We live in a country of 300 million people. Every permutation of things that can happen has happened. But that doesn't mean that large verdicts "where no or minimal injury was suffered" happen often enough to account for a significant share of medical malpractice payouts.
There was a Harvard study that looked at this issue pretty systematically, and concluded that the idea of "systematically" "large verdicts against defendants who are unlikely to have done anything wrong" does not fit the data: http://archive.sph.harvard.edu/press-releases/2006-releases/.... Most importantly, claims are about as likely to get denied despite the presence of error as they are to get paid despite the absence of error. (So improving the accuracy of the system would not necessarily decease payouts.)