Symptoms being over-reported in surveys

Here’s an analysis I did on data from React19 survey #2. That survey asked about:

  • Myocarditis as a symptom
  • Formal diagnoses (free-form). I used regular expressions to look for mentions of pericarditis and myocarditis, both of which have the symptoms of myocarditis.

This creates 3 groups:

  1. Reported myo/pericarditis as a formal diagnosis
  2. No formal diagnosis, but reported myocarditis as a symptom. Yellow/orange line. These people are likely speculating on which symptoms that they have. It is likely that some of them do have myocarditis and many of them don’t.
  3. No myocarditis.

In the chart below, you can see that #2 (yellow/orange line) is above the other line for many symptoms. Where that is the case, it is likely that a particular symptom is being over-reported.

The over-reporting would likely create large distortions where the blue and red lines are close to 0% (right-hand side of the chart). While anaphylaxis is a fairly objective symptom, it is likely being over-reported. There may be surveyees who are using an alternative definition of anaphylaxis (e.g. allergic reactions such as food intolerances, mild allergic reactions, etc.).