Compass on COVID: What test for whom and when—lab leaders talk

James Crawford, MD, PhD, professor and chair, Department of Pathology and Laboratory Medicine, and senior VP of laboratory services, Northwell Health, New York: Late April is when these discussions started, which is, with a huge therapeutic as well as survivorship population of cancer patients at Northwell, what should the testing regimen be? Our cancer programming was first in line to sort out what pretreatment testing should be. Once we had calmed down our oncologists in regard to the frequency and urgency of testing relative to when the cancer patients walked in, we were able to assure ourselves as a lab, and our oncology community, that we would be able to test every oncology patient for PCR prior to treatment, but that we did not need two tests 24 hours apart. We had to pull them off that ledge because at the time the CDC and the state of New York were big on two tests 24 hours apart.

The problem we continue to keep an eye on in the setting of high Ct values for what constitutes a positive test is the long residual viral shedding of a positive patient. The whole concept of being asympto­matic and afebrile, however many days or weeks post-resolution of symptoms, has provided assurance that cancer patients can undergo their therapy and don’t necessarily need testing if they’ve had a negative result.

We derive considerable benefit from publications coming out of Memorial Sloan Kettering. Even if a patient is PCR positive but asympto­matic, there is no detrimental effect on response to chemo or radiation therapy or recurrence of symptomatic illness. So it’s a steady-as-she-goes approach.

The other ledge we talked them off of was a PCR test 24 hours before treatment as opposed to 72. Seventy-two hours before turns into 49 hours before if the test sample is obtained late in the day. Seventy-two hours gives us the target two-day turnaround for these high-priority patients.

Where is Northwell’s cancer program now? Is it back to almost normal levels?
Dr. Crawford (Northwell): We are comfortably above normal right now for our surgeries. All chemo and radiation therapy is on full throttle and we were able to clear the backlog of the high-acuity patients through May and into the first half of June so that now we are back to what we hope is a trued-up sustained balance of early diagnosis and intervention. As all of you are experiencing, it’s a different world in which we’re performing our practice. That’s the huge impact of all of this.

Robert Carlson, can you tell us about your experience in this cancer space now?
Robert Carlson, MD, medical director, NorDx, MaineHealth: If you assume that the rate of new cancers is steady and you just delayed the diagnosis and all of a sudden you open back up, you’re going to get that bump of new diagnoses and I suspect that’s the primary cause. As our oncology practices open up, they go through the backlog and we have seen a spike in the diagnostic testing that’s done. I expect it’s going to go back to what it was pre-COVID.

Sterling Bennett, what’s your experience at Intermountain now on any of these topics?

Sterling Bennett, MD, MS, chair, Department of Pathology, Intermountain Healthcare, and medical director, Intermountain Healthcare central laboratory, Salt Lake City: Like many of the other institutions, we are back to full pre-COVID volumes in surgery and other areas. We’ve wrestled with the same questions Susan Fuhrman raised: Who do you need to treat, what are you going to do with the results? And we have triaged the various platforms that we have to different populations of patients and set up different priority levels even within a given platform. That’s all driven by the orders in the EMR. If a person orders COVID, they have to check a number of boxes to identify why they’re ordering the test, and then the order is driven to a particular platform or to a priority within a platform. That’s been a key step for us in being able to use the test ration­ally and optimize the turnaround times for those who need a result more rapidly.

Do you find that the science of testing, for lack of a better phrase, is much better understood today than it was at the end of April?
Dr. Bennett (Intermountain): Yes, by and large it is. Our senior executives are not only very interested now but also have a much better appreciation for the services we provide and for the essential nature of having certain capabilities in-house versus outsourced. That’s one benefit we’ve seen. The other is that physicians, in general, have a better understanding of the concept of predictive value. The testing isn’t simply black or white but needs to consider true positive rates, false-positive rates, prevalence, and then predictive values. So that’s been helpful, and the level of engagement we have with our physicians is the best it’s ever been in terms of test utility.

Greg Sossaman, is your experience similar in there being a greater understanding of what’s involved when you talk about tests and different test technologies?
Dr. Sossaman (Ochsner): Yes, absolutely. My colleague and I are involved in a daily call with our ID physicians and administrative colleagues, and we go through many issues. We learn if there’s any confusion from the front line or from anyone—it comes through that group and we vet these things and get communications back out. It seems like the communication lines are much more open and much tighter than in the past.

Many people are, from what they see in the media and other areas, well informed about testing struggles and what’s going on now, and we get a lot of well-informed questions about testing. Good, hard questions, though there are not a lot of answers. Our cache within the institution has gone up considerably, and we’ve also strengthened ties with our administrative leadership. It’s been great—the anti-commoditization of the laboratory. But, yes, there is more knowledge out there now.

You had mentioned the benefit of introducing rapid tests for businesses, schools, and so on. Is this greater understanding helping you feel better about that type of offering because you feel the players will understand better what you’re offering and not offering by not having a PCR test?
Dr. Sossaman (Ochsner): I think so. For a while when we first introduced the ID Now test there was skepticism about the accuracy. We’ve been able to help people understand the context in which they should use those tests, and that has opened the door to talk about antigen testing with our ID colleagues—where should those tests be performed?

John Carey, what is your reaction to what you’ve heard?

John Carey, MD, Pathology and Laboratory Medicine, Henry Ford Health System, Detroit: I am in line with what the others have said. We’re most intrigued at this point in time with the role of antigen testing. We haven’t implemented it to any extent and we’re debating now between a centralized testing plat­form versus the more rapid point-of-care approach. It depends on what the Abbott platform performance turns out to be. I don’t think we would be interested in going forth with antigen testing if it has sensitivity and specificity similar to early antigen tests. The juice isn’t worth the squeeze.

Like everybody, we’ve been fighting to have sufficient capacity for PCR and are using rapid testing in our community hospitals and in our flagship hospital to facilitate rapid admission and dealing with trauma cases and similar emergencies. Otherwise, we’re continuing to do centralized testing and we’ve been fortunate to get most of that done within 24 hours, 98 or 99 percent of the time.

From what I’m hearing, it appears there’s cause for optimism in the midst of this COVID crisis. Pamela Murphy, is that also your impression? Has the worst passed us now, and is there greater understanding? What is the situation in your area?
Pamela Murphy, PhD, APRN, system administrator, Pathology and Laboratory Medicine Integrated Clinical Center of Excellence, Medical University of South Carolina Health, Charleston: As far as positivity rates now, we’re in a lull, although many are expecting a rise in cases post Labor Day. As far as capacity now, we’re in a fairly good place, but that’s also because our volumes are low. We just got a second Abbott Alinity m, which will get us to about 2,200 to 2,500 internal tests a day. We also have three reference labs now that we can work with, although, as we know from experience, that can change quickly.

We’re getting more Abbott ID Nows in for flu season, where before we were hesitant to adopt those. And we’re working now on saliva testing. We’re looking at several different platforms for saliva testing—a Thermo Fisher assay, the CDC assay, and we’re partnering with the research arm of our academic medical center for this. We’re also looking at the Panther platform as well as the Abbott m2000 and Alinity m for saliva testing in our clinical lab. We’re still in the process of getting enough positive saliva specimens to see if saliva correlates with our NP swabs. But there’s a lot of pressure to validate and implement quickly.

The saliva news has been all over the place and veterans of the laboratory scene like me and Stan Schofield have been hearing about saliva most of our professional careers. Do you agree, Stan?
Stan Schofield (MaineHealth): We’ve been talking about validating it, but saliva is not the simple answer one would think. First, the procedures aren’t automated. Second, we were having problems getting decent saliva collection plastic devices. Third, the staff doesn’t like working with saliva. It’s thick. It’s problematic. Does it clog your liquid handling robots? Do they have to treat it? Did someone not follow the rules and have a chicken sandwich 15 minutes before they provided the sample? We’re struggling with all the relevant items. From a convenience and corporate testing perspective, it would be simple to do universities or colleges. But the other limitations are realistic: the need to validate an assay, to get the robots to handle it without clogging it up, and so on.

Bob Carlson, would you like to add something?
Dr. Carlson (MaineHealth): We looked at one container, a simple plastic one, but we had trouble with people screwing the cap on tight and leaking. We have a kit, which has a wide mouth, but it’s $11 a pop. We looked at another, which has an interesting thing you chew on; it absorbs saliva, and you put it back in and put in the centrifuge. But handling thick specimens with liquid handling systems is a real issue for us, as Stan said.

Susan Fuhrman, what is the view of saliva testing at OhioHealth?
Dr. Fuhrman (OhioHealth): We are also getting a big push to accept saliva testing. We’re bringing up the Thermo Fisher QuantStudio with automated liquid handlers, but they haven’t been delivered yet. I agree with what has been said about the specimen type; that’s the feedback I’m getting from the lab. Those liquid handlers and pipette tips are sensitive to viscous material. That’s going to be a challenge with automated pipetting systems.

Sterling Bennett, what can you tell us about saliva testing at Intermountain?

Dr. Bennett (Intermountain): There’s a lot of interest in saliva largely because it requires less PPE and skill for collection. And as the clinical volumes have returned, the people who were redirected to specimen collection have returned to their primary work.

We’ve done a couple of studies on saliva. For one, we pulled tubes off the shelf and had people drool in them at our collection site. We looked at 466 and had a 93 percent positive percent agreement. For an asympto­matic population that’s probably adequate. Then we got our hands on a commercial kit from Spectrum Solutions and we looked at 318. Interestingly, we had one false-negative, but we detected five additional cases on saliva that weren’t detected by the NP swab.

Saliva looks good from a diagnostic perspective, but we’re not certain we’re going to be able to get adequate supplies. There is a cost to it that may be offset by the reductions in PPE and use of less-skilled staff. But we’re also concerned about the issues of saliva with the automated liquid handling, and our liquid handlers are undergoing validation, so we have not yet tried them on saliva.

Is it fair to say that saliva is not yet ready for prime time in terms of high-volume testing with lab efficiency?
Dr. Bennett (Intermountain): That’s fair to say with automation, but we’re moving quickly toward it with manual processes that are in place today.

Susan Fuhrman, you nodded when I asked if it’s ready for prime time. Would you like to comment?
Dr. Fuhrman (OhioHealth): The issue is high volume. We’re doing 1,500 to 2,000 tests a day now through six platforms. And we do it similarly to what’s done at Intermountain, which is that we’re trying to get the right test for the right patient so it can be interpreted appropriately. What platform it goes to is dictated by the orders and the questions we ask on the orders; then it is sent to the right place. The saliva test is best for screening people when the right PPE isn’t available or you’re not in a hospital setting, et cetera.

If, on the other hand, you have a setup where you’re doing PCR for patients in your emergency room and turning results around within two to three hours, performing them one at a time on a less automated platform, that testing could be performed with a viscous saliva specimen. But it is going to depend on how you’re doing it.

When I look at all these tests and the Yale study, for example, there is no way I could put through the volume I have using that method. That’s kind of a research lab, low-volume method with a lot of manual pipetting. That’s where the rubber meets the road on this.

Robert Carlson, would you like to comment on the same question?

Dr. Carlson (MaineHealth): Susan nailed it. For high-volume, batched processing, using liquid handlers is going to be problematic, but it’s different in the one-off setting with a manual approach. We’re about halfway through our validation on saliva. Our positivity rate is so low that getting paired samples is a slow process. But we’ve seen they compare nicely. We’ve seen better sensitivity on some samples with saliva than some nasal swabs. So it’s not going to be a matter of sensitivity. What’s going to drive it is a matter of handling the sample.

Like Stan Schofield, I’m in contact with people on Wall Street, and in some of the startup biotech companies, they’re excited about T cell activation in COVID. Stan, can you explain what it is that is so exciting to them?
Stan Schofield (MaineHealth): They’re excited because the words T cell activation are new words they never heard before. Six months ago it was rapid testing. Three months ago it was antigen testing. The buzzword in the last six to eight weeks is saliva testing. Now people are talking about T cell activation. They’re asking who knows about it, who’s an expert, who can we call, and I’m in the middle of it all. They’re chasing the latest, greatest, hottest thing so they can try to make a buck.

Lauren Anthony, what are your latest thoughts about the conjunction of COVID and flu?
Dr. Anthony (Allina): We’re expecting that our flu PCR will be greatly affected by the COVID supply issues because the vendors are the same. And if the COVID test also becomes our flu test, it’s going to cause a lot of practice change. We’ve been working with our system expert groups and educating and planning for empirical treatment of flu. The CDC posted that you can treat empirically within 48 hours for low-risk patients and longer than that for high-risk patients. The CDC wants flu out of the picture as much as possible. So we’re educating on empiric treatment and then restricting PCR with decision-support alerts. We need the PCR for high-risk patients and anyone who is admitted to the hospital, and we want to reserve it for those patients.

Sarah Province, what is your outlook on flu and COVID?
Sarah Province (AdventHealth): We’re working through the development of an algorithm that will help triage the cases. We would start with a mini PCR panel that will have flu and SARS-CoV-2. And then from there offer a full respiratory panel that includes SARS-CoV-2 for an ICU-type patient who is symptomatic. Once this algorithm gets built, we’ll be able to know what platforms we’ll run on and how to triage these specimens.

Greg Sossaman, can you give us a recap of this discussion, which is our second round with the Compass Group?
Dr. Sossaman (Ochsner): It’s similar to the first round in that we all have seen the same challenges and struggles. The messages among our institutions seem to be the same again, though with a few different tactics since some labs are being pushed to be as efficient and as innovative as possible—for example, with the use of saliva and pooled testing. We all continue to have similar struggles but also similar optimistic outlooks that we can get through this. More testing can get us through this part of it until we get the vaccine.