Clinical Trials Can Waste How Much Money?

In the race to develop more drugs, the pharmaceutical industry may actually be wasting a lot of money for clinical trial procedures that are not absolutely necessary. In fact, data gathered from roughly 25 percent of procedures administered to patients are not only unnecessary, but can also be traced to as much as $5 billion in overall trial costs each year, according to a new study.

How can this be? Well, there is little to no data available on the incidence and magnitude of unused and irrelevant protocol data and associated costs, according to Ken Getz, a senior research fellow at the Tufts Center for the Study of Drug Development, who presented the findings at the recent Drug Information Association annual gathering. To be clear, core procedures were defined as those that support study endpoints or safety objectives.

With this in mind, Tufts set out to quantify some of the vagaries. For instance, 18 percent, or $1.1 million, of an average protocol budget is paid to perform non-core procedures, such as those that support secondary and exploratory endpoints. But this does not include staff time to collect, manage and analyze clinical data. The non-core procedures represent roughly 20 percent of a trial budget (here is a statement from Medidata Solutions, which sponsored the study).

As an example, Tufts identified amendments made to protocols as a problem area, since these can greatly add costs since each change has a ripple effect. To be specific, each amendment adds 61 days and costs more than $450,000 to implement. Overall, Tufts found there 6.9 amendments to each protocol and 52 percent occur in Phase One, and 69 percent of trials had at least one amendment.

In fact, Tufts found that 18 percent of amendments were completely avoidable due to flaws in protocol design, or inconsistencies or errors in protocols. Another 19 percent were deemed somewhat avoidable due to difficulty recruiting patients or issues concerning feedback from investigators or trial sites. More than 115 clinical trial protocols provided by 15 drugmakers and biotechs were analyzed, by the way.

Meanwhile, costs related to physicians associated with non-core procedures total as much as $4 billion annually. Why so much? The non-core procedures require data management, monitoring and statistical analysis, among other activities. And so, if one includes these related trial costs, Tufts estimates that total non-core procedures can eventually represent an estimated $3 billion to $5 billion each year.

What to do? Avoid making so many amendments and limit the identifiable non-core procedures. To improve protocol feasibility, Tufts suggests creating internal, dedicated assessment panels and committees; obtain 'real-time' input from primary investigators, study staff, and contract research organizations; use better protocol authoring tools; and match primary and secondary endpoints with procedures to be performed.

"The study not only demonstrates the magnitude of the problem, it also offers a framework that helps companies identify which procedures to consider removing from the protocol," Getz writes us. "If the procedure is not tied to an essential study objective, then they should be open to further scrutiny and, in some case, removal."

benjamins pic thx to amagill on flickr

13 Comments

Phil L Jul 10, 2012 - 11:16am
I think the real question and issue is, why all the waste and extra costs? Certainly those doing the trials don't want to waste money? is it an education issue, meaning that they just don't know how to do clinical trials optimally so save money? OR is it a regulatory issue, whereby the FDA has some type of policy that leads companies to do this? and in that case, how can this be solved? lobbying the FDA to change their regulations OR learning how to get around their rules?
Jul 10, 2012 - 11:24am
Here's how to save money on trial costs. Cement your reputation as the cheapest company in the industry. When I was at Abbott we had a reputation for paying 25 cents on the dollar. Guess what, when push came to shove most of the investigators took the money, and that was in the day when these guys could play one company off another.

They can't do that today. Therefore it should be easier than back then. Also, cap the allowable number of screen failures, and once the cap is exceeded the PI eats the cost. Amazing the results that produces.

Jul 10, 2012 - 11:31am
BTW, most changes to protocols don't have to be amendments, but administrative changes not requiring IRB approval. The general rule is that an amendment is required only when

1) you change the sample size 2) you make a significant change to study design that affects safety or the evaluation of efficacy.

I've seen amendments for things like changing the anticoagulant in the vacutainer tubes or the amount of total blood drawn in a phase I study. In the latter case as long as you stay under the monthly total allowed by the American Red Cross you're ok. You don't need to spend a half million dollars for an amendment that draws an extra 10 cc's of blood.

For a study to be ethical, it has to serve a worthwhile purpose and it has to be well-designed. Flaws in protocol design and carrying out unnecessary procedures are unethical.
Jul 10, 2012 - 11:39am
Ed: This is reminiscent of any number of studies about software development. The extra time / money spent in getting the requirements done correctly saves 5 to 20 times as much as reworking the program / system later "to get it right."

This is something that is a human proclivity and not something to blame on the FDA - except that they want to know WHY you made the change.... Which leads to more records, Of course!

Jul 10, 2012 - 1:19pm
Here's a novel way to reduce trial costs: Design a protocol to answer a single question with one primary endpoint and at most 2-3 secondary endpoints. This is how I was taught how to write protocol before the operation was hijacked by Marketing. Nowadays protcols are all larded up with the multiple "nice to have" outcome measures from Marketing.

Here's the rub: the study is always powered to answer a single question and the sample sizes are usually too small to answer the myriads of marketing derived question. God forbid in such case that such data trends positiviely. When that happens you now have a new hypothesis that marketing demands be tested in a new fully powered pivotal trial. They will also want you to take your non statistically significant underpowererd endpoints trending in a good direction, put on your cook's hat and see if you and your statistician buddy can transform Sh*t into Shinola that can be incorporated into a detail piece.

These factors actually drive up trial costs much more than do protocol amendments.

Jul 10, 2012 - 2:37pm
How much does study design/quality truly matter when the FDA requires only 2 positive trials and unlimited attempts to reach those?
Jul 10, 2012 - 2:53pm
from the article, "To improve protocol feasibility, Tufts suggests creating internal, dedicated assessment panels and committees; obtain ‘real-time’ input from primary investigators, study staff, and contract research organizations; use better protocol authoring tools; and match primary and secondary endpoints with procedures to be performed."

I'm going to have to dust off the - basically, Clinical Trials for Dummies - book I have that was published in 1985 and record an e-book version (read it into a microphone). Tufts must have found the same book at a storage-wars auction...

Site selection is very important. Standard of care (operations) is what matters - basically a protocol is embedded seamlessly into the operations of a site. When something new is required - like collecting genetic samples - that cost goes to the Sponsor who gets a tax break for research cost.

Bean counters cannot be allowed to make decisions about cutting costs while a protocol is in progress. The "learning moment" example for the whole industry was that Phase I dose-finding study being conducted at a Parexel facility in UK. Made perfect sense to the bean counters to do all four dosing arms at the same time to cut the cost of housing people over a possible two week period.

I also have a copy of a presentation done at THE premiere biotech investor yearly pooha about how only the AEs that the Sponsor wants to collect to prove the primary endpoint need to be collected.

Everyone wants a BIG piece of the ED rent money. That's what "saving money" means...

Jul 10, 2012 - 9:39pm
what dz is referring to is the attempt by Parexel to save money on a UK study of a new class of CD 28 superagonist antibodies which nearly killed every subject in the trial. Those who survived had to get new livers.

Pennywise and (British) pound foolish.

Jul 11, 2012 - 11:23am
Apologies, Ken Getz is the person and Tufts is the "corporation" - however they do share the same personhood status.

Bottom line - bean counters can't start fiddling around with an ongoing protocol's budget while the trial is going on - too dangerous to the safety of the patients and corrupting of the science (data).

Jul 11, 2012 - 11:30am
btw - always came in under budget and in less time - gee, wonder why? :-)
Jul 14, 2012 - 2:36pm
rom the article, “To improve protocol feasibility, Tufts suggests creating internal, dedicated assessment panels and committees; obtain ‘real-time’ input from primary investigators, study staff, and contract research organizations; use better protocol authoring tools; and match primary and secondary endpoints with procedures to be performed.”

Isn't this exactly what every sponsor does. I have worked for 3 companies, and we did this with every trial

Jul 14, 2012 - 4:40pm
@spencer - Getz's info from his LinkedIn - Investor and Senior Research Fellow - wow- wears many hats. Doesn't appear that he ever worked for a Sponsor so that's why he thinks that he thought it up all by himself :-)

Investor/Owner CenterWatch Privately Held; 11-50 employees; Information Services industry April 2010 – Present (2 years 4 months)

Publisher of business information services for clinical research professionals Special Advisor/Investor Atlantic Accord June 2004 – Present (8 years 2 months)

Private equity investment, board member and advisor in health care information services businesses. Senior Research Fellow, Assistant Professor (PHCM) CSDD, Tufts Medical School March 2004 – Present (8 years 5 months)

Direct grant funded research studies looking at pharmaceutical and biotechnology company R&D management practices; and clinical research landscape trends (e.g., outsourcing, investigative sites, study volunteerism, technology adoption). Founder and Chairman CISCRP Nonprofit; 1-10 employees; Hospital & Health Care industry October 2003 – Present (8 years 10 months)

Non-profit dedicated to educating the public and patients about clinical research participation and the role that clinical trials play in advancing medical knowledge. CEO and President Thomson - CenterWatch October 1993 – March 2004 (10 years 6 months)

Founder, CEO and president of newsletter and book publishing company focusing on the global clinical trials industry. Acquired by Medical Economics/Thomson in 1998.