Filed Under: ACS ·

Implementation of a Sensitive Troponin I Assay and Risk of Recurrent Myocardial Infarction and Death in Patients With Suspected Acute Coronary Syndrome JAMA. 2011;305(12):1210-1216

June 7, 2011 by · Leave a Comment

In Edinburgh they were changing to a more sensitive trop so they spent a few months using both but not telling the docs about the new cut-off. They then introduced it and looked to see if outcomes were different

Patients

Included were all patients admitted to the hospital with suspicion of ACS.

Interventions

For the first few months every trop sent to the lab was done on the new more sensitive troponin but results were reported according to the old cut-off. The new, intermediate rage was recorded for the study but not reported to the docs. This was the validation phase.

For the next few months the new results were reported; the docs now knew that their patient had a trop between 0.05 and 0.02. Previously this would have been a -ve trop. This was the implementation phase.

Clear so far?

Results

The overall rule in rate (in terms of +ve trops) was 35% which is pretty high. I hear figures of 10% are more usual over in the US.

As expected, the new more sensitive troponin, with a lower cut-off, increased the diagnosis of MI dramatically. A +ve troponin being a key part of the diagnosis of MI.

But more interestingly are the clinical outcomes.

there was a 10% fall in death rate in people who would have previously ruled out with the old troponin. In other words people who were picked up in the new intermediate range (o.o5-0.2) had an absolute reduction in mortality of 10%.

That’s pretty powerful but the weird thing is that the death rate in the group who previously ruled in with the old trop went up about 8% when they introduced the new trop! Even though the trop range they fell into shouldn’t have been affected

It’s not totally fair to say it helped one group if it killed another (and not mention it; they just say that death and MI together were the same at 12 months!)

Also interesting that the patients who “needed” and apparently benefitted from the new trop were about 8 years older than the others. Understandably they did worse and had a higher mortality over all.

Comment

The methodology in this trial is pretty neat, providing a pretty nice control group for comparison. The fact that they measured clinical outcomes and not just diagnosis of MI means they care about the important outcomes.

The results (with both +ve and -ve changes in outcomes) might imply that there’s a base-line imbalance in the groups that predicts the mortality changes before and after.

Also worth noting that lots of +ve troponins aren’t anything to do with ACS, but are coming from PE, heart failure or sepsis. This might explain the peak of intermediate troponins in the elderly patients.

Please leave thoughts, comments or corrections below.  

Filed Under: ACS, critical appraisal, diagnostics ·