3 Worst Case Clinical Trial Scenarios
Andrew Rohrbaugh, Associate Director of Project Management, describes three "worst case" scenarios he has encountered during his 6 years as an IRT specialist and how we were able to help sponsors avoid potential clinical trial disasters through a proactive approach to IRT project management.
Lex Raleigh: Hello again and welcome back to Cenduit QuickCasts. I'm your host Lex Raleigh. I'm really excited this time to have with me Andrew Rohrbaug, who is a senior manager of project management at Cenduit. You've seen a lot of things go right, but unfortunately maybe a few things go wrong. That's one of the things I would like to probe a little bit today, is what are the worst case scenarios. Some of the worst things you've seen happen in a clinical trial over the course of IRT involvement and what we've been able to do to mitigate that, hopefully mitigate that. If not, what we could do more proactively in the future?
Andrew Rohrbaugh: Absolutely, I'm excited to be here and excited to talk about this specific topic. I've actually been with Cenduit for just shy of six year at this point. In that time we have certainly seen an interesting set of scenarios and challenges to overcome and to speak specifically about some of the things that we can do as an IRT specialist. What we can bring to the table, what our services bring to the table, to help, not just identify those scenarios but how we can help proactively manage them. The first one that I was going to bring to the table is actually one of the first advantages that we have when someone uses an IRT system such as Cenduit's. The fact that we track where those materials are at any given point in the study. We're able to determine if they've been released or not, if they're at a depot, if they're at a site, if they've been dispensed to a specific patient.
The first situation that I was going to cover is one that really called that to service. It was a situation where we had a study that was enrolling very quickly and they were burning through the drug much faster than they had initially thought they were going to. They realized that they needed to quickly package, label, and distribute a new batch of their drug. We worked with them to do that. We had orders that were raised urgently. We had them expedited out to sites. They were actually going to be dispensed to patients starting on a Tuesday. Monday afternoon at four o'clock our project manager received a call from the clinical team in a panic. We had a quick conference call and they actually identified that the drug, the newest batch that they had just packaged, was found to have particulate material in it. They didn't know if it was safe to dispense to patients. They needed to somehow immediately stop it from being used by any patients whatsoever.
This is when having our services, having our IRT system, really came into play. We were able to immediately identify every kit within that batch of drug, confirm that none of it had physically been dispensed to patients yet and within minutes quarantine that entire batch in order to prevent it from being dispensed to those patients. We were subsequently able to work with the sponsor for replacement kits and replacement shipments as needed. Again, in that case, because we were able to do that with our real time data we were able to prevent that worse case scenario of the already bad situation getting worse by patients dosing with this drug.
Lex Raleigh: I actually had no idea of the complexity of that. I'm glad you hit upon two things that were close to my mind is this real time accountability. I think that's one of the things that you really hit upon there is that in real time we know exactly where the drug is worldwide. We had that accountability and visibility in our IRT.
Andrew Rohrbaugh: Absolutely. Again, knowing that focus, knowing that need, being an IRT specialist, one of the things that we've done is implement new functionality in our system where it's a self-service functionality. A drug supply manager would be able to go in and complete a batch block transaction down to either a kit level, a batch or a lot number level at a site, at a depot or even globally. They can complete that transaction in real time in our system. The moment that transaction goes through, it's already taken effect, all of those kits are now blocked and you can take next steps as needed.
Lex Raleigh: That's fantastic. As a technology person myself I understand and appreciate how that technology enables it. The one thing I also want to call out of your first worst case scenario is the service element of it. Yes, we're a technology company, we provide an IRT technology, but you mentioned that the client's, your CRO's, they called you up. You have you and your team of specialists there to talk through the situation and have your expertise of IRT and industry to help resolve it. Not just fend for yourself with our technology.
Andrew Rohrbaugh: Absolutely. Our project managers know IRT is our focus, that is our specialty. That's what we do every day, that's what we do for sponsors. We bring that to the table knowing that to continue that and to bring that forward we really need to have our clients best interest in mind.
Lex Raleigh: That's great, so much opportunity, just to have the technology and someone's thinking about using and IRT, I think that's one of the key points. It's not just a piece of technology, it's not just a website that you log into. It's the experts behind that, not only the experts that built the system but also experts such as yourself who service it in real time as trials are unfolding. Well great, that a great worst case scenario and even better because it had a good ending to it.
Andrew Rohrbaugh: Absolutely. The next scenario I was going to bring up is actually one that every sponsor thinks that they can avoid but somehow invariably always seems to arise. Unless it's a very short trial, chances are every study is going to run up against an expiry date where the drug that's in the study is only labeled up to a certain date, but we need to extend that date or replace that batch of medication with a drug with a later expiry date. Again, taking our expertise into focus, being proactive with it. Not just saying, well, we've had situations and we have knowledge to handle those situations when they arise. Looking at what we can do in advance to manage those situations when they invariably do come up.
Lex Raleigh: Let me jump in right there. This is coming back to accountability. You're saying that we know where the drug is, we know what drug is available, we know it's going to expire. You're saying that the team can then take that knowledge, extrapolate forward and say we're about to have and expiry date problem. We may be able to actively monitor that and bring that to folks' attention before it comes to pass.
Andrew Rohrbaugh: Absolutely. There are several ways we can do that. The first is trying to avoid the situation altogether with proactive management. Tools within the system, both reports as well as alerts that notify the user when expiry dates are getting closer so that we can manage that. Back to your point about exactly knowing where the drug is at any given time. As that expiry date is approaching, we can identify specifically which patients have upcoming visits that will need that drug or replacements of that drug that we may need to handle. When that situation does invariably arise, we also try to take a proactive approach even earlier than that in looking at different functionalities that may prevent that or better handle that situation when it comes up. One being what we call family functionality, where there may be the same drug type but in different packaging formats. Maybe a 5 mg and a 20 mg dose. The 20 mg is what the patient requires but if we build in that family functionality, if it's not available due to that expiry date, maybe we dispense four kits of the 5 mg to equal that 20 mg dose.
Lex Raleigh: From a technology perspective it's sort of the same. The idea is that from a project management or a study design perspective, your bringing that expertise and building it in to begin with. Before the study goes live you have the capability to alter this allocation on the fly based on how the drug dispensation unfolds. What else about expiry date management, is that the extent of the worst case scenario or could it get even worse than that?
Andrew Rohrbaugh: It could get worse than that, actually. Another possible preventative measure would be a partial dispensing. Again, back to looking at the study design and seeing what things we can implement at the beginning of the trial based on our knowledge and our experience in the IRT world. With that study design, if you need four kits, can we dispense one for this particular visit? At least to get the patient through another fifteen days until that drug arrives at the site. There are certain situations where there's not, that's just not possible. When that's the case, again, based on our experience, we can even go so far as to implement a manual process. It's certainly not ideal, it's not anything that the sponsor, the sites or, to be honest, the project manager want to have to work through. However, based on our vast experience in this area and with these situations we're able to so do proactively. We can look at those reports and look at that data that's in our system. We know which patients have visits upcoming, when they are, how many patients are at risk once we identify when the new drug will reach sites or relabeling will be completed. We're able to work with our help desk and the sponsor to coordinate reaching out to sites, identifying those patients, providing documentation as needed, in order to manually dispense to those patients.
Lex Raleigh: I think that's another great example too that with my having a technology background, us being a technology company. I think something easy to overlook is the service part of it. And to even talk about a manual process, I imagine that folks when they think about going into IRT technology, the idea of a paper process or a manual process is probably far from mind. As you mentioned, as a worst case scenario these things happen. You've seen it routinely over six years and your expertise and that of the team can come in and guide that to make sure that there isn't a major study disruption. What else have you got? What else in your six years have you seen that amounted to something that could have been a disaster?
Andrew Rohrbaugh: The third worst case scenario would really be a statistician's nightmare. It's once the study is completed. You've gone through the entire process of setting up your trial. You've built it, you've run it, you've run through your worst case scenarios that happen through the life and the maintenance cycle of that particular study. Now you're at the end, you've locked your database, everyone is feeling relief. We are pulling that data out and we're getting ready to review it. You're finding that your target was, let's say a ratio of 1:1:1. One active patient for every placebo patient for every comparator patient. That's really the focus that you're looking for in order to prove your data, to prove your hypothesis from your protocol. Suddenly, you're looking at it and you realize your ratio is closer to 3:2:1 and you're scratching your head saying, "what happened, how didn't we identify this, how did we get here?" Realizing that your data might be done, your trial might be a wash because you can't use that data any more to validify what you were trying to prove.
One of the things that we have, again, from our experience and from our focus on IRT is proactive monitoring of studies. It's actually something that we base out of our data control group which is staffed entirely of biostatisticians, the same people that will be looking at the data at the end of the given study. They look at the overall design of the study and they put various proactive checks in place. Sometimes they are very simple and standardized - if the patient is on an active treatment group, making sure that at every visit they receive an active material type. If for some reason that wouldn't happen an immediate flag and alert would be raised. Sometimes they get much more specific to a trial and several are around randomization itself. Again, being biostatisticians they know that's there focus and that's certainly one of the biggest focuses on the trial.
We actually had a study, it was one on my first studies that I had inherited and managed. In that study there were three treatment groups and they were going for a 2:1:1 ratio. It was very slow enrolling and it was a block randomization where it was balanced by site. Each randomization block was assigned to a specific site. They would bring in the patient, they would fill that first row in the block and the next patient would fill in the rest of the block from there. Ideally, it enrolls quickly enough that each site fills a block and then they move on to another block. It fills out and you know you've met your 2:1:1 ratio. In this particular study it was very slow enrolling. Each site was only enrolling one to two patients.
As the study started to progress, we started to notice, and actually these proactive checks identified that about six months into the study we were tracking to close to a 4:1:1 ratio. Again, looking at the end of the study and knowing that we needed to hit that 2:1:1 or as close to that as possible, in order to prove what they were trying to prove within the protocol, it raised a flag with the proactive monitors saying the 4:1:1 and it's something that had continued to trend to this and it wasn't self correcting at that point. It got escalated to the project manager, I was actually able to reach out to the sponsor, escalate it to their statistician to allow them to review the study, review the enrollment, review the randomization and decide what changed they wanted to implement, if any, to ensure that when the study was complete it was as close to that final ratio as possible.
Lex Raleigh: That one actually resonates with me from a slightly different perspective about enrollment. I know your first example was fast enrolling but this is slow enrolling. I think in general, in clinical trials, it generally enrolls slower than people expect. They always think that we're going to enroll fast and we're going to meet our targets, but slow enrollment can be a challenge. This one hits the fact that slow enrollment can not only delay from a time and cost basis but also from a study validity basis depending on the design. I think that's all the time we have. Andy, I really appreciate you coming and chatting with us today. It was great to talk through some of these worst case scenarios and see that not only have we seen them before as specialists, but our technology and our services can help provide answers. Thinking about these things up front gets us to a better solution in the end. Thank you for your time Andy.
Andrew Rohrbaugh: Absolutely. Glad to be here. Thank you, Lex.