Skip to main content

10 years of innovation with the FDA — and more in store

Dr. John Seeger reflects on Optum contributions to the Sentinel Initiative and looks forward to the opportunities ahead.

By John D. Seeger, PharmD, DrPH Chief Scientific Officer  | July 2024 | 5-minute read

Making a difference with real-world data

Early in 2023, we spoke with Optum Chief Scientific Officer Dr. John Seeger about the ways the Sentinel Initiative — led by the Federal Drug Administration (FDA) — monitors and evaluates the safety and performance of medical products.

Now that Optum is in its 10th year of working with the Sentinel Initiative, we caught up with him on how this work has evolved over the years.


Optum Life Sciences began working with the Sentinel Initiative 10 years ago. How did that work start?

In the early 2000s, the introduction and eventual removal from the market of rofecoxib (Vioxx™) made clear that for the sake of public health, long-term risks of drug products needed to be visible not just to manufacturers but the federal government and broader public.

A systematic post-approval surveillance process was needed. This would broaden the FDA’s authority to manage the evaluation and labeling of approved products.

This is where the Optum Epidemiology team, who had a fair amount of history with those kinds of projects, came in. Sentinel started small (it used to be called “mini Sentinel”) and in many ways, Optum was destined to be part of Sentinel. The initiative set a target of incorporating the observational data of 100 million lives into the system, and Optum could provide a sizable piece of this. 

How has the Sentinel Initiative evolved since then?

There were all these data sources — large players like Optum, plus some medium and small-sized data sources — that guard their patients’ privacy carefully, as they should. There was a lot of resistance to sharing data, so it was important to set up rules and infrastructure that all participants would agree to.

This was achieved by designing the system to work without individual patient data ever leaving the sites. Instead, Sentinel coordinators send analysis specifications to each site, those analyses are completed on-site, and results are shared back with researchers in a way that can be summarized across sites.

 

""

Sentinel has become very effective at addressing the concern that brought it into being, and the structure gives the FDA and all of us insights into the safety of recently approved drugs and vaccines that we didn’t have 10 years ago. That’s the triumph of Sentinel. We can say we are now much more prepared for the next Vioxx-like incident.

What are some of the differences between the work Optum was supporting in 2014 and today?

Sentinel is built to study drugs and vaccines, and the subject of study changes all the time based on what drugs manufacturers are developing and submitting for approval, and the indications for these drugs.

That’s what guides the research agenda for Sentinel — the safety of the drugs that have been approved. Along the way, Sentinel has encountered limitations where the existing methods are insufficient, and different methods need to be applied to address a range of questions being asked about novel drugs.

For example, many cancer therapies are built around very specific biomarkers that patients have and that physicians assess before starting a patient on a new therapy. This information tends to not show up in claims data, which creates a challenge to the existing Sentinel structure. This has led to a desire to include more electronic medical records in Sentinel, since these contain much richer clinical detail.

Has the COVID-19 pandemic reshaped collaboration with the FDA?

Before the pandemic started, we became involved with the FDA BEST program, a sibling of the Sentinel Initiative focused on biologics. The biologics division at the FDA wanted information on the safety and effectiveness of COVID-19 vaccines, so we needed to find a way to shorten the data lag from the time a patient receives the treatment or vaccine to when those claims are available to us for research.

When a person sees a doctor or goes to the hospital, a claim for reimbursement is submitted, and there is often back-and-forth between the provider and the insurer to correctly correspond the claim to the patient’s insurance benefit for payment. This process can take a couple of months.

We’ve shortened that time by accessing pre-adjudicated claims data. Our epidemiology team receives the claim just days after it has been filed by the provider, so we can run analysis months earlier than we would if we had to wait for the full claims adjudication process to complete. That has been one of the big triumphs we’ve had in working with the FDA, and we’ve used this approach for vaccine safety analyses.

These rapid-cycle analyses happen every couple of weeks and provide the FDA with up-to-date information on the safety of these vaccines across a range of targeted outcomes. 

""

Beyond data lag, how did you work through other claims-related challenges you saw during the pandemic — the absence of claims data from mass vaccination sites, for example?

It’s a challenge to study something that doesn’t produce a claim, whether it’s an exposure such as a drug or vaccine, or an outcome such as a specific condition that doesn’t have a unique International Classification of Disease (ICD) code. With the unusual manifestation of stroke that occurred with the J&J COVID-19 vaccine, there was a whole issue around how we code for this.

We also wanted to distinguish between subsequent doses and to separate the different manufacturers of vaccines. This created several different codes, plus some vaccines were given out in mass vaccination sites where no claim was submitted.

How can we conduct safety surveillance of a vaccine when there’s no code for the vaccine? We learned how to deal with this issue using data gathered within information systems that are run at state and local levels across all 50 states. We learned how to link these systems to our claims data to enhance our safety and effectiveness analyses.

You participated in the Duke-Margolis Center for Health Policy’s Fifteenth Annual Sentinel Initiative Public Workshop. What were you asked to share?

We are one of the data sources for Sentinel and the FDA, and I wanted to give the perspective of having boots on the ground. What does it take for the data partner sites to execute on Sentinel and BEST? There were other attendees and presenters from the Sentinel coordinating center and the FDA itself, and they recognize the value that the data sources provide to the system.

The FDA relies on knowledgeable researchers who work within a data source and who understand how to use that data source. The FDA may have questions they want to answer and ideas on how to structure analysis, but they rely on us to provide insights regarding the data and processing to make informed decisions, especially when developing analyses that are customized to very specific research questions. They need support from teams like Optum Epidemiology.

What are you most looking forward to as you continue work with the FDA?

It’s been really gratifying to be a part of this initiative. We’re at the front edge of pharmaceutical and vaccine public health challenges facing the country, and I find it exhilarating — and I know the large team of people within Optum who have contributed to the FDA work find it exhilarating. When we get together, everyone is engaged and contributing. Pardon the epidemiology pun, but the enthusiasm is infectious. It’s a group of self-selecting people who want to be involved in this sort of work, to be part of something that matters.

It’s a treat that Optum has the resources to participate and a team of highly skilled people. We can really make a difference in people’s lives and potentially the lives of our own friends and family by uncovering something new about the safety or effectiveness of the drugs and vaccines we use. We’re proud to have earned the trust of FDA, an institution that is entrusted with keeping our drugs and vaccines safe, so they can provide the most benefit for the people in our country who use them.


Partner with our epidemiology consultants

Connect with us


About John D. Seeger 

PharmD, MPH, DrPH, FISPE, Pharmacoepidemiologist and Chief Scientific Officer for Epidemiology, Optum

John D. Seeger joined Optum in 2001 and is also an adjunct faculty at Harvard T.H. Chan School of Public Health.

Dr. Seeger’s research interests are predictors of drug therapy and the application of research methods that incorporate these predictors.

He has authored or co-authored more than 100 articles in peer-reviewed medical literature and is a long-term member, past President, and Fellow of the International Society for Pharmacoepidemiology.