|dc.description.abstract||Low-resource countries are disproportionately affected by infectious diseases and residents living in rural areas in these countries are more likely to experience geographic or infrastructural barriers that limit their access to formal health care or information. For health interventions in these areas to be most effective, information should be tailored for their audience and then disseminated through relevant communication channels. Interventions that utilize existing social networks and that learn about how their audiences talk about the topic of interest are more effective than those that do not. This thesis used a case study in Sri Lanka to 1) identify central actors in wildlife, livestock, and human health information networks and 2) to examine themes and topics that arise during discussions about wildlife, livestock, and human health.
One-hundred and forty-three rural residents were interviewed to identify their main sources of wildlife, livestock, and human health information and to identify to whom they would report these health issues. Social network analysis of the responses revealed that government agency staff, such as the Grama Niladhari and government physicians, were the most frequently cited source of wildlife and human health information and the most common place to report health cases. A local indigenous healer was the most common source of livestock health information, the most common person to report livestock health cases to, and best positioned in each of the health networks to disseminate information and receive reports within the community. Women were more likely to be unsure of who to talk to and were considerably less likely to be nominated as a source of health information than men. Locally relevant and central leaders that are seen as key contacts for wildlife, livestock, and human health issues should be engaged and used to effectively disseminate information to and from the community. Government agencies should also engage with and maintain relationships with rural communities to facilitate information sharing. The gender differences shed light on the importance of engaging and accommodating all groups within a Sri Lankan community, perhaps by identifying group-specific opinion leaders that will appropriately communicate information to and from the group.
To learn about health discussion topics, a structural topic model was used to identify main topics that emerged in 7,412 survey responses and to examine gender differences among the topics. Seven topics were identified by the topic model: 1) Cost/benefits of living near forest, 2) Reporting/asking about animal health, 3) Diseases caused by animals, 4) Wildlife visits and consequences, 5) Issues and needs of the village, 6) Village societies, and 7) medicine. There were small but significant gender differences for Topics 1-6 which indicated that men and women were spending different amounts of time on different topics. However, given the small gender effect sizes, which ranged from 0.3%-1.6%, it was concluded that gender has a relatively very small influence on these topics. Further research should investigate the specific words and rhetoric males and females use to describe health topics to uncover small nuances that broader methods cannot.||