By McKenzie E. Cato* & Allyson B. Mullen –
Approximately one year ago, on New Years Eve, FDA surprised the device industry with a draft guidance on emerging postmarket device signals (see our post on the draft guidance here). On December 14, FDA issued the final guidance document: Public Notification of Emerging Postmarket Device Signals (“Emerging Signals”). FDA has made significant changes to the draft guidance, including a modified definition of “emerging signal,” a description of the process CDRH will follow in identifying and assessing emerging signals, and the prospect for interactions between industry and FDA as the Agency assesses emerging signals and decides whether to post a public notification announcing the emerging signal.
The guidance defines an emerging signal as “new information about a marketed medical device:
- that supports a new causal association or a new aspect of a known association between a device and an adverse event or set of adverse events, and
- for which the Agency has conducted an initial evaluation and determined that the information has the potential to impact patient management decisions and/or the known benefit-risk profile of the device.”
Notably, the final guidance adds to the definition that “[i]nformation that is unconfirmed, unreliable, or lacks sufficient strength of evidence is not an emerging signal.” As we noted in our post on the draft guidance, there is a significant risk with providing unconfirmed data to the public too soon, including the risk that such notification could lead to the public not using a beneficial device. This addition to the definition, therefore, represents an important, and welcome, change from the draft guidance’s definition, which stated in part that an emerging signal is information “that has not yet been fully validated or confirmed.”
The most substantial addition to the final guidance is a section which outlines the process CDRH intends to follow in evaluating “signals that support a new causal association or a new aspect of a known association between a medical device and an adverse event or set of adverse events.” The draft guidance did not include a description of this process; this section is entirely new. This process appears to be independent of and will occur prior to FDA’s assessment of whether to issue a public notification announcing an emerging signal. The guidance states that this process will typically include the following elements: (i) collection of information; (ii) interaction with impacted medical device companies; (iii) review by a multidisciplinary team of subject matter experts; (iv) stakeholder engagement; and (v) management oversight. Note: these elements are not necessarily listed in order of occurrence, as explained by the process overview described below.
In this process section, the guidance explains that FDA initially identifies potential emerging signals using information from a variety of sources, such as Medical Device Reports (MDRs), MedSun Network reports, postmarket studies, data from clinical trials or scientific literature, epidemiological research, health care claims data or registries, and inquiries or investigations from other health agencies. Once an emerging signal has been identified, a CDRH “signal management team” of multidisciplinary experts is convened.
Next, the emerging signal undergoes a “signal refinement stage,” in which CDRH “attempts to better understand” the identified adverse event or risk, including the likelihood of a causal relationship between the device and the adverse event or risk. CDRH will conduct a preliminary assessment to determine whether the issue is limited to a single device model or manufacturer or whether the issue is more widespread (e.g., across a device type). During the signal refinement stage, the Agency identifies, gathers, and evaluates additional information from other data sources. It may also consult with other stakeholders, such as clinical or scientific experts, patients, industry, or other government or regulatory agencies.
The guidance notes that the signal refinement stage will also include interactions with the affected device manufacturer(s) “unless time does not permit because of the risk of patient harm or it is not feasible, e.g., CDRH cannot reach all manufacturers.” This is an important and necessary addition. The draft guidance did not indicate that the Agency would consult with or obtain information from industry regarding emerging signals. Unfortunately, the guidance does not elaborate on FDA interactions with industry. Therefore, it is still unclear the extent to which FDA will communicate with industry and the type of information that FDA will seek or accept from affected manufacturers. Furthermore, the guidance does not elaborate on the caveat that FDA will attempt to interact with manufacturers “unless time does not permit because of the risk of patient harm.” It remains to be seen how frequently FDA will determine that “time does not permit” interaction with industry.
Finally, the signal management team will identify public health and/or regulatory actions to mitigate the risks or adverse events which led to the identification of the emerging signal. The guidance lists several “tools” the Agency may employ in attempting to mitigate the risks or adverse events, including public communication (described below), requests to manufacturers to modify product labeling (e.g., to include a new warning/precaution), development of Agency guidance documents (e.g., that relate to premarket testing or requirements), and/or ordering postmarket surveillance studies.
In regard to the public communication “tool,” the draft guidance included a list of factors that CDRH will consider in determining whether FDA should issue a public notification about an emerging signal. This list is largely unchanged in the final guidance, but FDA did include two new factors: “[t]he quality of the data or information” and the “[p]otential for patients to not receive treatments they should even in light of the new information.”
Additionally, the guidance includes the clarification that the purpose of a public notification is to give the public current information about a device which may help inform patient management decision making, but it does not mean that FDA has definitively concluded that a causal relationship exists. However, the guidance notes, in instances where FDA has not definitively concluded that a causal relationship exists, “CDRH generally does not intend to issue a public notification unless 1) credible scientific evidence supports a new causal relationship, but the Agency needs additional time to reach a more definitive conclusion . . . or 2) FDA has concluded a causal relationship exists, but the Agency needs additional time to develop recommendations . . . .”
The draft guidance permitted public notification when FDA had identified a “potentially” causal relationship between the adverse event or risk and the use of the device. The prospect that the Agency might prematurely release a public notification based on a “potentially” causal relationship, which could irreversibly affect a device’s reputation, caused us concern. Thus, we are pleased to see that the final guidance appears to reflect CDRH’s intent not to release a public notification until they are more certain of a causal relationship. We will have to wait and see, however, how much or how little information the Agency believes “supports a new causal association” in practice.
In addition, it is yet to be seen how much information FDA will share with the public regarding the information on which it relied in deciding to make its notification. For example, earlier this year, FDA issued a public notice recommending against use of screening tests for ovarian cancer. In this publication, FDA generally referred to data not supporting tests currently on the market and that relying on an unproven test could lead to a delay in treatment or unnecessary treatment. In this instance, however, there are no other options for potential early detection of ovarian cancer. Therefore, it may have been helpful to provide readers with more specifics as to the information on which FDA relied and specific instances of deaths or serious injury so that healthcare providers could more fully evaluate whether to continue to use available tests.
Regarding the process and timeline for issuing public notifications and updates to notifications, the guidance includes a new statement that, in addition to interacting with affected companies during the signal management process, FDA will “inform the impacted company or companies shortly before issuing a public notification, unless time does not permit because of the risk of patient harm or it is not feasible, e.g., CDRH is not able to reach all manufacturers.” This statement contains the same caveat as with FDA’s attempt to interact with manufacturers during the signal management process. Again, it remains to be seen how frequently CDRH will determine that “time does not permit” sending notification to companies. This addition, as with the prospect for interaction with industry during the emerging signal assessment process, is an important addition further acknowledging that FDA needs to work with industry while investigating and publishing emerging signal notices.
The guidance indicates that FDA will issue an initial notification within 30 days of receiving information generating an emerging signal. However, FDA has still not clarified how it will communicate the initial notice to the public (e.g., actively through letters/emails or passively on its website). The guidance states that “updates to the public notification should be posted to the FDA website at least twice per year,” but, as with the draft guidance, it is still possible that FDA could actively notify the public (e.g., through letters to doctors), but then only provide updated information passively by posting it on the website. Those who are initially notified of an emerging signal may not be aware of updates later posted on FDA’s website.
FDA has also added to the final guidance that it “may notify the public when FDA has completed its evaluation and determined that additional regulatory or public health actions are not required.” It is notable that this communication is possible (“may notify”), but not necessary. We would have expected stronger language if it turns out that FDA action is not required after an initial communication. Nonetheless, after issuance of an emerging signal notice, further notification from FDA, even if it says that there is no further action needed, may not be effective in repairing any damage already done to a device’s reputation.
In sum, we are happy to see that FDA has stated that unconfirmed or unreliable information will not meet the definition of an emerging signal, and that FDA has seemingly created stricter criteria for a causal relationship that would warrant a public notification. It should be reassuring to industry that FDA has added several references to communicating with manufacturers regarding emerging signals. In practice, we continue to hope that the Agency balances the need to provide up-to-date risk information to the public with the risk that the public will reduce or cease use of a beneficial device following public notification of an emerging signal.
* Law Clerk