Data—collected and used the right way—have the power to transform best practices in specialty pharmacy.
I recently joined Sonia Kasam, Director of Multimedia and Content Strategy for Becker’s Healthcare, to discuss how insights from real-world data and real-world evidence are transforming care delivery and clinical trials. Highlights of the podcast follow.
The terms “real-world data” and “real-world evidence” are frequently used but rarely defined. Is there a difference? Are they interchangeable?
A lot of people use the two terms interchangeably but they are very different. Real-world data (RWD) is a piece of data that has been extracted from an electronic health record (EHR), claims data—or even personal devices. There are many sources of RWD. Real-world evidence (RWE) is the actionable insight derived from those data pieces. Think of RWD as the bricks and RWE as the finished house.
What is the 21st Century Cures Act and how does it change the use of data?
The 21st Century Cures Act is legislation that was signed into law in December 2016. The purpose of the legislation is to accelerate FDA approvals and bring new drugs to patients. It is not meant to replace clinical trials but rather to enhance them.
Clinical trials have four phases; the fourth phase is post-approval studies of the general population. The first three phases are very narrow by design and don't provide a broad testing environment. Historically, as a result, you can have a drug out there for three years before problems in the general population are identified. The 21st Century Cures Act was created to reduce the time frame for approvals by using RWD and RWE to deliver information on more patients. That’s one of the legislation’s primary tenets.
Another goal is to use RWE for new indication approvals. For example, if drugs are used “off label” and you see outcomes that look favorable using RWE analysis, this enables the FDA to factor that information into a more rapid approval process for a new indication.
How much has the 21st Century Cures Act changed the landscape of new drug approvals, and what do you see going forward?
Adoption for new indications over the past five years has been slow and disappointing to many. There's a reason why this hasn’t taken off. For the first four years after the legislation went into effect, many researchers tried to use data they had sitting in their domains to create an outcome picture. But they were using claims data versus full unstructured EHR data. For example, a provider might write a narrative about how a patient is doing. This information won’t be in a structured data field where it can be easily extracted but that's where the gold is when you're mining for answers. Claims data can get you 80% of the way to an outcome but it will never get you all the way because it lacks the granularity needed. When people started looking at unstructured data sources, value emerges.
Let’s look at AstraZeneca’s TAGRISSO (osimertinib), a second-line therapy for small cell lung cancer and specific genetic subtypes. When the FDA mandated that RWE be analyzed as part of the approval process, findings showed that the drug was either equal to or better than the first-line product. The FDA re-evaluated its position and moved TAGRISSO from second line to first line, which changed the economics of the drug and the potential health outcomes of patients.
Other phase 4 clinical trials are now requiring RWE for post-approval surveillance. What you're going to see is genomic differentiation, where certain subtypes behave and respond much better to a drug than others. Since genomic biomarkers are not usually captured in distinct structured fields, you have to get into unstructured data to discover these distinctions.
One example is current research in the COVID space. There's a definite genetic component to how the virus manifests and research is underway using RWE to identify genetic markers that put patients at greater risk. The future for this use of RWE is very bright. You're going to see new laws mandating the use of RWE to drive better insights and, ultimately, better outcomes. These insights will be not only predictive in nature but also prescriptive in changing the way care is delivered.
So even though the adoption of the 21st Century Cures Act has been slow, it's likely to accelerate in the years to come?
Yes, now that a large pharmaceutical company has obtained a new indication at a much lower cost without full clinical trials, there is keen interest in looking for data sources that can provide RWD and RWE.
Since we're talking about costs, can you give me an idea of the duration and expense of traditional clinical trials? Could you also share the pros and cons of the new standard for evaluating drug safety and efficacy?
Clinical trials can take a long time. Seven years is the median and some can go longer. The expense of getting a drug to market can range from a few million dollars to more than a billion dollars. The problem with clinical trials is that you don't get a broad patient type. You're filtering out many patients based on eligibility criteria. A drug that performed well in clinical trials can have a very different effect in the real world. That's what phase 4 clinical trials are for—to see how the drug performs in the real world with many different patient types. Historically, phase 4 clinical trials are expensive because they're big. RWE can greatly expand the number of patients you're monitoring for post-approval studies at a much lower cost. That is one of the big benefits. Problem identification is also going to be huge. The FDA wants to use RWE to avoid approving drugs that have potential problems.
In what other ways can RWE be used to improve health care?
One of the benefits of RWE is that you have a much deeper understanding of the patient journey. Manufacturers are very interested in how their drugs impact the patient journey. But they’re not getting much feedback from traditional data sources, which may show that a drug was discontinued but doesn’t indicate why. Manufacturers want to know why. Was it financial? Was it an adverse effect? This information is usually found in the unstructured clinical notes.
In addition, if you're a manufacturer producing a drug for a rare disease, finding patients for clinical trials is expensive and difficult. Using RWE, you can more easily identify patients who might benefit from inclusion in a trial without the time and cost associated with hiring clinicians to manually review charts.
Finally, a crucial benefit of RWE is provider education. Let's say you have an NCI guideline for a particular type of treatment that should be used if criteria are met. You look into retrospective data and see that a provider is not following the guideline. Maybe there's a reason and maybe there's not, but you have the opportunity to talk about it. What studies have found is that many times the patient slipped through the cracks. Retrospective analysis is a great way to improve care and ensure that guidelines are being followed.
How is Acentrus using RWE in its role as network administrator for health system specialty pharmacies?
As a network administrator for many large academic medical centers, Acentrus has launched a new initiative called the Therapy-Specific Outcomes Coalition (TSOC). A number of hospitals have come together to determine what should be measured in specific disease states, with a goal of demonstrating two outcomes: clinical and cost of care. For example, if a patient is required to use a particular outsource specialty pharmacy because the payer won’t cover the drug otherwise, does the outcome look similar, worse, or better than if the patient’s hospital is allowed to fill the prescription? What is the total cost of care comparison between the two?
The hospitals where many clinical trials are conducted are often not allowed to dispense those drugs because of payer access issues. This is extremely frustrating, so instead of anecdotally talking about it, we're measuring both clinical and cost outcomes with a goal of demonstrating that there are better outcomes in hospital specialty pharmacies.
For example, the abandonment rate for specialty drugs is very high, with financial problems being the number one reason. There is anywhere from a 30%-50% abandonment rate for specialty drugs dispensed by big-box pharmacies. Many hospitals have foundation funds that support patients who can't afford out-of-pocket costs, and they have clinical liaisons to help access those funds. As a result, the abandonment rate for the same drug class when dispensed by hospital specialty pharmacies over a six-month period was 8%. This tells you that there are significant outcome differences between hospital pharmacies and big-box pharmacies.
What’s the biggest takeaway you’d like listeners to gain from this podcast?
People have a perception that data analysis is intrusive but much of it is de-identified. There's a methodology called tokenization that assigns a hexadecimal number to a patient. That way patients can be followed longitudinally throughout treatment in anonymity. There is fear about data overreach but the flipside is the tremendous value data can deliver in improving outcomes and people's lives.
