In the US, guidelines that seek to reduce cancer screening among certain patients often fail to impact clinical practice. Physician's social networks-the web of relationships among individuals-have been shown to be an important factor in the diffusion of innovation and may also be critical in settings where guidelines may conflict and where some guidelines may seek to roll-back, or de-implement, clinical practice. The objective of this exploratory R21 grant is to characterize how information from physicians' social networks defined as their colleagues, friends, and family members along with experiences from their own clinical practice may alter perceptions and recommendations regarding breast cancer screening and how this information affects population rates of screening.
In Aim 1, we will conduct a national survey of physicians to characterize the extent to which their breast cancer screening recommendations are influenced by their experiences with medical colleagues, knowledge of friends and family who have been diagnosed with breast cancer, and encounters with their patients.
In Aim 2, we will use the results of the survey to incorporate physicians into our existing agent-based model of patients' breast cancer screening decisions. The resulting model will allow study of interactions between physicians and patients, capturing the resulting feedback loops. Using this model, we will simulate different interventions by altering how social networks influence patient and provider decision making. With network interventions being increasingly deployed and the evidence base for cancer screening continuing to evolve, the proposed project will help prioritize aspects of interventions that may have the greatest likelihood of success in changing clinical practice.
Social networks have the potential to influence perceptions of risks and benefits associated with preventive health behaviors and may be a crucial factor in efforts to promote changes in clinical practice. This project will examine how physicians' social networks influence their recommendations for breast cancer screening (through surveys) and shape population rates of screening (using simulation models).