Reflections on our longitudinal study of healthcare organization change

This paper reflects on a nine year longitudinal field study of organizational change following the 1994 merger that created a large Midwestern U.S. healthcare system integrating hospitals, physician clinics, ancillary services, and health plans. Using an engaged scholarship approach, researchers and system leaders jointly tracked integration processes in real time through sustained field observations, repeated interviews with managers and physicians, and annual surveys of employees from 1994 to 2003. The paper highlights three sets of lessons. First, it describes a PhD training fellowship embedded in the research program that supported fourteen doctoral fellows and generated multiple dissertations, illustrating both the opportunities and constraints of cumulative, team based longitudinal research. Second, it summarizes methodological innovations for transforming extensive qualitative and quantitative materials into analyzable process data, including the construction of a coded chronological event database capturing changes in ideas, people, structures, relationships, resources, context, and outcomes. Third, it outlines two substantive contributions. Frontier and fitness landscape analyses using Data Envelopment Analysis provided a multidimensional view of clinic performance and revealed unintended consequences of system level policies for different clinics. Event sequence analyses yielded a grounded, cyclical model of integration in which well intentioned actions produced both virtuous and vicious cycles, often increasing within unit integration while eroding integration across units. The paper concludes with practical and scholarly implications for designing, conducting, and learning from longitudinal process studies of healthcare organization change.