There is only one report in the literature specifically addressing centralized review. It describes an initial year of experience with a centralized review process at a new hospital and offers anecdotal evidence of effectiveness. (1)
In contrast, our national study of peer review practices systematically examined more than 3 dozen aspects of peer review program structure, process and governance among 339 institutions. (2) Our primary outcome variable was the perceived ongoing impact of the program on the quality and safety of care. The organization of peer review activity in terms of centralization or de-centralization was unrelated to the quality impact.
On the other hand, the degree of standardization of the review process, the diligence with which it is governed, and the timeliness of performance feedback to clinicians were, among other factors, significant predictors of perceived effectiveness.
This makes sense. From a quality improvement perspective, there is no a priori rationale why a centralized program would be more effective, other than via some other parameter such as standardization.
There is a significant potential downside of centralized review in a larger hospital: lower case review volume. In our study, case review volume below 1% of hospital admissions was associated with lessened perceived effectiveness. Furthermore, in contrast to programs with de-centralized review activity, those with centralized or partially-centralized activity were less likely to report review volume over 1% with an OR (CI) of 0.46 (0.24-0.87). Centralized or partially centralized review activity was associated with greater perceived quality when controlling for case volume, but not when also controlling for the degree of process standardization.
I believe the explanation for this is relatively simple: many hands make for light work. A typical review committee only meets for 1-2 hours a month. This will often include the time for chart review. It is not possible to conduct a meaningful discussion on more than 10 or so cases in that interval. So, on average, one committee meeting 10 times a year might be expected to review roughly 100 cases. Thus, if targeting a 2% review rate, any hospital much larger than 100 beds needs either multiple committees or a more effective strategy to manage case volume.