A comparison of community detection methods found that modularity optimization (MO) significantly outperforms Dartmouth Health Referral Regions (HRRs) in delineating emergency general surgery (EGS) networks. An analysis of 1,244,868 patient encounters in New York and California revealed that MO identified fewer but more geographically accurate regional networks, with substantial reclassification of hospitals between communities. These findings suggest that network science methods can enhance the coordination of regional care systems for EGS patients.
Journal Article by Han J, Wan N, Horns JJ and McCrum ML in JAMA Netw Open