An early-offer/full-disclosure system may have reduced medical malpractice costs for the University of Michigan Health System in the last decade, but I find it hard to ascribe a great deal of confidence in the small-sample results because medical-malpractice results are "lumpy": all it takes is one outlier (or even just moderately right-tailed) jury verdict in the 1995-2001 timeframe before the apology system was put into place to completely distort the results. If anyone has a link to an ungated version, I'd love to see if this was a factor and/or acknowledged. [AIM abstract; WSJ Health Blog]
Regardless, it's probative evidence: it's entirely possible that early settlements plus the signaling that the hospital will quickly settle meritorious claims reduces the friction of litigation expenses (in part by deterring the meritless suits) so that total costs go down even as errors that previously would have been undetected result in settlements. And even if full disclosure did turn out to marginally increase litigation costs, one would expect it to result in more money going to pay compensation and less money to pay lawyers, as well as to improve the quality of medical care by discouraging defensive medicine, so that might well be a worthwhile tradeoff. One hopes that we see more institutions bravely experiment like this.