The Randomized Clinical Trial - Is It Time To Think Bigger?

By Ed Miseta, Chief Editor, Clinical Leader

The main purpose of a clinical trial is to generate data that will meet a specific end point and satisfy regulators. To accomplish those goals, the industry has relied primarily on the randomized clinical trial (RCT). The reason for this is obvious: the process is both strong and rigorous, and has a long history in clinical trials. But it also has its drawbacks: RCTs are expensive and each trial answers just one question. This has some in the industry asking if it is time to consider other options.

"Today there are tools available that generate good evidence but don't rely on randomization," says Nancy Dreyer, Global Chief of Scientific Affairs and SVP, Real World & Late Phase Research for Quintiles. "These tools are rigorous and they meet the broader needs of other stakeholders, including payers, clinicians, and patients."

Dreyer is a scientist at heart. She knows about research, data generation, and publications. In fact, the professional journal Epidemiology was founded by Dreyer. She certainly understands the process of generating a primary research question, hypothesis, and RCT. Secondary objectives can be considered in those cases, but they are not deemed as valid as the primary. Today, in a non-interventional study, there might be three questions, all of which are the main focus. It is legitimate to draw valid conclusions out of all three of those questions, as the paradigm is quite different.

Nancy Dreyer, Global Chief of Scientific Affairs and SVP, Real World & Late Phase Research, Quintiles

"Putting this model in place is a two-part process," says Dreyer. "The first part is to know what it is you are trying to do, and the second part is selecting the right tool. If you are bringing a new product to market, then you do need a RCT. But if you are developing evidence to show payers the value of your product, then perhaps you don't need a RCT. You can then look into whether the data exists somewhere, whether there is a registry, are there electronic health records or some other tool that might address it, or do you need to collect the data. Sometimes you can find most of the data and then supplement it with some direct data. Instead of going right to a RCT, you can go down a structured list of questions like that and find the best way to answer them."

Avoid The Common Mistakes

When trying to implement a non-interventional study, there are a couple common mistakes that Dreyer recommends you avoid. The first is more of a Big Pharma problem: in a non-interventional study, you will always have some missing data. You need to expect that and make plans for how to handle it.

"Everyone understands a RCT," she says. "They involve paying investigators a lot of money. With a non-interventional study what they are paying for is their time. When a CRO hits them with questions, there is no incentive to answer them. This is where the problem of missing data arises, and where it becomes important to know what is critical and what isn't. The success of the non-interventional study is dependent on differentiating between the information that you absolutely must have and the information that would simply be nice to have."

Second, Dreyer believes in a non-interventional study, it is important to understand why patients come to be treated. When you bring a product to market, why will someone want to use it? There are several reasons. Maybe your product is first to market and there are no other treatments. Maybe your product is better than existing treatments. Maybe your company has just done a phenomenal job of marketing the product. In many cases, the reason is some patients are resistant to the existing treatments.

Dreyer cites asthma as an example. A company bringing a new asthma medicine to market will attract the sickest patients of a disease that has a high mortality rate. If some who take the new medicine expire, safety reports may indicate the drug is causing the deaths, not the illness itself. The folks initially trying the product may not be the target audience, and firms need to understand that. Some amount of time should be spent trying to understand the target population.

Listen To The Patient Voice

Getting the patient involved in the design is important as well. Dreyer notes we want to end up with the right outcome, but an outcome that makes a difference to the doctors and patients as well as the company. Of course understanding the patient voice is not as simple as making them a part of the discussion. It involves understanding systematic research in the real world. Patients with different ailments will neither take medicines the same way nor be affected by them in the same manner.

"You also have to think about demographic factors," says Dreyer. "Patients in different age brackets are very different. There are instances where you have to start asking questions such as are you able to dress yourself? Are you able to get out of the house? Are you able to drive? Are you able to go to work? Those things drive healthcare coverage. We can talk about the optimal patient and the optimal setting, but ultimately it comes down to dealing with someone’s mother or spouse or child and helping them understand goals and possible outcomes. There are many factors important to patients and families that need to be considered."

Looking ahead, one trend that Dreyer finds herself being most excited about are electronic health records (EHRs). These records have the potential to generate better data for trials while at the same time further improving patient engagement. “We are getting the most data we can from EHRs, and then supplementing it with a few questions for the doctor and the patient,” she says. “So from the data we will have all of this information about how the patient is doing based on specific metrics. A typical question we might ask the patient whether they feel better or if they are more active. If we had one question to ask the physician, it might be why they decided to change treatments. Answers to those questions will never be found in the data contained in the EHR. This is how I believe we will gather information in the future.”

The switch to non-interventional studies will not happen overnight, and Dreyer acknowledges that. One of the first steps to making that transition will be education. A common misconception is that these types of studies are messy and not worth the time. Dreyer believes an effort will have to be made to help the industry better understand and design them. Dreyer might be the perfect person to lead that charge, as she literally wrote the book on this subject. Registries For Evaluating Patient Outcomes was penned by Dreyer for the Agency For Healthcare Research and Quality. The book, first published in 2007, has a third edition the hit the market in April 2014. It has been translated into Chinese and Korean, and has been cited by the European Medicines Agency. Dreyer was recently contracted to write the fourth edition.

“It is filled with problems that have been encountered and how they were solved,” adds Dreyer. “I feel it demonstrates how important this non-interventional tool is, and the amount of interest in it that exists worldwide. A lot of work still needs to be done, but this is definitely the wave of the future.”