Info & Opinion
April 24, 2019
With news about big data, Aetna, Covance, GNS, Scrip, Quintiles, PPD, Icon, BioClinica, Merge, Medidata and GSK.
With news about FDA, CTMS, PMG Research, Inclinix, EMA, Hemofarm, Parexel and the Korea Drug Development Fund
David Underwood of Quanticate says some firms are giving short shrift to the basics of clinical trials
With news about Roche, Quintiles, Allscripts, Janssen, SGS, Oracle, TriReme, OpenClinica and FDA
Husbanding drug supply as judiciously as possible is a no brainer. Often as not, even experienced trial managers throw a few numbers and assumptions into an Excel spreadsheet—and hope for the best.
But what about customs? Where are your depots? Is there a central randomization of patients? What is the visit schedule? In such cases, says Ed Tourtellotte, president and CEO of Tourtellotte Solutions, a spreadsheet is not going to be enough. A serious simulation and modeling tool is needed. “Most people are not that good with spreadsheets,” he says. “They are putting in random fudge factors.”
Using The Seat of Your Pants?
“The amount of supply that is needed and wasted can be greatly impacted by these decisions,” he says. Tourtellotte says it’s getting more common for supply and logistics people to become involved early in the planning of a trial.
That’s a trend he endorses, partly because of the rare but expensive disruptions that misgauging drug supplies can set in motion.
Having to turn a patient away because of a lack of supply is rare. Suddenly mobilizing an entire clinical operations team to avoid that scenario is all too common. “The costs are hard to quantify,” Tourtellotte notes. “But they’re all staggering.”
By the same token, the amount saved by more judicious consideration of drug supply (and the right systems to anticipate supply issues) can save large sums. In one case history on his web site, Tourtellotte estimates his firm helped to save one unidentified sponsor $13 million.
He is not comfortable, however, with the word “overage.” Too often, he says, it’s synonymous with “waste.” There can be an assumption that 100 percent overage or some other rule of thumb is “safe” or “standard.” In fact, he argues, the particulars of a trial dictate the proper levels of supply. Some overage is prudent and necessary. And leftovers above that are lamentable and avoidable.
The “O” Word
Tourtellotte says too many people in the industry have a magic number (whether it’s 10 percent or 40 percent) they believe is appropriate for any trial.
“They’ll say, ‘I’m already giving you 100% overage, why do you need more?’ To use overage as the primary way you think about your supply doesn’t do service to what is actually happening. All trials are different. Your overage is meaningless for planning. It could be entirely reasonable to need 400 percent overage. In my mind, what you really waste is whatever is left after the necessary overage you use.”
His company’s software, tcVisualize, is by his estimation just one of two products in the industry capable of modeling and simulation beyond the simple arithmetic possible in Excel. (The other is a services/consulting engagement from ClinPhone.)
In contrast, Tourtellotte’s tcVisualize is software that users can run (and probably play with) on their own. The application has a large number of dials and knobs that allow users to both forecast what is about to happen in a trial—and to run an instant replay of what just happened in the recent past.
“You can see the impact of your possible decision,” says Tourtellotte. “You may not care, but you should know. The trial design will determine how much you waste. A short visit schedule can wreak havoc with the amount of supply you’re going to need. You don’t know up front which factor is going to do what until you look at them all together.”
Monte Carlo & Pharma
Yes, there is a healthy amount of math under the hood of his tool. There are all sorts of algorithms to simulate a trial a thousand times in silico; the program’s world map can show your gyrating stocks at clinical sites in every country where your trial is recruiting patients.
But the underlying goal is to understand how each assumption, however minor or mundane, can affect levels of supply. “You might want to know a central rand is going to cost you a 200% overage,” he hypothesizes.
Obviously, the software can handle any number of real-world scenarios. You want to add a country? Two? Fine. You need to change the dosage or the packaging? Wonderful. But let the software show you the ripple effects of that decision.
“Protocol designers tend not to include supply until the later stages,” Tourtellotte sighs. But that’s changing. “There is a growing realization in the industry that you do have to simulate.”
Of course, simulation engines for drug supply cannot easily operate in isolation. Often as not, some other system provides some of the data for true electronic management of drug supply.
“Integration is actually quite easy as long as you know where your data is,” says Tourtellotte. “I would think the best place is interactive voice response. You could also get it from a clinical trial management system. You could also get it from electronic data capture.”