Sometimes I go off about the huge RAND study, which showed that giving people free care can lead to greater costs without improved outcomes, except for in a few areas like heart health and optics. Well, there's a moment in the interview above where Atul Gawande essentially agrees that it's a bimodal* issue, but he adds mental health issues to the list where cheap substitutes (or skipping care) is really, really bad. He then notes that those are the areas where most expenditures are found (maybe not optics, but literal hearts and minds).
*ie, there's two clusters in the data. (Here, one overconsuming unnecessary care and another underconsuming necessary care. Gawande would argue that, by the numbers, the second group is a bigger worry.)
My favorite NPR piece on health care is still from This American Life, on all the ways patients, doctors, insurers, drug companies and hospitals each make things worse in their own way. (Part Two)
And it's always important to note that medical bills cause two out of three bankruptcies. There's this common cycle where expensive consumers of health care have to miss large amounts of work. The incentives and the justification are there for the employer and insurance company to push this person out of work. The individual now has impossible bills and no job.
It's a fact that was common in the rhetoric of the bill, but wasn't directly addressed. Some costs are lowered and caps are used for consumers, but medical bankruptcy law remains essentially the same. It remains to be seen how this will turn out.