NAME: Naradah Tisdel
POSITION: Medical Technologist, Molecular Pathology, St. Luke's Episcopal Hospital, Houston, Texas
Faced with increasing workloads and shrinking budgets, clinical diagnostic laboratories are feeling the pressure to become more efficient. And while new molecular technologies designed to expedite testing are being introduced into the mix every day, many clinical diagnostic labs lack the financial resources to implement new platforms and workflows, and instead must find creative ways to work with what they have on hand.
One such organization facing these challenges is St. Luke's Episcopal Hospital, which is based at the Texas Medical Center in Houston and affiliated with the Texas Heart Institute. Known as one of the country's preeminent transplant medicine centers, SLEH and THU have since the inception of their transplant services performed more than 1,100 heart, 1,000 liver, and 1,800 kidney transplants. In addition, the centers offer left ventricular assist device implants and lung transplants.
In an effort to address an increasing demand for its services and the need for a faster turnaround time for transplant-associated virus diagnostics, technicians at SLEH developed a streamlined method of batching nucleic acid extraction and molecular testing of five such viral targets.
More specifically, the group consolidated quantitative real-time PCR protocols for Epstein-Barr virus, cytomegalovirus, and BK virus; and qualitative real-time PCR testing for herpes viruses 1 and 2, using Artus analyte-specific reagents and a Rotor-Gene Q thermal cycler from Qiagen. Using this approach, the lab was able to reduce its instrument utilization time by 33 percent, and can now perform its full menu of assays in just over 10 hours compared to the previous time of 15.5 hours out of 21 total hours available per day.
Outside of transplant services, SLEH also offers molecular diagnostic testing for a number of infectious diseases, and, like many of its clinical laboratory counterparts, was particularly overwhelmed by the H1N1 influenza outbreak of 2009 (PCR Insider, 9/6/2012).
To accommodate increased demand for flu testing following this outbreak, SLEH employed a QiaSymphony SP and QiaSymphony AS platform to automate nucleic acid extraction and assay setup in front of established PCR-based testing on Roche LightCycler instruments. Compared to the lab's previous manual assay prep methods, the QiaSymphony protocol afforded a slight but significant reduction in technician hands-on time, freeing employees up to perform other duties.
Naradah Tisdel, a medical technologist at SLEH, presented a pair of posters outlining these initiatives at the Association for Molecular Pathology annual meeting held in October in Long Beach, Calif. Following the conference, PCR Insider caught up with Tisdel to discuss his lab's efforts in more detail. Following is an edited transcript of the conversation.
Both of these posters involved Qiagen products. Has your lab been using Qiagen molecular testing products for a while?
Yes, we've been using their products for a while. We took on the QIASymphony SP in 2009 initially for the flu season – the H1N1 epidemic – and the QIASymphony AS more recently. From there, it provided a backbone for our testing, for [nucleic acid] extraction for viral load testing. And we had been using their products for manual extraction, as well.
So the goal for both of these initiatives presented in the posters was to speed up your molecular diagnostics workflow?
Yes. The [QiaSymphony AS] work was done in the hopes of not necessarily speeding up the process, but reducing hands-on time. We do have limited staffing. The streamlined protocol [of transplant-associated viral targets] was intended more to speed [workflow], because it actually combined several protocols into one, and normally we'd have to run those in kind of a staggered manner.
Regarding the viral target streamlining, these are assays that you typically test for in transplant patients?
Correct. Generally our clinicians ask for these particular targets the most out of all of them. This is kind of piggybacking on antibody-based tests that we do, which were instituted [a long time ago]. Those were the initial standards of testing before [molecular testing] to monitor treatment.
These are the main [targets], but in addition they often look at [Varicella-Zoster virus], which wasn't on the poster, but it's on our list of upcoming targets to validate.
These are all analyte-specific reagents from Qiagen for these assays? And you offer them as laboratory-developed tests?
Does multiplexing these tests change at all the way they are performed or how the results are provided?
We don't really change the test methodology, per se. We just adjusted it to make sure they ran on the same temperature profile [on the Rotor-Gene Q]. So they are multiplexed in that they [contained] the internal control and viral target, but they're not multiplexed in the sense that we put all of the reactions in the same well for every test. I would actually consider the Artus HSV kit to be the only truly multiplexed assay in this study because it detects both HSV 1 & 2 viruses and internal controls in a single reaction well.
Right, you're not really multiplexing targets in the same assay, you're consolidating several assay protocols into one.
And you found that the streamlining freed up a significant amount of technician time. So is this something you've already implemented at St. Luke's?
Yes, definitely. It's come in very handy. I think the number one impact we've seen so far besides [the reduction] in hands-on time and the instrument time … is the turnaround time, because we're able to get these results out a lot faster on days where we are offering our whole test menu. Currently we are running three thermal cyclers – a [Roche] LightCycler 1.2, [Roche LightCycler] 2.0, and the [Qiagen] Rotor-Gene Q – and normally, we basically have to wait. We put a test on and wait for the two-hour protocol to end, and then put on the next test. That can get kind of dicey when you add up all the tests in question. So this had already helped to reduce the turnaround time that we previously have incurred by offering so many tests. On the same token, it allows us to grow our test menu even further; for example … we just brought on Epstein-Barr virus with the same protocol.
Have there been any snafus or roadblocks to implementing this so far?
During the validation we occasionally had some problems, and I think the main thing we ran into was making sure that the instrument was calibrated correctly [using] the optical temperature verification kit, which Qiagen sells. That was of utmost importance from the start. We actually had several months of research that we had to trash because the temperatures just weren't correct. Running that before starting validation is really important, and after that it's really up to the user to streamline the kit to suit your needs. We're kind of lean here at St. Luke's, so we streamlined it so that we could use the smallest amount of reagents possible and still produce reliable data.
One person approached me at AMP and asked, "What's the purpose of this?" And I thought, "You obviously don’t work in a clinical lab." But he was in a research lab, and I understand that, because they [sometimes] have [more resources].
You mentioned consolidating additional transplant-associated viral tests, but would this be applicable in other areas beyond transplant testing? Just general infectious disease testing, for instance?
Actually, it would be a great benefit for a multiplexed stool pathogen assay, but for our purposes, we haven't gotten to that point yet. We can only do so much in the limited amount of time we have here. In addition to our test menu expanding, the actual volume of tests is also expanding.
But I think a stool pathogen assay would work well with this kind of format. Maybe several targets per well. I think the Rotor-Gene Q allows you to do [high-resolution melt] analysis, and there are five different channels you can choose from. [It would be great] to take advantage of its full potential.
Also, currently we run the [Becton-Dickinson] GeneOhm [assay] for [Clostridium difficile], but the [US Centers for Disease Control] is being more vigilant about hospital-acquired infections, so this might be a good way to go for stool pathogen detection in the future, because it's going to become kind of standard to help prevent the spread of diseases throughout the hospital.
Several companies are developing multiplexed stool pathogen panels, like BioFire Diagnostics and NanoSphere. But it sounds like you've been using an established set of protocols and products for a while. Is it easier to just come up with creative ways to increase the efficiency of your current methods than it is to invest in one of these new platforms?
Yes, at this point, at least until these other platforms are much more developed. This is much more well-established; the support system is there. Whenever I had any issues during the validation – for instance, the [optical temperature variation] kit – we just called Qiagen tech support and they were awesome. They helped us troubleshoot and come up with different ideas, and made suggestions as to what we might want to try as far as optimizing our experiments.
Do you sense that other labs like yours around the US are doing these kinds of creative protocols, perhaps because they are short-staffed or lacking financial resources? Did you hear from other AMP attendees that had an interest in implementing something like this, or already have?
Actually, I did. There was a guy from Children's [Hospital] in [Washington, DC], and we had a few conversations while we were there, and he said they were basically facing the same kinds of issues. The way they got through it was by automation. Qiagen's products came in handy for them, as well.
Two other people from a lab in Seattle were interested in the products because they were bringing on a Qiagen system and were looking at different options for CMV, BK/HSV, and EBV.
Your other poster focuses on implementing automation to reduce technician hands-on time. The poster noted that much of the setup for these tests was being performed manually, but was moved to the QiaSymphony AS. What's been the takeaway from this effort? The poster seemed to indicate that it afforded only a modest savings in time.
It wasn't so much the savings in time that mattered, it was the freeing up of hands-on time that mattered — setting up the actual test. For us it makes a difference, because we have nine or 10 tests on the menu now, and when you're juggling that many in the same day, that extra 15 or 20 minutes to set up 30 samples by hand is kind of crucial. It might be a good lubricant in the schedule that can help you slide between getting home on time and staying an extra hour or more.
That particular test, for Flu A and H1N1, [comprises] two separate targets. So, for instance, yesterday I had 44 samples, and the LightCycler only runs 30 plus two controls, so that extra ten minutes allowed me to do setup sheets and results from other specimens, while I could just go back, recap [a second set of samples], put them on and hit run. It definitely helps. It allows you to overlap your duties.