Network Influences on Chronic Illness Care in Large Physician Organizations: A study of the California managed care network in 2001 and 2006
- Author(s): McCurdy, Rodney Keith
- Advisor(s): Shortell, Stephen M
- et al.
Although the prevalence and burden of chronic illness is immense and growing, national studies have found minimal use of evidence-based practices to improve the management and care provided for patients with these conditions. A productive body of work has identified facilitators and barriers to the use of care management processes (CMPs) in large physician practices using an organizational capabilities and external incentives framework. Physician groups enter into multiple affiliations in order to best position themselves to achieve economic success and to maintain professional autonomy. To date, little is known regarding the impact of these interorganizational networks on CMP use in physician groups and many experts have called for the incorporation of social network methods to explore this area.
This study examines the relationships between network contagion and structural influences and CMP use among physician organizations in a managed care network. Specifically, the study addresses three main questions: 1) Do interorganizational ties influence the use of CMPs in physician organizations?; 2) Do external relationships influence how CMPs are used to treat patients with chronic conditions?; 3) Does change in network relationships influence change in CMP use in the practice setting?
To address these questions, the study examined the managed care network in California during 2001 and 2006. Two unique datasets were merged to allow network analyses of the entire network of physician organizations, hospitals and hospital systems, and health maintenance organizations (HMOs) that participated in the network during this timeframe. An exchange perspective was employed to examine interorganizational relationships and network position based on the functions and exchange needs of the various actors in the market. Statistical and visual representations of the entire network were accomplished for both years. Cross-sectional statistical analysis was performed to test for the influence of network contagion and structural effects on CMP use in 347 physician organizations and in homogeneity of CMP use among almost 30,000 pairs of physician groups in their treatment of four chronic conditions. Finally, a longitudinal analysis was performed to test for the effects of change in network affiliations on CMP homogeneity among almost 4,000 cohort pairs. This is the first known study to employ this methodology in the study CMP use in physician organizations.
Network analysis results highlight several important changes in exchange relation attributes among physician organizations in the California managed care network from 2001 to 2006. First, the overall size of the network in terms of the number of physician organizations participating in managed care declined considerably. Almost 25 percent of physician groups in 2001 had exited the network five years later. The majority of these organizations had closed due to financial hardships. The surviving physician organizations appear to be gaining network advantage in relationships with hospital systems and HMOs as these organizations have fewer structurally equivalent physician groups to meet their exchange needs.
Empirical results supported the influence of contagion effects on CMP availability and use in the physician organization. A physician group's portfolio of direct interorganizational ties with hospitals and physician organizations was associated with greater odds of CMP use at the organizational level. Results of the cross-sectional pairwise analysis found strong support for the contagion influence of shared physician group ties and shared affiliations with hospital systems on homogeneity in CMP use for four chronic conditions (asthma, congestive heart failure, diabetes, and depression).
Network structural effects were found to influence homogeneity in CMP use among physician organizations occupying similar locations within the network space. Physician groups sharing space within the core were found to be more homogeneous than groups occupying separate locations in the network. Conversely, CMP homogeneity declined over the five year period for among groups sharing space in the periphery. CMP use appears to be concentrated within the core of the managed care network and results of the cohort analysis failed to find evidence of diffusion toward the periphery. The findings suggest that peripheral physician organizations with few ties to core exchange partners may be at a structural disadvantage when it comes to obtaining the resources necessary to implement CMPs.
This study found that jointly interdependent exchange relationships are influential in the prevalence of CMPs in physician practices and in how they are used for chronic illness care. Implications of these findings suggest that current discussions surrounding the development of Accountable Care Organizations (ACOs) should be expanded in order to establish more linkages between core and periphery organizations in the health care network.