FDA safety alerts: Getting the response right!

The Food and Drug Administration (FDA) has issued more than 350 safety alerts for drugs and biological products, since January 2008.¹ Approximately 12% (40) of these alerts were regarding medications indicated for the treatment of neurologic or psychiatric disorders. FDA safety alerts are intended to improve patient safety by providing the medical community with timely new safety data and actionable information that may impact treatment choices by both healthcare providers and patients. Communication of the safety alert to health care professionals also generally provides a summary of the data that serves as the basis for the alert, although often unpublished and not accessible for in depth review by decision makers. Health care providers and institutions are expected to orchestrate a response to these alerts that is in the best interest of delivering safe care to patients. Responses may range from dissemination of new information, policy development, or changes in selection or use of a drug or medication class. As mental health care delivery systems vary widely in size, patient population, infrastructure and financing, and resources (personnel and technology), responses to safety alerts may also vary to same degree. We were interested in identifying commonalities for failures and successes, in the experiences of clinical pharmacy specialists in mental health settings responding to safety alerts.

The roundtable discussion, held at the CPNP Annual Meeting, had approximately 15 participants. Roundtable participants represented community hospitals, public mental health systems, large managed care organizations, information management companies and the pharmaceutical industry. Participants were asked to share their experiences in responding to the August 2011 safety communication, “Abnormal heart rhythms associated with high doses citalopram.”² We sought common experiences and best practices with a goal of developing this set of considerations when responding to FDA Safety Alerts that could inform diverse practices in the future.

CONSIDERATION 1: TIMING

A common frustration among participants was working with incomplete information and having early communication and work overturned by late revelation of facts that were not immediately apparent on reading a communication from the FDA. In the specific case of the citalopram QTc warning, the language of the safety alert differed from the product information label regarding recommended dosing in the elderly and lacked generalized language regarding maximum doses when used concomitantly with CYP 2C19 inhibitors (e.g., only cimetidine referenced in the alert). Both issues were adequately addressed in the product label changes but created confusion for practitioners that read only the FDA warning. For several of us, the differences between the warning and amended label were apparent only after incorrect information had been broadly disseminated. Participants agreed that delaying response to a warning may be considered in some cases to allow sufficient time to gather information as discussed below. On the other hand, FDA warnings are often reported in national news and patients and providers have an immediate need for information. It is important to balance the need for deliberation, with the needs of your constituencies.

CONSIDERATION 2: GATHER EXTERNAL INFORMATION

Participants responded to this and past FDA warnings by first gathering information. Routine approaches were to read the FDA warnings closely, study amended product labeling, and to seek out any related published scientific reports. Almost all reported using the CPNP listserv to discuss the FDA communication, discuss strategies and share experiences. Other approaches to information gathering included seeking minutes from FDA psychopharmacy advisory (or related) committee meetings, searching news sources for commentary from experts associated with the warning, seeking editorial commentary from experts in peer reviewed journals, and searching official statements from professional organizations. Pharmacists in larger integrated systems looked to other regions to model their response or waited on the experiences reported by other systems before acting. The delay of days, weeks or even months for some of these external information sources may promote a staggered response – delivering information to patients and providers early to support individual care decisions, but delaying complex activities that may have significant consequences to health systems and patients.

CONSIDERATION 3: GATHER INTERNAL INFORMATION

All participants worked to better understand the size and scope of the affected population within their systems. With the label change associated with citalopram, several participants examined the affected population by age, daily dose, comorbidities and concomitant medications and assigned levels of risk depending on these variables. At least one site chose intervention on the highest risk population as determined by this type of review.

CONSIDERATION 4: STAKEHOLDERS AND DECISION MAKERS

For most participants, Pharmacy and Therapeutics committee was the central body handling the response with delegation to medication safety committees and/or clinical pharmacists. In some settings though, individual pharmacists found themselves in the role of developing a response to FDA warnings. In either case, the consensus of participants was that input from experts outside of pharmacy or psychiatry, such as cardiology and primary care often was helpful.

CONSIDERATION 5: DISSEMINATE INFORMATION

Pharmacists working in all clinical settings responded to the safety alert by disseminating information to providers within the health care system using electronic-communication and hard copy letters, academic detailing, physician meeting presentations, and Pharmacy and Therapeutics (P&T) meeting minutes. Initial information sent out tended to be a general alerts about the FDA warning, while later communications were more specific to the local population served, and suggested actions or adjustments in delivery of care within the specific system.

CONSIDERATION 6: BUILD THE INTERVENTION

Most participants identified a struggle with the decision of “intervening” on patients taking citalopram in daily doses that exceeded the newly recommended maximum daily dose. Some systems chose not to intervene and allow the prescribing provider to make decisions on individual patients, other systems chose to systematically reduce doses to the recommended maximum of 40 mg/day, and some intervened on what they considered “high risk” patients or those with a contraindication to citalopram due to the risk of arrhythmias. One pharmacist reported outreaching patients, assessing depression status via PHQ-9, and determining if dose reduction was reasonable.

CONSIDERATION 7: ASSESS THE OUTCOME

Interestingly, none of the participants acknowledged using an outcome measure to determine if the strategy used in their facility was effective at improving safety risk. Potential outcomes measures discussed by round table participants included 1) evidence that all patients receiving high doses had a risk/benefit discussion with their provider, 2) reduction in the number of patients receiving above recommended daily dose or concomitant interacting medications, and 3) rate of relapse associated with systematic medication changes. Given the importance of FDA safety alerts to the provision of care to patients, measuring outcomes as a result of initiatives around that information, seems prudent.

In summary, this roundtable discussion focused on institutional strategies used in response to recent FDA safety alerts regarding citalopram as well as the role of the clinical pharmacists in those same settings. Responses to the safety alert differed across settings but participants encountered similar challenges in interpreting the data contained in the alert and developing an action plan around drug-drug interactions. None of the participants had measured or reported outcomes.