Facilitating practice change: lessons from the STEP-UP clinical trial.

BACKGROUND: The Study To Enhance Prevention by Understanding Practice (STEP-UP) clinical trial (1997-2000) resulted in sustainable increases in preventive service delivery in primary care practices. However, the process by which practice change can be facilitated has not been well described.
METHODS: Comparative case studies were conducted of eight STEP-UP practices with the largest increases in preventive service delivery rates and compared to seven practices with the lowest increases. A multidisciplinary team (research nurse, nurse facilitator, physician principal investigator) used an editing analysis approach to create individual case studies. Then, using an immersion-crystallization approach, the team identified pragmatic lessons for people working to improve primary care practice, and validated these lessons with a participating practice and an additional facilitator.
RESULTS: It is not always possible to predict which practices will change based on understanding initial practice conditions. "Malleable moments" can be identified during which practices become open to change. It is important to tie change strategies with existing motivations, or to develop new motivation among potential change agents. Motivation can be developed by discrepant information that challenges self-image, aligning change plans with existing values, or identifying feasible ways of responding to outside pressures or internal demands. Instrumental interventions (such as office systems, tools, new processes) are useful when motivation to change exists, and can build motivation when they meet a perceived need. Disruption in previously workable approaches, either by purposeful information seeking or unanticipated changes, promotes openness to change.
CONCLUSIONS: Despite limited ability to predict which practices will change and when, understanding practices' initial conditions and evolution can identify opportunities to craft individualized approaches to positive change.