In short, the nocebo effect is like the placebo effect but for side effects. If a patient consumes a lot of articles and anecdotes about negative side effects of a certain drug, they are far more likely to experience those side effects for themself, even if they receive placebo.
This is becoming a greater problem as more patients read up on side effects through the worst of the worst side effect stories shared on social media.
As an anecdotal data point, I'm a redhead with a periodic liver condition and a fast metabolism, and I don't metabolize medications normally. I take an antispasmodic for pain relief in doses that leave some people catatonic and it has little to no negative effects on me, over a decade of use. I've tried low doses of muscle relaxants that leave me reeling drunk for hours. Had two awful experiences on SSRIs, I will never try those again should I ever have the need of an antidepressant. I suspect there are sub-populations of people like me who are statistical outliers.
1. we tend to overlay our own interpretation over what is described. Because that's pretty much the only way if we are to hope to do something about it.
2. In cases where something really unusual (meaning, not described in the literature) happens, modern medicine is an exercise in flying blind. Our profession is completely bound by best practices guidelines. The days of individual experimentation are long gone, and no one will ever forgive you for taking the off-path if there are negative consequences in this day and age.
Apparently the only reason the link link between tick bits and the allergy was because the researcher studying the pheomenon got bitten by a tick and developed the allergy himself.
There's a great Radiolab episode on this.
There are a number of medical problems, side effects, interactions, etc that only a very, very small number of people experience. Doctors, rightly, seem skeptical of anecdotes that haven't been borne out in large studies - they're likely to go with the most common explanation for something and usually be right.
But how do people who really are experiencing rare problems get the right care for them? Doing their own research, talking to doctor after doctor until someone believes them? Why should the medical community do research into problems that only affect a few people? Even if they do decide it's worth it, is it possible if the sample sizes are going to be small? Say I could volunteer to prove through experimentation that if you give me medication X, I get symptom Y, and when you give me a placebo, I don't get symptom Y, but no one else has ever experienced symptom Y from medication X, does anyone care?
All hard questions that I wish there were answers to.