Interpreting these findings is difficult. The authors appropriately note the major limitations of their evaluation, including bias introduced by the nonrandom selection process, voluntary participation, the possible influence of regression to the mean in a before–after study, and the inability to distinguish the effects of QIO assistance from the impact of other concurrent quality improvement activities and exposures typically experienced by providers (an issue that is particularly relevant to hospitals, many of which do not participate in the QIO Program). An additional barrier to interpretation is the paucity of information about how the QIOs recruited provider organizations; the nature of the “general guidance” provided; and the content, timing, duration, and frequency of QIO interventions. We don't know which of the reported differences are statistically significant, although given the large number of organizations involved, one would assume that some were. Furthermore, the authors' analyses do not account for potentially relevant variables, such as organization size and ownership, availability of resources, degree of market competition, and the regulatory environment in the relevant state. The authors note that they tried to control for some of these factors, but they don't report the results.