A New Way To Find Side Effect Info: Brian Explains
14 CommentsBy Ed Silverman // October 11th, 2011 // 7:05 am
Anyone who has attempted to wade through the FDA Adverse Events Reporting System can readily list its shortcomings, starting with difficult to navigate. But faced with a personal reason for mining the data, Brian Overstreet and a partner came up with some algorithms which they have developed into their own web site for finding drug side effects. Called AdverseEvents.com, the site has just launched and might force some change in the pharmaceutical industry and among physicians. We spoke with Overstreet about his new, alternative universe. This is an excerpt…
Pharmalot: How did you get started on this?
Overstreet: Well, the background is that, in 1999, I’d started Sagient Research…and ran that for 10 years. I stepped down at the end of 2009. And almost immediately after I left, one of my partners there (Robert Kyle), told me his wife had gotten very sick from an adverse drug event. So like anyone would, we went to Google and searched on the drug and tried to figure out what was going on. But we couldn’t find any relevant information on side effect incident rates or what the outcomes were. The only data we could find after a lot of searching was housed by the FDA in AERS (the Adverse Events Reporting System). It’s not an open search capability, though. You have to download unstructured data files and know what to do with them to make them useful. Fortunately, that was our business and so we didn’t think twice about downloading them and tried to run searches. But what we found very quickly is that it couldn’t be done. The database the FDA maintains is so poorly structured that you couldn’t run any search successfully..
Pharmalot: So from this AdverseEvents.com was born?
Overstreet: Yes, although it took about a month to restructure the database. But the bigger problem was that the actual data in the database was corrupted – there a ton of misspellings, key stoke errors, padded spaces. There was no standardization of data in any field. That’s when the light bulb went off and we realized how big this potential problem was. There’s no way they (the FDA) could know what was going on with drug safety in this country. And it’s taken us 18 months to fix the data. We built a proprietary algorithm that goes into each field so data is standardized and normalized…Anyway, as we got deeper into the problem, we realized how bad the situation was and we also realized we had the background and expertise to fix it…
We think it’s the only system that can tell people find the incidence rates and outcomes for adverse events… The FDA has always had this data and as far as I can tell has always made it available, but taking it and putting it into useful, cohesive format takes a huge exercise. ..It was born out of necessity and we wanted the info and we were just able to do it…
Pharmalot: I would imagine some drugmakers were not happy to hear this.
Overstreet: We’re not trying to be activists. We’re not trying to be Chicken Little claiming the sky is falling. But what we are saying is that having information is not in and of itself a negative thing. There will be cases where a pharmaceutical manufacturer will be less happy data is available on their drugs and cases where they are more happy. We’re releasing something in two weeks, comparative data on statins, and that will be good for some and less so for others. But I don’t think, in this day and age, that you can make a cogent argument that a lack of information is a good thing.
We’ve actually been talking to pharmaceutical companies for a few months …Before launching publicly, we showed them what we were going to do. The initial, knee jerk reactions, in some cases, were negative, but the more they saw what we’re doing and why the data is important and ultimately good for them, they came to understand it. They’re not big, evil conglomerates. They’re trying to make a product and sell it. But they’re also trying to make products that help patients. And if these data can help do that, it has a good purpose. And if it helps them get ahead of the FDA and lawyers, that can also be a good thing.
Pharmalot: And consumers will be able to access the data? There’s no pay wall?
Overstreet: The consumer can go there and do a simple search for a drug name and find the actual side effects that are out there and how often…At a consumer level, the only premium service we have that requires payment is if you want an alert in the future to any new side effect problems. You can sign up. On the professional level – for pharmaceutical companies, insurers and hospitals – we have a bunch of sophisticated analytic tools they pay us for to query in different ways beyond a simple drug search. And that’s how we make our money.
Pharmalot: How do you think this might change prescribing?
Overstreet: I don’t want to sound overblown, but I think, ultimately, where we’re going is that this has the potential to change prescribing trends. If you can show clear data that one of two comparative drugs is showing more of a rate of side effects than the other, why would a physician continue to prescribe it? And why would a patient continue to take it? I think, ultimately, as people adopt the data and understand it, it has to drive prescription trends.
Pharmalot: Have you talked to the FDA about this?
Overstreet: We’ve not. We’ve submitted papers… At some point, we’ll have to sit down and discuss it with them… But the data is ours, as taxpayers. It’s freely out there and we’re making use of it and, hopefully, this will benefit public health…we’d love to sit down and work with the FDA, but they have their own issues. That’s something we don’t want to get sucked into… But if the FDA were to come along and say they can do it better than we can, I’d gladly be put out of business. As much as we’re trying to make money, we’re trying to improve public health… I have other things I can do and spend my time on….
Pharmalot: But will you charge the FDA, too?
Overstreet: (laughter) I would if I could… But I don’t know…We’re not having that conversation….
Elmore
Seems like essential information that nobody wanted to make easily available, for various reasons. These guys have done us all a huge public service.
Anonymous
Thank you … a huge public service as the VAERS site is next to impossible to access and read.
Interesting...
AWESOME!!!!!!!!!!!!!!!!
Jill McMahon
As someone who has done research using AERS,it is unwieldy, but the main problem is that it is voluntary.
Jennifer Wolfe
I recently finished a documentary about parents who didn’t want to give their kids medications because of their concerns about side effects. They had concerns about long term effects, but even more so about short term. There is a lot of conflicting information - doctors don’t always have the exact same goals as parents, or the same definition of “manageable” side effects. Also, effects can vary widely from one person to another, making things even more confusing. Something like this will be invaluable!
Yvonne
This is important information to have for clinical trials on investigational and approved drugs. Both investigators and IRBs can use this info for consenting volunteers for trials especially if it’s kept up-to-date. Great to have this site.
Disappointed
Looking up Lipitor on the website, I only find a summary of AERS case reports. For example, one can see that there are 4,048 primary case reports for Lipitor and Myalgia (16% of all Myalgia reports).
The website owner claims that this is the first website to report side-effect incidences - but there are no incidences reported on it! A summary of case report numbers is not an incidence (i.e. a measure of the risk of developing some new condition within a specified period of time).
A table of case numbers (even when written as a % of total cases) doesn’t help a consumer who wants to know what their risk of developing a side-effect is…
Why, Pharmalot, are you promoting something that doesnt have an epidemiological basis? You are supposed to be a champion for fact checking and rigor.
Stacey
Disappointed - check again you need to click on the links its all there at least when I went, funny enough I was looking up Lipitor for my dad.
Brian Overstret
Ed - first of all, thanks for the interview. I’m excited to see all of the comments on this board, we appreciate your support and feedback.
To Disappointed - I appreciate your comment as well. You are correct in how you are reading the data on our site. We do not publish incidence rates on the site. The reason for this is that the number of scripts or patient population numbers are tightly controlled by a couple of large companies. We are not allowed to directly display their information on our site or even show the calculated rates on our site. We try to make that clear throughout our site, and have hopefully succeeded based on your finding and comment.
Where we do discuss and show specific side effect incidence rates is in our deep-dive research reports, called Monitor Reports. Coincidentally, we are publishing a new report on all of the major statin drugs next week where we compare the side effect incidence rates for all of the major statins across three major serious side effect types. In that case, we absolutely look at the patient populations of each drugs to do a side-by-side comparison.
We hope you look for that information when it is released - likely next Tuesday (10/18) and I invite anyone to contact me directly at brian@adverseevents.com with any other questions or concerns.
Best regards,
Brian Overstreet
President, AdverseEvents
Geeta
It is with cautious optimism that I read this potentially hopeful news. It has always been my view that the FDA website, in toto as well as adverse event reporting, is deliberately confusing and confounding.
Perhaps this new system will shed a bit of light on the truth. Even though the Adverse Events Reporting system is voluntary, at the very least one should be able to make some sense of it and draw some conclusions about the safety (or lack of it) about a drug.
One should not have to wait for leaked information to the New York Times, another country requiring an American pharma to put a warning on a drug in that country (and news of this - after the fact), and so forth, to learn to truth about potentially lethal side effects or - pray tell - to have the FDA finally act and require some sort of warning after many deaths.
Tanith
Got to agree with Disappointed. The website does not provide incidence just a summary of case report numbers
Gerd Kassel
Sorry, but I also have to agree with Disappointed. There is no (and cannot be)any useful information on incidence of AEs.Pure numbers of reports without background information on exposure, reporting bias (e.g. due to temporary publicity regarding a specific AE or product)do not tell you a lot.As a doctor I would be cautious to base any prescribing decision on this kind of information.
AnnePME
Ed and Brian, thank you for this. Has there been any response from the FDA?
dzieczko
@Kassel - a guest speaker from the FDA gave a lecture to those of us in clinical research at Sandoz in early 1990s about how it is possible to understand side effects (AEs) correctly from the Phase I - First in Man data. The lecturer was in his 50s (years old) at the time, so I’m sure he’s retired by now. But it just goes to show you how kicking everybody out with experience only creates lost time until people re-invent the wheel - or in this case, they don’t - they invented a square circle :-)