Daniel Hudson, Chistine Holzmueller, Peter Pronovost, and others have published an article entitled, "Toward Improving Patient Safety Through Voluntary Peer-to-Peer Assessment."  It is in the American Journal of Medical Quality where, regrettably, only the abstract is available without paying:

Health care has primarily used retrospective review approaches to identify and mitigate hazards, with little evidence of measurable and sustained improvements in patient safety. Conversely, the nuclear power industry has used a prospective peer-to-peer (P2P) assessment process grounded in open information exchange and cooperative organizational learning to realize substantial and sustainable improvements in safety. In comparing approaches, it is evident that health care’s sluggish progress stems from weaknesses in hazard identification and mitigation and in organizational learning. This article proposes creating and implementing a structured prospective P2P assessment model in health care, similar to that used in the nuclear power industry, to accelerate improvements in patient safety. 

The article notes that there has been a lack of progress in patient safety, merely 1% between 2000 and 2005, despite major efforts by many advocates and participants.  Health care also lacks safety-related performance measures in most clinical areas. "Most errors result from good clinicians working in complex and hazardous systems."  Thus, we need to recognize the fallibility of people redesign systems and the manner in work is done to "anticipate and mitigate inevitable human error."

The authors note that other high-risk industries have recognized similar patterns and have put into place more effective strategies for mitigating them.  I discussed the airline example below.  These authors suggest looking at the nuclear power industry as another example:

Health care’s marginal improvement in patient safety contrasts sharply with the remarkable success observed in the nuclear power industry. After devastating nuclear accidents, this industry implemented a voluntary and proactive peer-to-peer (P2P) assessment program to improve plant safety and reliability throughout the world.

Where health care falls short is its approach to identifying "latent factors" that can provoke or create weaknesses in system defenses. "When latent factors combine with unintentional slips or lapses or with intentional acts that bypass safeguards, an accident can cause an adverse event."  But latent factors can be identified in advance and systems can be redesigned before adverse events occur

One approach to this is to engage in peer review, but the kind of peer review that occurs in health care is "grossly inadequate" to this task.  That is the case because of the methods used but also because of the context of such reviews:

Inevitably, most physicians associate “peer review” with aspersions of negligence, misconduct, or malpractice and feelings of blame, shame, and fear. This culture has promoted an atmosphere where there is a fear of judgment and humiliation rather than one conducive to learning, improving, and protecting patients.

So the authors instead propose a structured peer-to-peer ("P2P") assessment process to supplement the usual case reviews that take place in hospitals.  The necessary conditions for success of such a program would take some work to put in place, but they would include establishing or identify organizations that could coordinate and oversee an independent, confidential, and external P2P assessment process; developing validated tools and a reliable process; and establishing a training model and training peer evaluators.  Of course, a sustainable funding model and cooperation from hospitals would be necessary.

This is all good stuff, but there is one thing that the authors might have missed.  They note that peer evaluators potentially would have to devote valuable time to P2P assessment activities, time "away from work" that might be a major barrier.  I would urge them to restate this.  It would not be time "away from work," in that the insights gained by participating in a P2P process at another hospital would inevitably bring benefits to the evaluators' home institution as well.  Doctors often spend time out of hospital delivering and hearing professional papers.  Certainly, the P2P activity could be determined to be equally valuable if hospital chiefs of service and other physician and nurse supervisors rewarded participation in the same manner they reward travel to conferences.