As of this writing, 35 states require health care facilities to request a Certificate of Need (CoN). If you want to start a new hospital, nursing care facility, or dialysis center, or eve buy a new MRI machine, the CoN application process puts the burden on you to prove that there is sufficient demand that existing providers won't be harmed by a new entrant. Along the process, existing facilities can typically give input as to whether there is a need for new entrants.
Some of the rationales for CoNs given in the 1960s were based on that Soviet-era belief that economies of scale are such that you're better off with one giant hospital instead of several small ones. Other policy goals, like how we want equitable access and not just hospitals in the rich parts of town, are often added to the CoN process, even though in a world with no CoN there would be other mechanisms to achieve those policies.
But in my mental word cloud around health care, the two biggest words are expensive and necessary. Are hospital, dialysis, and MRI beds really the place to be carefully controlling supply? I can't think of any other industries where new entrants can't operate until they prove they won't take business away from the existing.
If you want the hard numbers, you won't quite get them. There have been many studies hoping to find an effect on health or cost outcomes one way or another—maybe skip to the bullet-pointed list of appendix C of This sort of a metastudy—but the big impression I get from the literature is that there's not enough of a resounding effect to overcome all the ways people have to bend linear regressions.
From there, I'll just pass you on to Wikipedia, because my goal here is not to give you a long persuasive essay on why or how supply restrictions support incumbents or can raise costs, but to let you know that this little piece of friction exists.