While this article points out some potential problems in the methods used to derive that headline political figure, its own methodology seems to have pretty serious flaws. The most I'd take from this is "This is such a bad problem that even when you try pretty hard to look in the wrong place you find it anyway".

This study deliberately does not - just for example - count it as a "medical bankruptcy" if you weren't hospitalized. So magically every person who was bankrupted by the cost of their must-take prescription medicine, they don't count because they need to be admitted to hospital to count for this study.