Following Up on SAEs: How much is enough?
A very common refrain amongst drug safety specialists (“pharmacovigilantes” we are sometimes called – affectionately I presume) is how many follow up tries does one have to make on a post-marketing SAE before saying enough is enough?
There is, as you know of course, no official or standard answer to that question. Alas.
The way to approach this is to consider the importance to public health, the patient and the company of getting complete information. If the SAE is new (unexpected, unlabeled) and is appears to be quite important and perhaps truly due to the drug, then a major effort will need to be made to validate the SAE. That is, if we have a case of agranulocytosis, aplastic anemia, torsades or Stevens Johnson or perhaps any of the FDA’s 2003 list of always expedited cases[*] (which never went into effect but is a good list nonetheless), a “full court press” is very appropriate. If the SAE is labeled and well known, it is less important to make herculean efforts to obtain more data. If the cases is non-serious, many regulatory agencies do not even require follow up.
So what is it that one should do for a serious, unexpected AE? It can start off with the usual standard operating procedure of the drug safety specialist phoning, faxing, pdfing, emailing the reporting physician. Usually most SOPs, when they do state a number of attempts, will require two attempts to get follow up information. If this fails, it may be worthwhile escalating the effort by having, say, the MD medical reviewer making an MD to MD phone call which may break through the reporting MD’s firewall of nurses, aides, voicemail etc. Sometimes a certified, return receipt or express mail letter is sent. If, however, the physician or a member of his staff has to go over to the post office to pick up the letter, the reaction may be one of fury rather than willingness to cooperate.
Sometimes very creative means are used. The author remembers a case some years ago of torsades from a major heart institute following the use of a rather benign and very safe (in retrospect) anti-histamine. The spontaneous report came in first as a fax with no cardiogram. The specialist made several calls to the institute reaching only a senior resident and not the treating physician(s). The resident was of no help and was not sure he had the right even to speak with the company let alone sending the hospital chart and ECGs. So the key issue was whether this drug did indeed cause this very nasty ventricular arrhythmia. The sales and marketing team immediately noted that this anti-histamine was in a very hot and competitive market and was a blockbuster. They also noted it was paying the salary of the drug safety department (which they also pointed out was a cost center and not a profit center. We argued against this point but never made any headway). So, an SAE like this would not quite kill the drug but would severely hurt sales.
So next I made calls (as the MD head of the drug safety group) and got nowhere either. Finally we had our medical information team hunt down the senior reporting physician’s medical background (where he went to school, did his residency etc.) and it turns out our Chief Medical Officer was a friend of a medical school classmate of the reporting MD. Our CMO called his friend who called his classmate at the heart institute and explained why we needed the follow up information. It turns out our entreaties never even reached him (a great firewall) and once he realized the issue he was more than happy to send further data.
When we received the complete dataset including multiple ECGs, it turned out that the patient had had several episodes of torsades (as well as multiple other arrhythmias) and that the one in question was not torsades but supraventricular tachycardia with aberrant conduction (our senior cardiologist and outside consultants confirmed that). Thus, what could have been a major problem for the patient and the drug turned out to be less severe.
Other stories and anecdotes include sending one of the physicians from the company to camp out on the “doorstep” of the reporting MD until the reporter would consent to see our physician. This doesn’t always work and sometimes engenders resentment on the part of the reporter.
The point here is simply that some cases require enormous effort to validate or refute the initial report.
What do the regulations and the laws say? Alas very little.
ICH E2D says: “Follow-up information should be obtained, via a telephone call and/or site visit and/or a written request. The company should provide specific questions it would like to have answered. Follow-up methods should be tailored towards optimizing the collection of missing information. Written confirmation of details given verbally should be obtained whenever possible. In exceptional circumstances, if requests for information have been refused by the reporter, a regulatory authority might be able to assist an MAH in obtaining follow-up data.”
Very reasonable but not too helpful. In my experience the regulatory authority will not do the follow up for the company.
The US FDA regulations state in 21CFR314.80: “Postmarketing 15-day “Alert reports”–followup. The applicant shall promptly investigate all adverse drug experiences that are the subject of these postmarketing 15-day Alert reports and shall submit followup reports within 15 calendar days of receipt of new information or as requested by FDA. If additional information is not obtainable, records should be maintained of the unsuccessful steps taken to seek additional information.”
Not too helpful but at least the FDA understands that follow up cannot always be obtained.
The UK MHRA states in its publication “Best Practices in Reporting of Individual Case Safety Reports (ICSRs) that: “Reports should be routinely followed up to ensure that information is as complete as possible with regard to early symptoms, treatment and outcome.”
Again, not too helpful.
The EMA in Module VI on the Management and Reporting of Adverse Reactions to Medicinal Products states: “Follow-up methods should be tailored towards optimizing the collection of missing information. This should be done in ways that encourage the primary source to submit new information relevant for the scientific evaluation of a particular safety concern. The use of targeted specific forms in the local language should avoid requesting the primary source to repeat information already provided int eh initial report and/or to complete extensive questionnaires, which could discourage future spontaneous reporting. Therefore, consideration should be given to pre-populating some data fields in those follow up reports to make their completion by the primary source easy.”
Nice point but still not too helpful.
So that is pretty much the state of the art. Do what you have to do to get complete information in order to make a good medical and scientific judgment regarding the case.
Bottom Line: What then is to be done?
In my experience over the years having seen many companies’ methodology and SOPs, there is more or less a consensus that two (rarely three) serious attempts are made by a medical professional to get information. Leaving a voicemail on the MD’s machine is generally not considered a serious attempt. If a phone call is done, the usual protocol is for the company caller to actually speak to someone in the physician’s (hospital’s, ambulance’s etc.) office, get the name of that person and leave a clear message to call back (collect) or write or email etc. This is then documented by the company caller in the patient’s record in the company.
Also considered to be a serious attempt is a letter (with acknowledgement of receipt). Faxes and emails are less well thought of as one cannot be sure the fax machine is on or that the email is not immediately dumped into the spam folder.
If this fails, the case is often escalated to the medical reviewer or a senior physician in the company who attempts a phone call. If that fails, a personal letter to the physician is done by the company MD again with some sort of acknowledgement of receipt. This can be repeated once or twice if appropriate.
Sometimes we have contacted the local company rep (detailer, salesperson, medical liaison) to personally call on the physician or go up to his or her office. This sometimes works.
Finally, if all this fails, the extraordinary attempts may be made if warranted. If not, one must concede that this is one of those times when the data will not be obtained.
That’s probably the best answer one can give. Will one sometimes get dinged during a health agency inspection for not doing sufficient follow up? Usually no but sometimes yes if there are too many cases that should have follow up but don’t.
The future actually promises that this mechanism of obtaining post-marketing information will end. If and when “big data” comes into effect and the nation or world is all digitalized, there will be algorithms that will be used to “pull” data from the databases automatically based on some sort of flag or trigger. The data will come automatically and completely (presuming the computer program or algorithm is set up well) and follow up information will arrive on the specialist’s “doorstep”. Of course, there are privacy and security issues here but that is another issue. For the specialist who hates making these types of calls, this will be paradise!
[*] Congenital anomalies, acute respiratory failure, ventricular fibrillation, torsades de pointe, malignant hypertension, seizures, agranulocytosis, aplastic anemia, toxic epidermal necrolysis, liver necrosis, acute liver failure, anaphylaxis, acute renal failure, sclerosing syndromes, pulmonary hypertension, pulmonary fibrosis, transmission of an infectious agent by a marketed drug/biologic, endotoxin shock.