Laura Leviton of the Robert Wood Johnson Foundation discussed positive deviance and similar methods from other areas. In education literature, for example, the term “school improvement methods” is used. Some of these methods have been successfully implemented in other fields and could be useful in health care, Leviton said.
The difficult portion of reducing variation, Boat said, will be allowing medical practices to account for context while treating individual patients and understanding the various confounders of genomic background, environmental exposures, and psychosocial contexts.
Ignatius Bau of the California Endowment agreed that the discussion of context is critical. There is the assumption that context does not matter in quality improvement, he said, but this assumption conflicts with the research agenda that attempts to produce generalizable processes. What is known about changing culture, such as changing provider behavior and changing processes of care in team-based environments, should be a necessary component of this conversation. Bau also discussed the need to understand both resistance to change and why, according to performance measures, best practices are not followed every time a patient is treated.
There is a large co-occurrence of mental, addictive, and general health conditions, Pincus said, citing the findings of the IOM report Improving the Quality of Health Care for Mental and Substance-Use Conditions (Institute of Medicine, 2006). But resource investment in a quality improvement infrastructure and improvements in quality that have been made in health care have not been paralleled in mental health care. Partnerships, among other strategies, should also be considered, he said. AHRQ’s Integrated Delivery System Research Network1 could be applied in mental health settings, for example. Quality improvement methods have not yet permeated the areas of mental health and substance use, where they could potentially have a great impact.