FDA Was Wrong About Google Functionality

I didn't say very much about the use of Google text ads in the Regulatory Alert that I released last week about the FDA space-limited risk & benefit presentation guidance. Partly, that's because I'm working on an article about Google text ads that updates the chapter in my primer.

There is, however, one extremely important item to point out. FDA got sitelinks wrong. I don't mean that I disagree about the FDA's view on the use of sitelinks. I mean they don't understand the technology, and what they say in the guidance demonstrates that.

The FDA guidance recommends using sitelinks to include some risk information that the FDA says people must include if they are presenting product benefit information (and the product name). The example they give is:

In this example, the benefit information "For severe headache from traumatic brain injury" requires corresponding risk information to be balanced.

The FDA accomplishes that risk presentation via the various sitelink descriptions "Potential for brain swelling," "Life-threatening drop in heart rate," etc. Importantly, the words in the sitelinks themselves "Boxed warning," "Warning," and "Risk information" do not themselves constitute risk information, just links to risk information. Consequently, those links in themselves do not suffice to provide the risk information.

The first problem with this is that FDA is assuming that a product manufacturer can guarantee that their ads have the sitelinks and the link descriptions appear when they run the ad. But you don't have that control.

Google does not guarantee that sitelinks will ever appear. In fact, Google's page dedicated to sitelinks in paid search campaigns, includes a section header called, "Why your sitelinks aren't showing," and the first sentence under that header reads, "Keep in mind that your ads won't always show sitelinks."

That's a problem because if your ad is running with benefit information presented in the body of the ad, the ad requires the risk information to meet the fair balance requirement. If your ad displaying the benefit information does not include the sitelinks, then the ad would violate the fair balance requirement.

So, if the maker of HeadHurtz were to actually run the ad that the FDA provides as an example, there's a significant chance that the ad would actually appear as follows:

This ad would definitely be violative because as we learned in 2009, and as FDA reiterates in this guidance, if the product name and benefit information is presented, then FDA requires that some risk information be presented at the same time with comparable prominence as the benefit information.

So following the FDA's provided example could result in violative promotional activity. Moreover, even if the sitelinks appear, there's a chance that the sitelink description would not appear. But again, the sitelinks provided in the example ("Warning," etc.) do not in themselves provide risk information. So, just providing the sitelinks would not meet the fair balance requirement.

Further compounding the problems with the FDA's example is that the order and number of sitelinks provided varies without complete control by the advertiser. So, if all four sitelinks were not provided, there's a chance (based on FDA's guidance) that the advertiser might not be complying the fair balance requirement because FDA might deem it necessary to provide ALL four sitelinks to meet that requirement.

As if these problems were not enough, there's one more difficulty: mobile devices. If this ad were enabled for display on mobile phones (which is not required for Google ads but is encouraged), then you would definitely not have all four sitelinks presented, and none of the descriptions of the sitelinks would be presented.

So, the sample ad presented by the FDA could definitely not run on mobile phones, but of course, mobile access to the Internet is exploding, and in fact, some people (hint: me) believe that providing information via mobile devices is even more important than doing so via desktop.

A Different Twitter Proposal

I've been thinking further about the FDA social media guidances delivered this week and especially the "Twitter" guidance, which isn't just about Twitter, but which does have significant implications for brand communications on Twitter.

What if the FDA had proposed a different framework?

Here's the FDA's sample compliant Tweet:
NoFocus (rememberine HCl) for mild to moderate memory loss-May cause seizures in patients with a seizure disorder www.nofocus.com/risk 
This Tweet takes up 134 of the 140 characters available on Twitter. It includes
  • Brand name (NoFocus--7 characters)
  • Established name (rememberine HCl--17 characters including space and parentheses)
  • Link to risk that includes both brand name & indication that risks are presented (www.nofocus.com/risk--20 characters)
  • A hyphen to separate benefit from risk info (1 character)
These are the mandatory elements of the Tweet, and they take up 43 characters with the balance available for communicating both the most serious risks and a non-misleading indication statement.*

Of course, the exact length of the brand and established names will change the number of characters that are set aside for accommodating the FDA's proposed mandatory elements. 

One challenge is that almost no brands will be able to provide a non-misleading indication statement and all of the most serious risks** in the space that's available after meeting FDA's mandatory elements. 

Another challenge is that although the FDA might not object to the sample Tweet they provided, users of Twitters certainly would object. Their objection wouldn't be to the regulatory compliance of the Tweet. Their objection would be to its value, as in, it has none. While such a Tweet might occasionally be useful for some products, and perhaps be a promoted Tweet on occasion, not many people would be inclined to follow an account peppered with this type of content.

The sample Tweet provided by the FDA is analogous to someone going on Twitter and saying, "I'm great. Read more about me here." And although some people do in fact use Twitter to do that, people who primarily send out such communications are rarely successful. Instead, communications need to meet the needs of the other users of Twitter, who might be looking for information about a product, news, etc. 

Responding to such information seeking behavior and providing customer service are two of the best uses of Twitter by companies (not just pharmaceutical companies), and it is not unreasonable to believe that most of these needs can be fulfilled without providing any benefit information about the product. So, if, for example, I want to know how to sign up for a co-pay program for your brand, just getting a reminder-style Tweet with a link to the co-pay program link can be very useful. Some companies are already engaging in this type of activity.

That's one reason that it was so disappointing that the guidance didn't address reminder ad formats.

However, on occasion, there might be a need to include benefit information, and when that happens, the format set forth in this guidance is greatly limiting and not likely to be accessible to most brands. So, instead, I was considering a format along the following lines:

The brand name would be presented in the Twitter handle, not the body of the Tweet itself. The Twitter handle has a 15-character limitation, and limited risk information could be presented in the Tweet in a manner that makes clear that additional risk info is available at the destination URL.

Modifying the FDA's Tweet to meet this proposal would yield this sample Tweet:
For mild to moderate memory loss-Risks incl seizures in some ppl bit.ly/isi
Assuming this Tweet was sent from @NoFocus, let's see how that Tweet compares to FDA's format.

Because both the brand name and the generic are omitted from the Tweet itself, that eliminates 25 characters from the Tweet that were taken up by those two elements plus the space between them. I expanded from the hyphen to include the phrase "Risks incl" so that's a growth from one character to 11, but I did that both to provide explanation for users about what the next few characters were going to present as well as to set up the expectation that the link that is provided will have further risk information. It is extremely common on Twitter to abbreviate the word "include" with "incl" so that is likely to be understood by most users of the platform.

Then, I made use of a URL shortener (bit.ly) and a common abbreviation for the page name (isi) where the risk information is provided. This meets the FDA's criterion for the URL that it not be promotional in content or tone, and because I preceded it with the phrase "Risks incl" users expect to find product risks presented at that page, which the FDA accomplished in its example by including the product name and word "risk" in the URL itself.

The phrase "Risks incl" also accomplishes the goal of clearly communicating within the Tweet itself that the brand is NOT providing all of the risks associated with the product within the Tweet itself. So, users are alerted that there are other risks to the product and they can learn more by clicking on the link. 

The revised Tweet is only 75 characters with 21 characters devoted to providing mandatory risk information, not including the risk statement itself which would vary by brand:
  • The phrase "-Risks incl" (11 characters)
  • A shorter url linking to the risk information (bit.ly/isi--10 characters)
And in that case, I would be able to provide additional links or other information such as directing this Tweet to another user. Imagine the following exchange:***
@phillycooke: My father was prescribed Brandufate by his doctor. What does Brandufate do?
@Brandufate: @phillycooke It's for mild to moderate memory loss-Risks incl seizures in some ppl bit.ly/isi
In the conversation, I have now presumably provided some value to the person who was looking for information about my product, while meeting almost all of the FDA's requirement.

I write "almost all" because I didn't include the generic/established name in this example. My belief is that in this type of space-limited context, it would be better for the FDA to permit that the established name be provided via the landing page with the risk information, not in the Tweet itself. I believe that meets the needs of users better and also provides extremely valuable characters for communication in the Tweets themselves.

The FDA's guidance explicitly points out that it is not addressing the landing pages, home pages for brands on Twitter, etc.. It is worth noting, though, that a brand's participation on Twitter is not limited to the Tweets it sends, but also includes the brand's home page, description, and a Twitter icon. These other locations provide what I believe are more appropriate locations for presenting the full brand and generic names in compliance with all of the FDA requirements for presenting the established name.

And because Twitter has a verification service available, it would be possible for users to understand via the blue checkmark icon that the account providing them information is an official representation of the brand, who is subjected to FDA regulatory oversight and therefore meets a higher standard for accuracy.

I'd love to get some additional feedback on this proposal, so post your comments, send me your emails, and let me know.

* Keep in mind that the FDA is saying that both of these elements are required. It is not OK to omit either the risk or the indication statement.

** And the guidance is unambiguous about the need to include ALL of the most serious risks. Quoting from the guidance
At a minimum, a firm should communicate the most serious risks associated with the product together with the benefit information within the individual character-space-limited communication. For a prescription human drug, the most serious risks would generally include all risk concepts from a boxed warning, all risks that are known to be fatal or life-threatening, and all contraindications from the approved product labeling (the PI). (p. 9)

*** For my modified Tweet, I'm changing the brand name because @NoFocus actually exists on Twitter as a user. I created an account for @Brandufate for demonstration purposes. 

FDA Updates Mobile Apps Guidance DHLB Regulatory Alert Available

Updated to correct important error: Correcting "not" to "now" in opening paragraph.

On Friday, CDRH came out with a new guidance that included revisions to the previously finalized guidance from September 2013 about mobile medical apps. The most significant aspect of this guidance is that systems that simply transfer medical imagery and data from one platform to another will now face a significantly lower regulatory burden.

You can read the Digitas Health LifeBrands Regulatory Alert about this guidance here.

In many ways, this activity was foreshadowed by the update made earlier this week to the list of mobile apps that would receive enforcement discretion.

On a tangential note, the recent activity from CDRH on this topic is welcome because, in my view, it correctly demonstrates a willingness to update FDA's position as new information is released, but it does present some challenges for industry and some problems with the existing Good Guidance Practices, and FDA is clearly aware of some of these difficulties.

According to GGP, Guidances must be released in draft status with an initial period of comment of at least 90 days. Then, FDA has the option of finalizing the guidance unaltered, updating the guidance and releasing a revised draft, updating the guidance and releasing a final version, or doing nothing (in which case the guidance stays in draft status).

What we have in this case is FDA proposing a new draft guidance with edits to a previously finalized draft guidance, and at least as of today, there's nothing in that finalized guidance to indicate to readers that it no longer represents FDA's current position. I understand the desire to not retract a final guidance or change it back into a draft guidance for every update, but in this case, someone could easily read the final guidance and not realize that it is outdated.

It seems the far better practice is what CDRH is actually adopting, namely, maintaining a webpage that is more easily updated that a formal guidance document, but the status of the webpage versus the guidance documents themselves seems to create some tension.

Perhaps moving forward, FDA should limit the guidance documents themselves (because of the time-consuming process needed to update them) to the principles FDA is applying with more easily updated lists referred to in the guidances themselves.

Of course, if FDA's view about the regulatory status of a product, device, etc., is subject to revision too frequently, then clearly there's a deeper issue that needs to be addressed.

Busy Week in Rockville (and Philadelphia)

CDRH has released a new draft guidance about mobile medical apps. Here it is.

I'm reading now and will have more later.

DH LifeBrands Regulatory Alert Available Now

The Digitas Health LifeBrands Regulatory Alert about the two social media guidance documents from the FDA is now available.

You can read it here on my Scribd.com page. As a reminder, everything I publish is accessible via that page, with the convenient link on the right of the blog.

Would love to hear any feedback about this in the comments, including the correction of any misinformation on this third-party website, should the FDA choose to apply its guidance.

Let's have a chat!

Some of you might recall that in the fall, CDRH released a finalized version of its mobile apps guidance (see here and here on that). A few days later, they hosted a live Twitter chat on the topic. They used the hashtag #FDAapps and have actually kept that going for ongoing updates.

I participated, as did I'm sure many of you, and it extremely informative. I commend CDRH for setting the precedent and embracing new platforms for outreach.

In that same spirit, I suggest that now is the perfect time for OPDP, APLB, and CVM's advertising arm to join Twitter (none has an official Twitter presence so far as I'm aware and based on the FDA's official listing of social media accounts) and talk about the new social media guidance on presenting benefit and risk information in space-limited contexts. This seems especially appropriate given that the guidance dwells so heavily on Twitter usage.

I floated the suggestion on Twitter yesterday, and @FDAmedia picked up on it. So, now, I'm hoping you'll help me keep the conversation going by Tweeting, posting, sharing, etc., to see whether we can get FDA to participate.

And even if we can't get FDA's involvement, let's plan on holding a Twitter chat of our own on this topic soon.

It's Social Media Guidance Day!

Update: FDA's Tom Abrams has a blog post explaining the new guidances here

Well, the FDA has kept its promise to deliver additional guidance prior to the July 1 deadline set in FDASIA.

Two new guidances addressing the topics of presenting risk information in space-limited formats and correcting misinformation on third-party websites are now available. Read them in full here and here.

Combined with the previously released guidance on fulfilling the 2253 filing requirement for social media (Guidance here, my take here), that brings us to three guidances on the topic, and one more promised on the guidance agenda for the year covering the topic of links.

I'll have a more detailed take available later today, but right now, I have some reading to get back to.

FDA will not regulate Apple's HealthKit

In case you missed it, FDA has been updating its list of mobile apps where it intends to exercise enforcement discretion (on that concept in the FDA guidance on mobile apps see here and here). MobiHealthNews has a more extensive story here.

The most significant fact is that FDA has apparently added a new category that seems designed explicitly to include Apple's new HealthKit that was announced on the WWDC a few weeks ago.

h/t @wapgeekboy

FDLI Advertising & Promotion Conference

I have accepted an invitation to speak at this year's Food & Drug Law Institute Advertising and Promotion conference about social media issues.

I'd like to solicit your input on topics for an interesting session. Of course, we'll have to cover OPDP's first social media guidance (see here and here), but what else do you think is particularly interesting or relevant to the participation of prescription-only products in this space.

Are there specific issues that you have not seen previously addressed? Have you seen anyone do something especially interesting in their use of social media that you think the industry as a whole needs to learn more about?

Send me your suggestions, or post a comment below.

June PHL RAPS Networking Event

The Philadelphia chapter of Regulatory Affairs Professionals Society (RAPS) is holding a networking meeting on Thursday, June 19 at the Primavera Pizza Kitchen in Ardmore, PA.

Here's the link to the RAPS Philadelphia group announcement.

I hope you'll join us.