For four and half years now FDA’s OPDP has endeavored to devise answers to 5 questions raised during a public hearing on the Internet, Social Media and Pharma and Device Manufacturers held in April 2009. Finally, on a day when I am travelling, the agency poses at least some partial answers to those questions in the form of two guidances. Ed Silverman at Pharmalot broke the news (“Tweet This: FDA Finally Proposes Social Media Guidelines”).
FDA’s approach to the guidance materials is not to lay down commandments. Rather, the agency presents their thinking or principles in a section labeled “Background” and then to provide examples that lend insight into how a company might consider approaching a situation. There are, I think, some real surprises for companies. I will note them below.
Here is my first take on the contents of the two draft guidances, with the proviso that what is rendered herein is not legal or regulatory counsel, but rather observations about the impact of the guidances in communications considerations and any regulatory or legal question should be posed to counsel:
1. – “Correcting Independent Third Party Misinformation About Prescription Drugs and Medical Devices” provides some very useful information, both direct and indirect.
- Companies CAN Correct Misinformation. The first surprise is this. Companies can correct misinformation. “If a firm voluntarily corrects misinformation in a truthful and non-misleading manner and as described in this draft guidance, FDA does not intend to object if the corrective information voluntarily provided by the firm does not satisfy otherwise applicable regulatory requirements regarding labeling or advertising…” Up until this point, I believe that many firms believed that correcting misinformation on a third party site was not an option. Apparently it is. That is part of the principle laid out in the background.
- Guide for Making Corrective Comment. FDA made specific recommendations for going about a correction. Corrective comment has to be relevant and responsive to the misinformation and tailored to it specifically. A corrective message should apply to the indication being discussed in the original commentary, not additional indications for the product that were not mentioned. It must not be promotional in nature, tone or presentation and must be accurate and consistent with the label, supported by evidence and should disclose that the person posting the correction is affiliated with the product manufacturer. These are the standards by which a correction will be assessed by FDA.
- Companies Can Host Discussions. Another point the agency makes is that when the firm itself is responsible for the product communication that contains the information, then the guidance does not apply. So, no surprise there – I think everyone understood that they are responsible for content that they generate. However, here is another reveal – in providing examples as to how a firm is responsible for its content and where the draft guidance would not apply – one example involves a firm that is hosting a discussion group on its own website – and is monitoring that discussion and in the course of it, removes some of the unflattering comments or postings and adds positive postings. Obviously that content is under control of the firm and so is not something covered by the guidance. But the interesting part here is that the example seems to take for granted that most firms would do this in the first place, but I think that is most definitely not likely the case. And here is a statement I bet most did not know – “Firms are generally not responsible for third-party UGC (User Generated Content) about their products…regardless of whether the firm owns operates the platform on which the communication appears.” In the example of a firm hosting a discussion about an FDA approved product on their corporate website where they have provided a clear message that they are not controlling content and did not create the content, then the “firm is not responsible for the information that is posted by independent third parties and can, if it so chooses correct misinformation according to the guidance.” Now comes a very big BUT. What does it mean that a firm is “generally not responsible”? In other words, the firm is not responsible for misinformation, one supposes – but what if a participant in such a circumstance starts talking about an adverse event. Here the guidance is silent – and presumably that is the subject of another guidance perhaps yet to come. But I believe that most in industry would not hold such a discussion on their corporate site in the first place, so the example is murky here.
- Not obligated to correct. FDA also uses an example of a blogger supplying misinformation and states that a “firm may decide to attempt to correct misinformation, but is not obligated to attempt to correct it.” That raises an interesting point to this for manufacturers. If there is misinformation out there, they may correct it, but in doing so, must follow guidelines for doing so and they could mis-step. That means that there is some inherent risk in correcting misinformation, but there is no obligation to correct misinformation. In that case, many conservative regulatory cultures are possibly going to take the position that misinformation should not be corrected as a matter of taking the safest policy course. That means, in essence, that in seeking to address the public health concern about misinformation, the guidance may fall short of the mark.
- Correcting a Part of a Website Does not Obligate One to Correct All of the Website. A company should correct all of the mistakes in a posting, but in doing so, is not obligated to correct misinformation that appears on other pages of the site. To many, that will be welcome news.
There is more to the guidance, such as the fact that companies can contact the original author to suggest corrective information (once again, which is corrective and not promotional in nature). Companies are not responsible if, once contacted, the author of a piece does not issue a correction.
2. The second draft guidance “Internet Social Media Platforms with Character Space Limitations – Presenting Risk and Benefit Information for Prescription Drugs and Medical Devices” is a title that I am pretty sure is over 140 characters.
This guidance addresses the situation where a communications platform is of limited space and therefore poses a challenge for communicators regarding medical products where benefit information about the product has to be accompanied by information on risk. Previously through regulatory action letters, OPDP has taken the position that inclusion of the risk information by link is not sufficient. With this guidance, OPDP appears not to have changed that position. This guidance addresses primarily microblogging platforms (Twitter) and sponsored link promotions. For my purposes, I am going to focus on the news medium – Twitter.
And here is the rub on this guidance regarding Twitter. Prior to the guidance, Twitter – a major communications hub in healthcare – particularly at medical meetings – was essentially off limits for branded information about pharmaceutical and medical device products specifically. There are many pharma Twitter feeds addressing everything from recruitment to corporate news, but product-specific efforts are extremely rare, though there is at least one product-specific Twitter feed run by a pharma company. However, a company could not tweet about the approval of its product by naming the product and the indication and a link to full information without running the risk of regulatory action from OPDP.
In other words, if FDA would tweet a drug approval – “XYZ Drug approved by FDA for diabetes Type 2″ – the manufacturer would presumably not be allowed to re-tweet the FDA’s announcement. After this guidance, nothing has changed.
In the upfront section of the guidance, FDA outlines the traditional regulatory parameters that any promotional labeling that makes claims about a prescription drug product must include the risk information conspicuously, present fair balance, contain risk information in each part, as necessary to qualify any representation about the product, etc.
General factors that go into the equation – first of course, the benefit information must not be misleading, which means that material facts about the use of the product such as limitations to an indication or patient population must be included. Further, the information must be accompanied by risk information within the character-space-limited communication. A link should also be provided to allow direct access to more complete risk information about the product. Finally, the brand name and the generic name need to be included.
The agency then sets up some hypothetical example to construct a tweet using a fictitious drug called “NoFocus” which is indicated for mild to moderate memory loss . Further profile of the product – it has no boxed warnings and no known fatal or life-threatening risks in the PI. The most serious precaution is that it may cause seizures in patients with seizure disorder. The chemical name of NoFocus is remembrine hydrochloride, but scientific abbreviations may be used for a chemical ingredient name so it could be written as HCI instead of hydrocholoride). Now, how does that translate to a tweet in compliance? Here was FDA’s result:
“NoFocus (remembrine HCI) for mild to moderate memory loss-May cause seizures in patients with a seizure disorder www.nofocus.comrisk” – 134 characters, with 140 the limit. While presumably you could cut this down by making patients “pts” and making with “w/” there is essentially little to no room left for any news, such as including the words “FDA approves” before the tweet which would take it to 146 characters and allows for no link to the press release about the approval.
Yet twitter is a news forum. The way that the example is constructed in the example is nothing more than an advertisement – not a communication particularly appealing in the Twittersphere and in fact, might backfire. The inclusion of so much information in the text of the tweet and the link being consumed by the link to fuller risk information leaves the tweet nearly unmanageable from a news point of view and would require a lot of review internally by a company to ensure compliance. So in the end, while product specific information was nearly impossible before the guidance, nothing has changed with the guidance.
In the end, that may be the lesson of both of these guidances. After such a long wait, little may have changed. While technically firms are allowed under the first guidance to correct misinformation, there is regulatory risk if the parameters for doing so are not met and therefore the most conservative course of action is to not correct. Remember, there are two pathways to getting regulatory guidance – the first is guidances and the second is warning letters when the nuance of the guidances is misread. And for the matter of the second guidance, while twitter was not a medium that appeared friendly for product-specific discussions by manufacturers, that would appear to continue to be true. Still, only time will tell. At least the guidances do offer some new insights but there are still outstanding questions and there remains work to be done.
Since beginning this posting, the WSJ has published another piece on the guidances that can be found here.