If we find that up to 10% of people report insomnia after taking Panadol, does that mean it was a side effect of the drug?
The correct answers, if there are such things, are 8% (nausea), 19% (upset stomach), 12% (abdominal distress) and 28% (cough). I don’t know how many people have these complaints simultaneously.
I’ve taken these figures from a study of 1000 New Zealanders, who took part in a telephone survey by University of Auckland health-psychology professor Keith Petrie and colleagues five or so years ago. The “symptoms” were part of a list of 46 complaints, and topping the list were back or neck pain (two in five said yes), fatigue, headache, congested or runny nose and joint pain or stiffness, all of which about a third of participants had experienced in the previous week.
In my opening paragraph, I asked specifically about the complaints because, although they’re not the most common seven-day ones, they are listed among the most frequent side effects of Panadol. Nausea and vomiting are experienced by up to 34% of people and 15% of those taking Panadol respectively, so are “very common” side effects. The others are considered “common”, because between 1 and 10% of Panadol users have them.
So, if 28% of us say we’ve “coughed” in the past week, and as many as 10% of Panadol users have also coughed, does that mean coughing is a side effect? Superficially, you could argue that taking Panadol means you’re less likely to cough than if you don’t take it. At the same time, it seems as if nausea and abdominal distress are more common among Panadol users than in our seven-day baseline.
I’m intrigued by this, because it suggests that we have to be careful how we think and talk about drug side effects – the generally unpleasant and unintended effects of the treatment in question. If we find that up to 10% of people report insomnia after taking Panadol, does that mean the drug caused it? What do we make of that in relation to Petrie’s finding that 286 of his 1000 New Zealanders also said they’d had insomnia in the previous week?
Maybe reading the side of the container after you’ve taken a Panadol tablet cues you to interpret your otherwise “normal” trouble sleeping as being attributable to the drug? In other words, a nocebo effect (the opposite of a placebo).
Petrie is also obviously fascinated by this, because he’s done more work on the impact of telling users about possible side effects. For example, you probably don’t remember that in about 2008, minor changes were made to the formulation of the only publicly available thyroid hormone replacement treatment. The active component remained the same and nothing weird was added, but there was a strong spike in reports of adverse drug reactions.
In 2012, the Auckland crowd published a paper in the medical journal BMJ, which showed that this spike followed a small number of TV news reports about the change to the formula. They noted that the specific adverse side effects showing the biggest reporting spikes were those mentioned in the news items. Put another way, not all side effects are equal. Although there are other side effects associated with the drug, those not mentioned in the news items failed to show a corresponding jump.
I don’t work in the pharmaceutical industry, but I imagine the possible creation of complaints as a result of reading drug companies’ precautionary information can leave the companies in a bit of a bind. They are required to tell potential users (and prescribers) about potential side effects, and there is a clear obligation to make sure users are proving informed consent, but these studies show that doing so can “create” the experience of side effects.
This article was first published in the March 23, 2019 issue of the New Zealand Listener.