Networks Influencing MRI Use
This article discusses a study, conducted by researchers at the Yale School of Medicine and Johns Hopkins School of Medicine, which found that surgeons’ use of imaging tests is influenced by the practices of their peers. The researchers found clusters of surgeons by looking at Medicare data to identify which surgeons frequently share patients with one another. Additionally, they decided to study MRI use because, within the medical community, there are some professional societies recommending against its routine use as part of planning for breast cancer surgery. They then looked at surgeons within these clusters to determine whether surgeons who did not originally use MRI, but their peers did, would be more likely to use MRI. The researchers found that “women whose surgeons were in peer groups that had the highest baseline MRI use were more likely to receive MRIs compared with women whose surgeons’ peers did not use MRI (24.9% vs. 10.1%).” Finally, the researchers in the study concluded that this social influence can “increase the use of medical approaches that are not based on sound evidence — or alternatively, this social contagion might also be used to promote best practices among physicians.”
This relates to our class and our discussion of the diffusion of ideas in networks. Novel ideas or in this case controversial medical techniques have an element of risk associated with them. Therefore, it makes sense that surgeons in clusters with the highest baseline MRI use would adopt MRIs at higher rates than surgeons in clusters with lower baseline MRI use because when these doctors choose between adopting the practice or not adopting the practice, the decision is partially based on what their neighbors are doing.This relates to our discussion of how the spread of new behaviors requires not just awareness but an actual threshold for adoption to take place. Furthermore, in class, we discussed that the diffusion of ideas in a network depends on the initial adopters as well as the network structure. The researchers in this study focused on doctors that shared patients, which suggests that the ties that these doctors had to each other were more likely to be strong ties instead of local bridges, therefore this network structure is conducive to the spread of this technique because of its clustered structure. Their findings can also be explained by our discussion of how when a behavior is risky or costly to adopt, then there is a higher threshold of neighbors doing it before the node switches. Finally, the higher rate of diffusion in one group versus another can also be explained by the concept of compatibility. If a doctor who doesn’t use MRIs receives many patients whose previous doctors used MRIs as part of planning for breast cancer surgery, consistency and compatibility with previous practices might explain why a doctor would adopt the practice.
Source: https://news.yale.edu/2017/05/15/social-contagion-exam-room-peer-influence-and-cancer-surgeons-use-breast-mri