I guess the study used the modifier “wealthy” along with developed to explain their choice of reference class. I looked at the list and it didn’t seem obviously cherry picked. What countries would you add?
Without looking at the data, of course.
The proper reference class probably depends on what this is being used as a proxy for: the ratio of self-described health to medical quality-of-life metrics is an odd enough figure that I assume it’s being used as a proxy for something. If we’re looking for degree of overconfidence in health care efficacy, which seems like the most likely candidate, using the first N countries ranked by per-capita health care spending might be the way to go: that gives you a list that’s not too dissimilar from the one in the article_per_capita), although some of the details are different.
That being said, once you actually start getting into the statistics, the US ends up in the middle of the rankings for most categories of disease and accident—it’s obesity-linked diseases, automotive accidents, and violence where it really shines. All of which isn’t too much of a surprise, but I don’t know if it’s much of an indictment of the American health care system on its own.
(There’s some odd features buried in there, though. For example, the US is ranked highly in deaths from chronic obstructive pulmonary disease and lung cancer, both correlates of smoking—but it’s middling-to-low in deaths from other cancers, indicating good oncology, and has a fairly low smoking rate. Air pollution’s also low. I have no idea what’s causing this.)
Re: parenthetical statement
Perhaps past smoking patterns are the important detail for some of smoking’s effects while more recent smoking determines the others. I wouldn’t be surprised to find that the US in the 50′s had (relatively) high smoking rates. Also the differences in female versus male smoking rates and disease susceptibility could be significant.