Hey Mik (and other people who know EHR systems)
-
Question about the big EHR systems, say Epic....
I have a question about the extent to which they are configurable at the site or installation level. But let me give some background first.
One of the things my foundation does is to try to increase detection. We think there are between 100-200k people in the US with my little disease, but only ~10-20k of them are diagnosed. One main reason is that the symptoms look similar to much more common diseases. If someone has Alpha-1 lung disease but is late middle aged and smoked 30 years, the Dr will assume the symptoms are "vanilla" COPD and not test for Alpha-1. Thus that patient gets substandard care - treated for COPD but not the underlying disease.
We have some data that suggest 1~2% of people with COPD have full blown Alpha-1, and several fold more have at one abnormal allele.
We successfully got the Global Initiative for Chronic Obstructive Lung Disease (GOLD) to put into their recommendations that ALL COPD patients be tested for Alpha. Few pulmonologists do this, and almost no GPs do it (and GPs are often the point of care for COPD patients).
So one idea we've had is to get Epic et al to recommend a test for Alpha when a COPD diagnosis is entered into the system. To nudge the physicians in other words.
Ideally we would get the big EHR vendors to implement something like this, but that might be tough to pull off.
Is this the kind of thing that could be part of a local configuration? If so, we could conceivably get a mid-sized health system to make the change as a proof-of-concept to collect data we might use to convince others.
Thoughts?
@jon-nyc said in Hey Mik (and other people who know EHR systems):
We have some data that suggest ~2% of people with COPD have full blown Alpha-1
That's amazing.
-
@jon-nyc said in Hey Mik (and other people who know EHR systems):
We have some data that suggest ~2% of people with COPD have full blown Alpha-1
That's amazing.
-
It’s certainly configurable to make that recommended protocol. The EHR vendor would probably be the place to do so, not the individual health system. You would have to get consensus on it which would be difficult. But any health system putting it in an order set for COPD would be no big deal.
But That’s not where the issue lies. The testing is genetic, and quite expensive I’d imagine. Could one test screen for the various conditions you describe?
I think your obstacle is probably more payor than physician related.
That said, I support your effort. Is there any symptom , especially earlier, that would lead a physician to suspect Alpha 1?
-
It’s certainly configurable to make that recommended protocol. The EHR vendor would probably be the place to do so, not the individual health system. You would have to get consensus on it which would be difficult. But any health system putting it in an order set for COPD would be no big deal.
But That’s not where the issue lies. The testing is genetic, and quite expensive I’d imagine. Could one test screen for the various conditions you describe?
I think your obstacle is probably more payor than physician related.
That said, I support your effort. Is there any symptom , especially earlier, that would lead a physician to suspect Alpha 1?
@Mik said in Hey Mik (and other people who know EHR systems):
The testing is genetic, and quite expensive I’d imagine.
I forgot to comment.
There is the problem. At the moment it's expensive, although it'll probably become accessible in the future.
Look at BRCA testing.
-
I’m working on BCRA now. Interesting. There are so many ways you can assess a patient’s risk. Luckily I only have to get the family history across and the assessment back.
“I am fond of pigs. Dogs look up to us. Cats look down on us. Pigs treat us as equals.” ~Winston S. Churchill
-
Actually the initial screen is usually a blood test - cheap. It checks for levels of AAT in the blood, doesn't perfectly correlate with genotype but come enough for government work.
Also one of the treatment providers just got approval for a DTC cheek swab genomic test that is free to the end user. The same test has been free to healthcare providers for some time. They pay for it since each patient on product is about 100-150k top line revenue per year.
-
Actually the initial screen is usually a blood test - cheap. It checks for levels of AAT in the blood, doesn't perfectly correlate with genotype but come enough for government work.
Also one of the treatment providers just got approval for a DTC cheek swab genomic test that is free to the end user. The same test has been free to healthcare providers for some time. They pay for it since each patient on product is about 100-150k top line revenue per year.
-
@Mik said in Hey Mik (and other people who know EHR systems):
Seems pretty cut and dried.
Thank you for writing 'cut and dried instead of 'cut and dry' as is taking over the dang world.
Grump.
-
@Mik said in Hey Mik (and other people who know EHR systems):
Seems pretty cut and dried.
Thank you for writing 'cut and dried instead of 'cut and dry' as is taking over the dang world.
Grump.
@Catseye3 said in Hey Mik (and other people who know EHR systems):
Grump.
I could care less about your opinion.
(this is how threads go down the crapper at TNCR)
-
@Catseye3 said in Hey Mik (and other people who know EHR systems):
Grump.
I could care less about your opinion.
(this is how threads go down the crapper at TNCR)
-
And now, for something a bit different...
Small labs are not going to do AAT testing without volume. The screening test is going to be a nephelometric method based on turbidity, looking for the absence or presence of the protein.
A typical low to mid volume lab box that would do the screening test is this:
https://www.beckmancoulter.com/en/products/chemistry/au480
I assume any positive result would be followed by PCR for a definitive diagnosis.
It's kind of a chicken and egg thing. If the volume is there, most mid-level hospital labs could offer the test. But right now, they don't have the volume, and volume depends on doctor ordering patterns.
University systems are often where ordering patterns change. You get those hospitals and med schools to change, and it changes the world. I saw it with A1c testing, and then the drive to standardize the test. Once the ball got rolling, even smaller labs offer it as a routine test.
At what point does it become economically viable for a hospital lab to offer a chemistry/special chemistry test? The old rule of thumb used to be twenty patient specimens/month. I suspect that number is probably 30-40 nowadays.
That's my side of the court, at least for testing done localky, in-house. What I really don't have a clue about is what would be the need for the screening test in any basic patient population? If fully educated, can physician ordering patterns generate enough volume or will it remain a reference lab test in most areas? Prior experience tells me the faster the TAT and the ease of interacting with the lab down the hallway or street, means any test will be ordered more.
-
Jon, what's the accuracy of the AAT test? Is it more likely to throw false positives or false negatives?
Complicated answer. Most of the inaccuracies are false negatives due to the complexity of the disease and the way the tests work. Most genetic tests look for certain common mutations - 23andme finds the S and Z mutation only, the Grifols test used to look for 5 alleles, the one they just got approved looks for 14.
Two problems - you can have a rare allele that is not screened for by these tests (300 have been identified! But almost all are extremely rare). Also, there exists what they call the 'null' allele - literally there is no encoding for the AAT protein at all in these people's genome. They have the worst lung outcomes - no protection whatsoever from elastase damage - but come back 'clean' from a genetic test (since the genetic test didn't find the sequences it was looking for).
The more simple test is a blood test to check serum levels of AAT - how many micro moles of AAT per deciliter of blood? Seems straightforward, but there are two issues, one pretty common, one less so.
First of all, AAT is an "acute phase reactant". When you have any lung inflammation, your body instructs your liver to make a lot more. So you might have well above your 'normal' amount at the moment of a blood test. The could put someone who is mildly deficient (say an MZ) look like they have normal levels (MM).
The second problem is some very rare mutations don't decrease the amount of AAT, it just isn't effective (i.e., it doesn't work well at inhibiting neutrophil elastase in the lungs). Thus a serum level check makes you look normal, but the AAT functionally doesn't work at all.
The way around all of this would be to do a functional assay along side of a genetic test - if you find a deficiency in function, but the genetic test looks clean, keep digging. This happens, but not at scale. Usually by a single scientist in a lab when they're trying to figure out a strange case.
Another complication relevant to newborn testing (and one of the reasons we can't get on the newborn screening panels) is that AAT crosses the placental barrier, so there's lots of the mom's AAT in the kid's bloodstream in the early days of life, leading to very messy readings.
As for false positives because of spurious results in one or another type of machine, I haven't heard of it being a big problem.
-
So the AAT test can throw inaccurate results (don't many tests?).
What percentage of Alpha 1 cases would we expect to be found through the AAT testing on suspected COPD patients?
Hell, my mom's supposed COPD was so severe, it might have been that. Pretty early onset for COPD - about 45.
-
So the AAT test can throw inaccurate results (don't many tests?).
What percentage of Alpha 1 cases would we expect to be found through the AAT testing on suspected COPD patients?
Hell, my mom's supposed COPD was so severe, it might have been that. Pretty early onset for COPD - about 45.
@Mik It could easily have been, especially a single allele (MZ, where M is 'normal' and Z is severe deficient). MZ people have about 5-6x the risk of lung disease of MM.
Several studies have shown 1-2% of COPD (and bronchiectasis) patients having severe deficiency (ZZ). And mid-single digit percentages having one deficient allele (MZ).
-
I’m working on BCRA now. Interesting. There are so many ways you can assess a patient’s risk. Luckily I only have to get the family history across and the assessment back.
@Mik said in Hey Mik (and other people who know EHR systems):
I’m working on BCRA now. Interesting. There are so many ways you can assess a patient’s risk. Luckily I only have to get the family history across and the assessment back.
come here. if you are ashkenazi and have boobs, the test is free and doesnt need any genetic counseling. just drop by your local clinic and get a blood test on the spot.
-
@Mik said in Hey Mik (and other people who know EHR systems):
I’m working on BCRA now. Interesting. There are so many ways you can assess a patient’s risk. Luckily I only have to get the family history across and the assessment back.
come here. if you are ashkenazi and have boobs, the test is free and doesnt need any genetic counseling. just drop by your local clinic and get a blood test on the spot.
@bachophile But they charge for sherpardim?
