by Erin Warner Guill on Oct 1, 2024 8:33:57 PM
Introductions (00:00)
Erin Warner Guill (Auxilius): I am Erin Warner Guill, the COO and President of Auxilius. Thank you so much, Mike, for joining us. Mike is currently the Chief Financial Officer at PMV Pharmaceuticals, where he's been since 2020.
Prior to PMV, Mike spent a decade on the Celgene finance team, culminating in the role of Executive Director of FP&A for Celgene's global R&D practice.
I thought his perspective would be very interesting because he has that broader biopharma perspective and has transitioned into the world of biotech.
One other thing I find interesting about Mike's background, and which has certainly been a benefit to us here at Auxilius, is his experience as a systems implementer and tech adopter.
From prior conversations, I know Mike has implemented everything from enterprise finance systems to point solutions like Auxilius. So, I think we'll have a great discussion, not only on clinical trial financial management and Auxilius specifically, but also on your broader experience with implementing technology.
I wanted to start by teeing up a couple of slides that I hope will frame some of the challenges our customers face when thinking about the best source for expense estimation, considerations and trade-offs of various expense sources, and then we'll dig into the calculus Mike has done to determine which is best for his organization.
About Auxilius (02:14)
Before we get into sharing best practices with you, I want to briefly explain what Auxilius is. I see a lot of familiar faces here today, but for those who don’t know, I’ll give a brief overview of the Auxilius system. Auxilius is the first-to-market FP&A and accounting product built from the ground up to manage the complexity of outsourced R&D spend.
We answer three main questions as they pertain to clinical trial financial management.
-
What do I expect to happen in my study?
This is the forward-looking forecasting question. It’s very hard to forecast clinical trials given all the changes that can occur. It’s also difficult to forecast in typical, industry-agnostic tools when you need to drive cost distribution based on actual clinical drivers. -
What has actually happened in my study?
This is the backward-looking accrual question. While it shouldn’t be as challenging as it is, we’re reliant on the outsourced ecosystem. We are under pressure during the close process, and Auxilius is purpose-built for the end-to-end workflow to determine accrual and prepaid positions at the end of each month or quarter. -
What’s driving the variance between what I expected to happen and what is actually happening?
This is crucial for driving accuracy on the accrual side, and we’ll talk about that a lot today. It’s also essential for managing the vendor ecosystem, ensuring transparency and accountability, which can be tough in this industry.
I think it’s important to note, and we won’t go in-depth on this today, but Auxilius also provides investigator intelligence.
What we recognized early on is that there isn’t enough information from vendor estimates and forecasts alone to make informed decisions about the investigator side. The Auxilius system takes in your contracted data from site CTAs and EDC data to translate, in real-time, what’s happening in that volatile variable cost bucket of investigator spend.
That’s the Auxilius system.
Now, I want to dig into the big question and topic of this webinar.
Comparing Accrual Methodologies (05:08)
Erin: Now I want to dig into the big question and topic of this webinar:
"What should I use as the source of my expense estimation at the end of any period?"
What’s fascinating about launching Auxilius and now serving upwards of sixty-five customers is that we get to see what’s happening across the industry and your peer set. You want to understand what your peers are doing. We all know the accounting regulations and filing requirements, but it can be hard to implement that guidance operationally to ensure complete and accurate accruals.
What I’m showing on the screen is a variety of expense estimate methodologies that we’ve seen employed across our customer base and the broader ecosystem.
Now I’ll caveat this by saying that this is dependent on the conversations you have internally and with your auditors.
Our big takeaway is that there are many trade-offs being made, especially when doing this work in Excel, between efficiency—getting the answer quickly in your close window—and accuracy—ensuring every single patient visit and invoice is accounted for.
You’ll note that approaches on the left-hand side may work in other industries. When we demo or talk to folks new to biotech, their inclination is that it should be easy, saying, "I can straight-line my PO, I don’t understand why everyone says this is so hard."
Similarly, using historical trends is an accepted methodology in other industries. But when you introduce the notion of an enrollment curve, that historical trend no longer works as an accurate estimate.
As we move to the right-hand side, most of our customers fall in the middle, between efficiency and accuracy. Many rely on vendor estimates without question, literally pressing "equals services cost" in Excel as the basis of their expense estimate.
Auditors will say you need to talk to your vendor, but also exercise judgment to ensure it’s correct.
While we see full reliance on vendor estimates, it’s crucial to have your own internal view to calibrate against what the vendor is reporting.
That brings us to the right-hand side of the slide. This may not be feasible in Excel, but best practice in this industry is multi-source triangulation. It’s important for each cost category to have at least two sources of data to calibrate against what your vendor is reporting and your internal forecast, especially on the investigator side, leveraging clinical data.
Mike, I’d love to hear your thoughts.
Mike: Absolutely, almost every flag resonates. Straight-lining is the easiest method, and we use it for smaller studies, like ClinPharm that doesn't require the rigor of going into patient drivers, but it doesn’t work for standard clinical trials, especially for PMV’s oncology and more complex multi-part studies. We need to move to the right for accuracy.
Erin: When we map our customers across methodology axes, there’s no single right answer. Financial maturity, clinical complexity, and study or vendor specifics all play a role.
One core factor is financial materiality. For example, is this vendor contract, investigator invoice, or overall study spend material relative to my P&L and financial reporting? Some hospitalization costs, for example, can fundamentally change a study’s cost trajectory, so they must be tracked closely.
I always defer to auditor feedback. As you switch methodologies mid-year, the key consideration is documentation to keep your auditors comfortable.
We’ve found that auditor feedback varies firm by firm, and even partner by partner, depending on their experience with biotech companies.
One big factor is availability of information. We’re optimistic that the industry norm is vendors providing granular information at least quarterly. Many organizations are now contracting upfront for this level of detail to support financial reporting.
Resource constraints are real—G&A budgets are small, while R&D gets the big funding. If you can nail this on the finance side, it benefits the clinical organization downstream. But if you don’t have a clinical ops team that can build a patient-by-patient cost tracker, you won’t be able to implement that methodology.
Finally, access to automation is key. Without a tool, you’re likely not getting that patient-by-patient view or forecasting against your vendor. While these tasks are possible in Excel, they become difficult when scaling a portfolio. Automating those tasks—whether using Auxilius, another system, or a homegrown tool—is crucial when weighing different methodologies.
Any thoughts, Mike?
Mike: No, all valid points. Definitely resonates with me.
Erin: Alright, so the last slide for me. We introduced the notion of multi-source triangulation on the right-hand side to drive accuracy in financial reporting.
The reality is, what we’ve consistently heard as a pain point is that your forecast sits over here, while accounting sits over there. FP&A may want to forecast at a rolled-up level, while accounting needs the details. FP&A might want to understand why site monitoring results in change orders, but accounting says it doesn’t matter for materiality.
Auxilius was designed with the notion that finance, accounting, and ClinOps should all benefit from reading from the same sheet of music.
Our product’s ethos is that all the information should be at the same level of granularity and accessible across the organization. In practice, that means having your forecast and actuals in the same system, with consistent recalibration.
Outside a system, it’s very hard to do this. A change order is a perfect example. FP&A may not need to update the forecast except quarterly, but accounting needs that information in time for the expense estimation.
Auxilius builds that budget data model in real-time, integrates contract changes, change orders, and out-of-scope costs from the vendor file, updates the forecast in real-time, and calibrates it against the vendor’s reporting. You make your best judgment call on the data, book the most accurate accrual, and the forecast recalibrates automatically.
This creates an ultimate feedback loop, where one system answers all those questions across the organization.
Now, we’ll have some discussion topics because Mike will talk about how he’s thought through methodologies, specifically changing methodologies mid-trial or mid-year. We’ll also show how Auxilius tees up various expense sources for judgment in expense preparation.
Lessons from PMV Pharma (17:33)
Erin: Okay. With your background in mind and the context we’ve just shared on the various methodologies, let’s dig into your approach, systems implementation, and working with your auditors.
So, let’s start with a softball question: When did you onboard Auxilius, and why?
Mike: Yeah. We started with Auxilius in late 2021. PMV had IPO’d in 2020, and our trial started that same year. We were using the spreadsheet model at first. We're a small biotech, so we didn’t have the budget for a big enterprise-level system. We went down the spreadsheet path initially. But in late 2021, we came across Auxilius. It was definitely the right system at the right time for the industry, and that’s when we started engaging. However, we also ran into the fact that 2021 was our first full 10-K filing year, so we paused while filing the K and getting our ducks in a row. By Q2 2022, we were live with Auxilius and have been using it since.
Erin: Awesome. So, to your point about using it since, I will be transparent—Mike, you were an early customer of Auxilius. I’d love to hear what has kept you in the system since those early days.
Mike: The big benefit was the improvement in processes. Maintaining an Excel file with various POs, contracts, and drivers across multiple studies or parts of a study was laborious and error-prone. What really appealed to us, even with our ClinOps team, was that we didn’t need to spend a million dollars to build an enterprise-level tool. Auxilius allowed us to replicate and improve the exact process we were using in Excel. It gave us controls. We’re a public company. We were a large filer, and now we’re a smaller filer, but the system kept controls in place. We could upload data from our systems—POs, invoices, patient-level information, and vendor files—into a tool with a consistent process. That process improvement saved us time, and it was a much better process. The reporting was consistent, and the archive of each version could be leveraged.
There were a few other benefits, but those were the main ones that kept us using it. It’s part of our tech stack now. Clinical trials are one of the most highly scrutinized areas by auditors, so you don’t want to change your process too many times. Auxilius has been perfect for that, and we’ve been using it ever since.
Erin: Great. And I loved your point about onboarding while leveraging your existing methodology. One of the big points in your implementation was ensuring we trued up to where you were from a forecast standpoint.
It would be great to hear more about transitioning to vendor reporting. How did you initially settle on a forecast approach for clinical trial expenses?
Mike: We originally evaluated several options, similar to what we saw on the earlier slide—should we straight-line for simplicity, rely on vendor estimates, or use a driver-based approach? We definitely used the driver-based approach in Excel. We brought in vendor POs, except for the CRO, and went one level down. Everyone knows those CRO budget grids can be hundreds of lines long. We weren’t going to maintain a forecast and accrual methodology at such a granular level, so we used higher-level groupings like data management, clinical monitoring, and project management, and applied drivers at that level. That’s how we approached it. We leveraged our forecasting process to drive accruals and cross-referenced reports from vendors and actual invoices.
Erin: Awesome. And, going forward, is that the approach—leveraging forecasts and cross-referencing? At one point, I know you switched CROs. Can you talk about that process and how you handled it?
Mike: Yes. We were in Phase 1 and transitioned from one CRO. That CRO used a milestone-based contract. You either have milestone, time and materials, or a hybrid approach. The milestone reports from that CRO didn’t give us the granularity we needed to tie back to the original budget grid, which caused issues.
We transitioned to another vendor, where we had the luxury of asking for more granular, monthly reports. We received a detailed monthly report with hundreds of lines, which allowed us to review everything. It was tricky, though, because the CRO manages investigator grants, and you need to close out with the previous vendor. Auditors want you to reconcile with the old vendor and ensure there’s a specific cutover date. As you can imagine, investigator grants can still trickle in with pass-through costs from the old vendor. We set a cutover date, but managing two CRO contracts added complexity.
It’s definitely not recommended if you don’t have to do it, but operationally, we wanted to, and as a finance team, we figured out how to absorb it into our process.
So, the potential silver lining is that switching enabled you to get granular reporting.
Mike: Yes, and switching to a CRO that provided that more granular level of detail was crucial.
Erin: Mike, you make a great point about milestone-based contracts. Our early adviser, David Fenske, always talked about the ebb and flow between unit-based grids and milestones. The best advice we received from our customers was to ensure that even for milestone budgets, they’re giving a sense upfront of the units going into those milestones.
Also, full credit to Tony Carita from Danforth Advisors—he always emphasized that milestone-based contracts are tough for accountants to track for accruals and vendor accountability.
Mike: Yes, absolutely. You go into it thinking, "the milestone is what it is," but then change orders come in, and suddenly, justifying them becomes tricky.
I always find it interesting when you ask for unit-based backup on milestones and aren’t given it, but as soon as a change order comes in, they can tell you exactly how many units are involved. So, you know that information exists somewhere.
Erin: Yes, that’s very true. Also, Tony Kurita from Danforth is no longer there, but he is like the godfather of clinical business operations. His wisdom on vendor accountability in this industry is invaluable.
We have a thought leadership topic we did with him years back, and I can share that.
Mike, do you want to share how many studies you have?
Mike: Sure. We have one study with three parts—a Phase 1, a 1b, and we’re now in Phase 2. It’s the same protocol, so we manage it as one, but it’s a multi-part study.
Erin: Alright.
Let’s move on. Mike, thank you for providing great insights.
Getting Auditors Comfortable With Auxilius (29:08)
Erin: let’s talk through getting auditors comfortable with the system.
You mentioned implementing enterprise-level systems at Celgene. Especially when you were implementing Auxilius, there was probably less name recognition at that point. I think now most of the audit community is aware of us. Let’s talk through how you went about that in general and some of the documentation you did around your expense methodology in Auxilius.
Mike: Absolutely. It was definitely in the early days, and not only was it about getting auditors comfortable with the tool, but also switching mid-study, which is tricky in itself.
You have algorithms that you originally used to spread costs, whether they were inception-to-date calculations or drivers. When you switch tools, auditors want to know, “What’s different?”
The good thing our auditors acknowledged was that any clinical accrual is an estimate. It’s not perfect science. No one can guess exactly how many patients you’ll have or how the costs will come in. So it’s an estimate, and that’s really what any Excel-based system or tool does for your accrual.
What they wanted to know most was the logic change. We had a similar but slightly different approach. In Excel, our model went back to the beginning of the study, using inception-to-date calculations. Sometimes, there’d be a retroactive effect of that versus Auxilius, which locks prior periods and allows the accrual to calculate from that point forward. This is actually more indicative of how a change order or new cost would flow when processed.
I actually prefer it because it spreads out future time periods. But the auditors had to get comfortable that it was similar but different logic and that the differences were okay. The way we went about it was by doing two months of parallel processing—one with the old methodology, one with Auxilius—then flipped it.
We showed each of the major drivers, like patients or different milestone drivers, and how the formula worked similarly. While there were differences, we noted them and explained why, and the auditors were comfortable.
The second part they liked were the controls around the system. You’re no longer dealing with an Excel file where finance or ClinOps might be manipulating data points. With Auxilius’ workflow processes, ClinOps has their dashboard to update information, and finance is responsible for drivers. We showed them that, and they were able to produce reports.
The last thing we did, for both processes, was confirm that everything tied back to our accounting system. Whether we were leveraging POs or invoices, we made sure those two tied out, and instead of Excel tying back, Auxilius now tied back. Once we completed two months in parallel, they were comfortable.
They blessed it. There were minor differences in the numbers, so there was a slight adjustment during the transition period, but then we were fine moving forward.
Erin: That’s helpful. I should also note that was before Auxilius’ SOC 1 Type 2 certification, which would have made Mike’s life a little easier, and now that’s available.
Collaborating With ClinOps (33:17)
Erin: And then, Mike, I’d love to understand because, when we first started out, we were cognizant of the bandwidth of the ClinOps team. Initially, we were mainly having finance interact with the tool. I’d love to hear any insights on how ClinOps is now interacting with the tool and how you got them comfortable.
Mike: Yeah. ClinOps are busy, especially in a small biotech. They’re doing a lot to run the trials. The last thing they want is finance hitting them every month, especially since we do a hard close every month.
We wanted to make sure they adopted the tool as much as we did. So, we had them at the table from the beginning. They were part of the meetings.
They update patient and site information directly into a table. We also have feeds from our EDC that we cross-reference, and we double-check that information. We generally do those checks quarterly for validation and confirmation.
For the most part, the ClinOps team is in the system monthly, just updating. It probably takes them less than an hour to update.
They’re in the system and use it for some of the reports. We’ve exported investigator grant grids to help them do some analysis at one point. The nice thing is, all of that’s in the tool—your contracts, your information, all in one spot. It’s nice to be able to go in and get that type of information.
Erin: I remember you saying at one point, "This is great structured data." When we were looking at the by-visit cost at each of your institutions, you mentioned, "You’ve got to make this table downloadable. This in itself is a data asset." So, we pushed that export button in the next sprint and updated our demos, showcasing that incredible data asset in the system. Always helpful feedback. Always great user feedback.
Closing (35:36)
Answering a question in the chat: We generally sit within that ecosystem between the clinical tool set and the ERP. We integrate with the ERP to pull invoices, keying off the PO infrastructure.
We’re pulling in data from the clinical tool sets to drive some of that insight. The JEs are available for export and import into your ERP.
Also, generally speaking, what you probably noticed on each of the screens was the idea of attribute tagging. We give a lot of flexibility with tagging at whatever level of granularity and attributes you want. That gives our system the flexibility to roll this data up, and a lot of our customers are exporting this data and importing it into their planning tool.
I’ve heard from various CFOs who say, "I’m not really sure why I am investing in an industry-agnostic planning tool when eighty percent of my P&L is comprised of R&D spend."
I’d argue planning tools are probably inevitable across the board, but the Auxilius system will give you that granular view on the clinical trial side. So, what we see even for folks that have a planning tool or FP&A tool stack, they’re still doing that planning in Excel if they don’t have Auxilius.
Mike: Yeah. I can confirm that. We have a separate FP&A planning tool, and that complexity—you wouldn’t want to try to replicate everything Auxilius is doing inside any of the standard FP&A tools. It’s too much of a challenge and too granular. Auxilius really helps you with your accruals and forecasts. It’s easy to take a summary report out of Auxilius, load it into a planning tool for a project or clinical trial, and then go back to that tool for the details.
Chris, we’d love to add recruitment fee calculations.
I can’t speak exactly on where that sits within the roadmap, but we have a lot of flexibility in doing that. I’d love to connect with you offline about it.
Alright, this has been a helpful session. Yes, a lot of audit trail information is captured.
Hopefully, we answered folks’ questions. Please follow up with any other questions, and thanks again to Mike. Always appreciate it.
Mike: Thank you for having me.
Erin: Take care. Bye.