Follow Up of SAEs and NSAEs
Performing follow up on spontaneous and clinical trial SAE and non-serious AE cases is a critical function and it is not always done optimally.
Does follow up have to be done?
Yes. This is clearly required by law.
In clinical trials FDA clearly states in 21CFR312.32 “Followup . (1) The sponsor must promptly investigate all safety information it receives.”
FDA actually goes into a further explanation in its 2012 Guidance on Safety Reporting Requirements for INDs and BA/BE Studies (https://www.fda.gov/downloads/Drugs/…/Guidances/UCM227351.pdf).
“Most IND safety reports are derived from observations from clinical trials. In the setting of a clinical trial, information is collected in a controlled environment so that the information needed to evaluate the suspected adverse reaction (e.g., information that would be contained in a narrative report or on FDA Form 3500A) is generally readily available. If any information necessary to evaluate the suspected adverse reaction is missing or unknown, the sponsor should actively seek such information from the source of the report.”
Companies are expected to make full efforts to obtain complete information. Similarly, in 21CFR314.80 FDA discusses in several places the requirements for follow up of expedited reports and for reports in PADERs.
For OTC products in the US, FDA noted its policy in a guidance: Guidance for Industry: Postmarketing Adverse Event Reporting for Nonprescription
Human Drug Products Marketed Without an Approved Application (https://www.fda.gov/downloads/Drugs/…/Guidances/ucm171672.pdf).
All follow-up information received within 1 year of the initial report must be submitted within 15 business days. Note that the law says states only 1 year of follow-up, but FDA has indicated that it wants no time limit. That is, report all follow-ups forever.
The EMA similarly requires follow up on cases from all sources. There are over forty references regarding required follow up in Module VI of the EU’s Guidelines on Good PV Practices – Management and reporting of adverse reactions to medicinal products (https://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2012/06/WC500129135.pdf).
In general, the only cases that do not require follow up are those that are absolutely positively complete and/or no further information can be obtained as there is no contact information. This situation may occur with cases received from health agencies where the contact information is redacted or from other manufacturers who send reports with the contact information similarly hidden.
Secondly, it is highly unlikely that a complete set of information will be received on the initial report, particularly if the SAE has just begun and is evolving. The initial reporter may not have all the information, particularly if it is the patient doing the reporting. The health care professionals (HCPs) involved in the case will have a more complete data set. Finally, the outcome is needed and that may not be known if the case is still active.
Thirdly, signaling must be done and it is very hard to make reasonable, scientific judgments if the data is incomplete.
So, it is clear follow up has to be done.
The next question is how much follow up should be done and by whom?
The answer to the question on how much follow is needed is the Goldilocks principle. This was originally described with porridge: not too hot, not too cold but just right. What that means here is that one can go to extraordinary extremes and obtain every morsel of data or get minimal data, or one can get the appropriate data needed to understand the case. Obviously, the latter course is the best. This is clearly a judgment call and thus follow up should be done by someone who is both familiar with the case and also has medical expertise such that the right questions are asked and full and useful answers are received. Some companies have “canned” or scripted follow up requests which are usually not very useful unless they are created for specific AEs or situations such as a myocardial infarction script or a GI bleed script etc. In general these don’t work well unless they are clearly tailored to a narrow issue.
It is hard to know how much effort should be invested in the follow up. The exact amount is not written in any of the regulations that I am aware of. The broad rule of thumb is that for serious cases, two well-documented attempts should be made. This is usually more than just a phone call to the secretary or administrator asking for information or a call back but rather an attempt by someone who is likely to get through (e.g. a company physician, nurse or pharmacist) to the busy reporting physician or health care provider. It is sometimes not productive doing a phone call which won’t get through but rather sending a letter (sometimes certified return receipt requested) or email. For clinical investigators phone calls are more likely to be returned but for spontaneous reports less so. Sending a letter tends to call attention to the importance the company puts on the case. On the other hand, if the physician has to run down to the post office to get the letter and sign the certified receipt, the response will be one of fury!
What this means, is a medical professional should, in general, do follow up on all serious cases (and perhaps some non-serious ones too) such that the appropriate questions are asked and if the response is incomplete or not useful additional follow up questions can be posed. This is most often the case if the follow up is done by phone to the physician or HCP involved with the case.
All follow up efforts must be documented. By this, one means that the request (including the questions asked or documents requested) are listed in the electronic or paper case file along with the date of the request and the name of the requester.
The sender should track the follow up request and if no response is received in a week or two then the sender should try again. This should be clearly stated in the SOP covering follow ups. Although it is generally felt that two follow ups are sufficient, the number really should correspond to the importance of the case. A Stevens-Johnson or Torsades case should get strong and repeated follow up. A penicillin skin rash perhaps less so. The author remembers a case of alleged Torsades right after a patient took one of our anti-histamines (a very benign and safe product). This was so important, both for medical and commercial reasons, that multiple follow up attempts were made. However, when they failed to produce the needed information (cardiograms and previous cardiac history) the chief medical officer of the company flew on the corporate jet to the physician’s office some 800 miles away to personally see the treating physician and obtain the cardiograms and needed information. By doing this we discovered that the patient had long-standing Torsades which had occurred both before and long after taking the drug in question thus showing the drug was highly unlikely to be the cause.
So the bottom line is for serious cases, probably two follow ups but more if the case is important in signaling, is the first time that this was seen, was more severe etc. For non-serious cases, some companies do no follow up especially if the case is totally benign and the problem quickly disappeared such as a headache lasting an hour. Again, medical judgment and common sense should dictate the follow up.
The mechanics of follow up.
As noted above, there are several ways to do follow up: phone, email, letter, personal contact/visit. If in writing, the requests should be as explicit as possible (“the ECG of April 28, 2013 during the patient’s chest pain”) though sometimes this is not possible (“the ECGs relating to the patient’s chest pain in April 2013”). Be careful not to overask (“all lab reports during the patient’s 7 week hospitalization”). If the case is complex, consider asking for summary reports such as the discharge summary or the treating physician’s letter to the patient’s private physician.
Some companies now split up case processing into multiple components such as initial data entry, triage, coding, narrative writing, follow up request etc. whereby a different person may handle each step and no one owns the case. Although this may make for efficient case processing, things often fall through the cracks and no one remembers to follow up to see if the request for follow up was answered. Cases must be tracked.
Follow up should be done while the case is “hot” which usually means within a week or two at most after receipt of the case. If one waits longer than that the trail will get cold, the records shipped to the archive room, the treating physician is now working elsewhere, no one remembers the case etc. The likelihood of getting useful information in this situation is much lower. All new data must be triaged rapidly and handled per SOP procedures.
Documentation that is obtained should be as complete as possible. Ideally the actual cardiogram and interpretation should be obtained rather than a verbal report that the cardiogram showed an anterior wall myocardial infarction. The appropriate redaction of patient and reporter identifiers should be done to ensure compliance with HIPAA and other data privacy requirements.
- There are instances where follow up is difficult or not even feasible. The most common example of this is litigation. If a patient is suing the company, the company’s attorneys may require that all requests for follow up go through the law department and that there be no direct communication by drug safety with the physician or anyone else involved with the case. This obviously makes life more difficult and the person suing may refuse to send the information.
- As noted, cases from governments and other companies may not have any contact information. Companies often do this for business or competitive reasons especially in clinical trial cases. If the suspect drug is the comparator marketed by a competitor, the company doing the study may not want to send a MedWatch or CIOMS I form to the competitor as it would reveal some information about the study to the competitor. This is not good medicine but is the reality.
- One of the worst situations occurs when a case is reviewed at the time of the PSUR/PADER or DSUR or during signaling when someone asks for more information on a long ago case. This is often very hard to obtain; the drug safety department should have thought of this and tried to obtain the information when the case was fresh. Thus, do it right the first time.
Bottom Line: Follow up must be done and the follow up should be well thought out and assiduously done up front.