Cell-phone traces reveal infection-associated behavioral change

摘要: Epidemic preparedness depends on our ability to predict the trajectory of an epidemic and human behavior that drives spread in event outbreak. Changes during outbreak limit reliability syndromic surveillance using large-scale data sources, such as online social media or search behavior, which could otherwise supplement healthcare-based outbreak-prediction methods. Here, we measure change reflected mobile-phone call-detail records (CDRs), a source passively collected real-time behavioral information, anonymously linked dataset cell-phone users their date influenza-like illness diagnosis 2009 H1N1v pandemic. We demonstrate use differs measurably from routine behavior: Diagnosed individuals exhibit less movement than normal (1.1 1.4 fewer unique tower locations; P 3.2 × 1 0 − 3 ), average, 2 4 d around place calls (2.3 3.3 calls; 5.6 ) while spending longer phone (41- 66-s average increase; 4.6 10 usual day following diagnosis. The results suggest CDRs health may be sufficiently granular augment efforts infectious disease-modeling lacking explicit behavior-change mechanisms need revisited.