ABBOTT'S NEW COVID-19 TEST

[y2k_free_radical]

Any thoughts and could use help with the best contact number- P.S.-I was just successful in reaching a good contact number

---

Members don't see this ad.

 

[Napoleon1801]

As much as Abbott drops the ball, I would not pursue. I've not been impressed by them or their service.

 

[bronx43]

Have they determined the sensitivity/specificity for this yet? I can't imagine it to be too high...

 

[gbwillner]

No clinical validation has been performed on this test. Their performance specifications were limited to contrived spiked samples.

 

[dingdong28]

I was going to make a thread about this, but glad I searched beforehand. This is kind of a long and somewhat funny read, sorry:

I'm extremely skeptical about this test considering the iSTAT recall just a couple months ago. We went live with beerbug testing with the ID NOW yesterday. The ID NOW reminds me of a Hyundai: It's shiny and somewhat sophisticated and appealing to the eye, but it is made of cheap plasitcy parts when you're touching and using it. I can talk smack about Hyundai because I own a 2013 Elantra, although I really do like the 2021 Elantra and Sonata.

So far in my short career, I've been lucky in that I've only had to use Abbott instrumentation once for flu, RSV and strep A testing. I can't remember what instrument it was, but let's say that we had these instruments for less than a year before switching over to the Cepheid for flu, RSV and strep A testing. The instruments we had from Abbott were such garbage that they had to send out 2 replacement instruments because they couldn't remedy the issues we were having. Imagine only being able to run flus, RSVs and strep A tests on only 1 instrument during flu season at 60+ ED bed/500+ bed hospital. These instruments are not user friendly in that if something is wrong with it, you cannot do any troubleshooting and end up having to discontinue testing (or send to another facility) until someone can fix or replace the instrument; the same thing applies to the Abbott ID NOW.

I think the biggest issue (for our facility) we're going to have is receiving a properly collected specimen. There have been several instances where the first swab collected in the ED was negative. After the patient was admitted and the first swab came back negative, we'd receive a 2nd swab for verification (I was so annoyed with this because I thought it was extremely wasteful) and the 2nd swab would come back positive. If these patients are being admitted without any PPE precautions or as a PUI, we can have tons of staff going in and out of the patient's room without realizing that they are being exposed to the beerbug for several days. In addition to patient exposure, I absolutely despise how open the instrument is when it comes to aspirating and dispensing your specimen into the various kits on the instrument for testing. Spilling and contamination is exponentially high with this instrument. After each use, the whole instrument must we wiped down with bleach and 70% EtOH before continuing any testing.

Is this a great resource for our clinician colleagues? Sure is since the TATs have been reduced from 12-24 hours to 1-2 hours, but how accurate and reliable are these results? How are the clinicians and nursing staff collecting these NP specimens? Now more than ever, specimens need to be properly collected to prevent any false-negative results and exposing staff/whoever to patients we think are truly negative for the beerbug.

That's my take on the instrument. I may have more in a few days, but this is all that I could think of so far. Hopefully you enjoyed hearing beerbug instead of coronavirus.

 

[LADoc00]

No clinical validation has been performed on this test. Their performance specifications were limited to contrived spiked samples.

This. In fact I have heard very very bad stuff comparing this isothermal cycler (which is why its so fast) vs. true Taq PCR like Cepheid.

Given Cepheid's NP swab sensitivity isnt great, if Abbott is worse than that sends shivers down my spine.

I love how people will push "evidence based medicine" when its convenient to them and will abandon all pretense of science when they want...

 

[dingdong28]

This. In fact I have heard very very bad stuff comparing this isothermal cycler (which is why its so fast) vs. true Taq PCR like Cepheid.

Given Cepheid's NP swab sensitivity isnt great, if Abbott is worse than that sends shivers down my spine.

I love how people will push "evidence based medicine" when its convenient to them and will abandon all pretense of science when they want...

I was going to mention something about this test platform only having 55-60% accuracy, but couldn't find the literature to back that figure up (besides the medlab subreddit) nor would I know the reason why besides the speculation of the test only taking 10 minutes for negatives and 3 minutes for positives.

 

[gbwillner]

I was going to mention something about this test platform only having 55-60% accuracy, but couldn't find the literature to back that figure up (besides the medlab subreddit) nor would I know the reason why besides the speculation of the test only taking 10 minutes for negatives and 3 minutes for positives.

This is because (if this test is based on similar methodologies like RT PCR) the test is dose dependent on the amount of DNA bound to the probe to generate a fluophore or other detectable event. If there is a lot of stuff for the probe to bind, you will see it after the appropriate number of cycles. To be negative, the detection must not be triggered after all cycles are complete. A positive test will trigger (if a test can be called mid run) as soon as the detection is made, which will be before 10 minutes.

 

[yaah]

Abbott just released a statement saying swabs in viral media should not be used because it reduces sensitivity (or something like that). Dry swabs only. The ones in our labs are coming back positive a reasonable percentage of the time. Have no way of knowing yet whether it's truly a significant underestimate or not. The validation worked fine and included swabs on known positive volunteer inpatients.

 

[dingdong28]

Abbott just released a statement saying swabs in viral media should not be used because it reduces sensitivity (or something like that). Dry swabs only. The ones in our labs are coming back positive a reasonable percentage of the time. Have no way of knowing yet whether it's truly a significant underestimate or not. The validation worked fine and included swabs on known positive volunteer inpatients.

Do you have a link for that statement or was this an emailed statement from the company? I can't find any statements on the interwebs. I asked my manager about the issue after you posted this. Funny enough, our main hospital/pathology/lab leadership already received the statement and will discuss if the community hospitals will start testing dry NP swabs tomorrow. I just read a in the Abbott procedure that the swabs should be thrown away after they've been introduced to the media/saline for 10 seconds, but no longer than 1 hour. I can't understand why a swab still being present in the specimen would be an issue.

 

[Torsed]

We inquired about getting the Abbott instrument for COVID 19 and was told we were further down the list and it would first go to hot spots like Seattle / NYC / etc. Glad we haven't got it yet given the above comments.

 

[gbwillner]

One physician I consult with told me her hospital was deploying this for all patients with respiratory issues and the results of the test would determine if staff would need PPE. She asked if this was a sound plan. I told her that even the collection of specimen only has a sensitivity of up to 85% or so, and this particular test has not been clinically validated by the manufacturer. Additionally, even with contrived samples used in the performance documentation they never tested the system near its stated LOD. AND, the positive samples used for testing (for spiking) were derived from patients with severe disease and known extremely high viral counts. The conclusion: don't bet your life on this.

 

[deleted314957]

One physician I consult with told me her hospital was deploying this for all patients with respiratory issues and the results of the test would determine if staff would need PPE. She asked if this was a sound plan. I told her that even the collection of specimen only has a sensitivity of up to 85% or so, and this particular test has not been clinically validated by the manufacturer. Additionally, even with contrived samples used in the performance documentation they never tested the system near its stated LOD. AND, the positive samples used for testing (for spiking) were derived from patients with severe disease and known extremely high viral counts. The conclusion: don't bet your life on this.

I really am very proud of how y’all are the lab experts that will eventually determine how well this works and how it will be best applied while this crap lasts. Watching from the side lines makes me feel like a retired 3rd base coach.
And, I really believe elective surgery path will be fine.


Sent from my iPad using Tapatalk

 

[dingdong28]

One physician I consult with told me her hospital was deploying this for all patients with respiratory issues and the results of the test would determine if staff would need PPE. She asked if this was a sound plan. I told her that even the collection of specimen only has a sensitivity of up to 85% or so, and this particular test has not been clinically validated by the manufacturer. Additionally, even with contrived samples used in the performance documentation they never tested the system near its stated LOD. AND, the positive samples used for testing (for spiking) were derived from patients with severe disease and known extremely high viral counts. The conclusion: don't bet your life on this.

This makes me so nervous. I’m not an expert by any means, but I hope the staff at your facility stay healthy. Today, 2 of our phlebotomists were exposed to a patient that was admitted with a negative result and came back positive 2 days later. No precautions, isolation protocols, nothing. Both of them are quarantined at home for a few days now.

Im starting to understand the rigorous process it takes for a test to become approved. I never really thought of nor bothered to read about the process one particular test goes through in order for us to start using. It’s amazing how much red tape is removed when used in an emergeny and we get this “test” that’s supposed to help. I guess I think of it as a physician requesting emergency released uncrossmatched blood, rather than waiting for a type and screen to finish up.

Can I just say you have my absolute respect for your expertise and knowledge about this testing. Thank you for sharing

 

[deleted314957]

Last paragraph;

 

[Vicinal]

From my understanding although these are RTPCR they are not quantitative but binary. I don’t know how many cycles are required by controls to define sufficient amplification for detection. On a non quantitative but binary test even amplification of small amounts of target do not have a cycle cutoff based on control runs. Isn’t it arbitrary? What is the minimum amount of target necessary when you don’t know quantity of target on the swab? Don’t you just run it for as long as it takes to get a possible detection? I did quantitative PCR for years and there was always a baseline presumed amount of target and this was assessed when the controls were set up for the number of cycles necessary for presence or absence of target when validating. This is the problem I have with these binarized tests. They have no real relation to viral load and disease state they are only demonstrating presence or absence of target.

 

[gbwillner]

From my understanding although these are RTPCR they are not quantitative but binary. I don’t know how many cycles are required by controls to define sufficient amplification for detection. On a non quantitative but binary test even amplification of small amounts of target do not have a cycle cutoff based on control runs. Isn’t it arbitrary? What is the minimum amount of target necessary when you don’t know quantity of target on the swab? Don’t you just run it for as long as it takes to get a possible detection? I did quantitative PCR for years and there was always a baseline presumed amount of target and this was assessed when the controls were set up for the number of cycles necessary for presence or absence of target when validating. This is the problem I have with these binarized tests. They have no real relation to viral load and disease state they are only demonstrating presence or absence of target.

I would think there is still a relation to viral load, even if it is not stated or reported. Based on the input material and controls, and the number of cycles needed (ct) to trigger detection, you can backwards calculate what the original amount of initial target material. The binary test may simply not tell you what ct is, but is still measured.

 

[Vicinal]

I would think there is still a relation to viral load, even if it is not stated or reported. Based on the input material and controls, and the number of cycles needed (ct) to trigger detection, you can backwards calculate what the original amount of initial target material. The binary test may simply not tell you what ct is, but is still measured.

Then why doesn’t any paper, report, data set make any comment about number of cycles to detection. What is the theoretical limit or realistic limit for of number of cycles necessary in a negative results to determine definitive negativity on a binary test. I have never seen any reports of death by COVID discussing viral load or amount of target. Wouldnt this be an important piece of data?. Does presumed viral load have any relatoinship to clinical symptoms or cause of death since the virus presumably invades epithelial cells through the widely expressed ACE2 enzyme? Or would this have any relationship to mounting a presumable immune response?

 

[gbwillner]

Then why doesn’t any paper, report, data set make any comment about number of cycles to detection. What is the theoretical limit or realistic limit for of number of cycles necessary in a negative results to determine definitive negativity on a binary test. I have never seen any reports of death by COVID discussing viral load or amount of target. Wouldnt this be an important piece of data?. Does presumed viral have any relatoinship to clinical symptoms or cause of death since the virus presumably invades epithelial cells through the wildly expressed ACE2 enzyme? Or would this have any relationship to mounting a presumable immune response?

I know several studies have looked at viral load, but to your point, the assessment between clinical severity or death and viral load is probably assumed at this point with no firm connection established to date. This has been very problematic looking at clinical studies with antiviral medications that measure viral load as a determinant of response. Additionally, this is often used for determining low long the virus can remain on surfaces or in the air, without knowing if those detected particles are actually still infectious. Viral load, by the way, is determined using the methods I outlined above.

To your first point, however, the answer is likely that this device has a proprietary and internally tested ct cutoff as part of an algorithm for a "yes" answer, so that is the output. As it is proprietary, Abbott may have a reason to not share that information, and it should not be relevant as the "yes" answer is pre-determined based on the inputs to the machine to calibrate the answer. Of course, in light of the fact that this was not very rigorously tested with COVID-19, there is possibility that this "yes" is not a very accurate "yes" in real clinical samples.

 

[Vicinal]

I know several studies have looked at viral load, but to your point, the assessment between clinical severity or death and viral load is probably assumed at this point with no firm connection established to date. This has been very problematic looking at clinical studies with antiviral medications that measure viral load as a determinant of response. Additionally, this is often used for determining low long the virus can remain on surfaces or in the air, without knowing if those detected particles are actually still infectious. Viral load, by the way, is determined using the methods I outlined above.

To your first point, however, the answer is likely that this device has a proprietary and internally tested ct cutoff as part of an algorithm for a "yes" answer, so that is the output. As it is proprietary, Abbott may have a reason to not share that information, and it should not be relevant as the "yes" answer is pre-determined based on the inputs to the machine to calibrate the answer. Of course, in light of the fact that this was not very rigorously tested with COVID-19, there is possibility that this "yes" is not a very accurate "yes" in real clinical samples.

So the ct is proprietary that makes sense but I still don’t understand how this is determined and validated in the current setting of “asymptomatic carriers” and individuals with comorbidities that can have other causes of pneumonia that leads to ARDS and death. Many patients with lack of reserve and comorbidities impacting the respiratory and cardiovascular systems succumb to pneumonia that may involve multiple microorganisms. How are investigators assessing whether COVID is among other viruses or bacteria resulting in ARDS, ventilatory dependence and death. In any compromised state any number of microorganisms can colonize and depleted reserve resulting in lethality. Especially when the majority of those dying are older. I guess investigators can extrapolate and determine quantity of viral load based on the proprietary Ct (19-26 cycles in most test I have seen) but there are so many variables here including method of procurement, the fact that this is an RNA virus and thus more readily destroyed by RNAses present everywhere. It seems like there is too much we don’t know and some of this data is extrapolated from SARS which is a start but possibly a very different virus in terms of infection and behavior. Did SARS have asymptomatic carriers? Probably not as it was much more lethal and less infectious.

 

[gbwillner]

So the ct is proprietary that makes sense but I still don’t understand how this is determined and validated in the current setting of “asymptomatic carriers” and individuals with comorbidities that can have other causes of pneumonia that leads to ARDS and death. Many patients with lack of reserve and comorbidities impacting the respiratory and cardiovascular systems succumb to pneumonia that may involve multiple microorganisms...

I think you know the answer to this question. It is not validated by the manufacturer in this setting.

 

[D1b()4]

This test appears to be fantastic: Univ. of Washington ramps up Abbott Labs’ ‘fantastic’ test for COVID-19 antibodies

 

[chooks]

Abbott has been very much a double edged sword in the acute care setting. It is being touted as a panacea, but there are significant concerns about its sensitivity and applicability to the asymptomatic carrier population. The EUA applies to patients suspected of having covid and has not been studied in the asymptomatic population as far as I know. The Abbott covid "positive" control actually contains ZERO covid sequences and is instead a flu specimen that merely verifies the algorithms are working. Again - the control doesn't even have target sequences on it!

Specimen collection is an issue and while I have been told that "Dr. Brix says nasal swabs are ok", I have a hard time believing that viral load in nasal samples in asymptomatic patients is the way to go. Meanwhile we are scrounging for anything that can be used as NP, and are even starting to take our Cepheid flu kit swabs and sterile microcontainers for use with Abbott in the situations where we are deploying Abbott in the hospital. We have yet to resort to 3D printing (3D Printed COVID-19 Test Swabs).

I have been able to push back in the acute care setting for Abbott based on our own internal validations studies (which admittedly were based on the original EUA and not the post "Oooops! Yeah you can't use UTM with this! My Bad!" Abbott world). I am losing that battle amid the desire to screen asymptomatic populations in a POC manner in non-ED/non-PUI patients. Deploying Abbott in a POC fashion gives me heartburn in thinking of environmental contamination and employee exposure. But you know "any test is better than no test!". Clinicians claim to work with nebulous clinical situations all the time. Except that this test when wrong takes out half your ICU and a few patients along the way.

Our ambulatory side is using them and while I worry for their safety, I am less concerned about performance in the non-acute setting (presumably prevalence is lower in the general population than in the hospital population) and with shelter-in-place, a false negative has far less impact to the overall health system (until those shelter-in-place orders are lifted).

Very frustrated here. We have cepheid and diasorin platform, both of which at baseline I would say have better Sn/Sp than Abbott, but not nearly enough reagents. The claim that we have enough testing supplies is a lie. The claim we have enough collection media is a lie. It is sickening. Literally.

edited for formatting and some wording.