Are my symptoms due to my medication?

Sometimes when I receive questions by e-mail, I would like to share my answers with everyone, not just the initial writer. The following e-mail concerns a question I'm frequently asked: whether a medical finding that's new for the patient is related to a diabetes drug.

The e-mail asked: "I have been using [drug's name deleted] for 2 years. I have developed shaking of my hands and lost a lot of my hearing. Has anyone else complained of these symptoms?"

In this case, the author implies that his "symptoms" (which would be called "adverse events" by drug safety specialists) are new since he started taking the drug, which would increase the likelihood that they are due to the drug, rather than being a coincidence. Most adverse events arise almost immediately after starting a drug, in the first few days or weeks or months, but some are delayed and might not show up for several years after starting the drug. And sometimes preexisting conditions are worsened by starting a drug, so the timeline of starting drug then having the adverse event isn't always necessary.

First things first in order to decide if there's a relationship between the drug and the events: if the events have been reported frequently in the past, they will be listed in the approved product labeling. Approved labeling is variously called the "PI" or "USPI" or "SPC" or "SmPC" or "leaflet" or "label" in different parts of the world. For instance, for many diabetes drugs, hypoglycemia is a listed event. For some, liver damage is listed, for others, cardiac adverse events. But I am generally surprised to hear of shaking or hearing loss as side effects of diabetes drugs, and looking at the label for the patient's drug, I couldn't find any reference to these specific events.

Second thing to do: check with your physician about what they think; if he/she has a reasonable alternate explanation, then it's much less likely that it's a drug side effect.

But it still might be an adverse effect of the drug that simply hasn't been reported frequently enough to make it into the label. That's the problem with drug surveillance after a drug is approved and available on the market: reporting of new adverse events is dependent on a high level of suspicion on the part of practicing physicians that an event that had not been seen in clinical trials might indeed be due to the drug - either by itself or perhaps in some interaction with other drugs that had not been observed in clinical trials.

If it's not in the label, and if your physician doesn't have a reasonable alternative explanation, either he/she (or you, or your pharmacist or diabetes nurse) should submit a report of these "adverse events" to either the manufacturer or to the FDA's MedWatch program. (There are similar reporting processes in almost every country: in every country, physicians and pharmacists would know how to report events in that country.)

Of course, there are other ways to ascertain if an event is related to the drug: the easiest is to stop the drug, and see if the event goes away. (In drug safety lingo, that's called a "positive dechallenge".) That's pretty strong evidence of a relationship, but even better would be to subsequently resume the drug and see if the event comes back ("positive rechallenge"). This concept of "dechallenge" and "rechallenge" are great if the physician and the patient agree that it's safe enough to stop the drug in the first place, and if it's safe to resume it if there's a "positive dechallenge."