Evidence-based communication on climate change and health: Testing videos, text, and maps on climate change and Lyme disease in Manitoba, Canada

about it?”). They proceeded in the same way for the two other materials. The materials were presented in randomized order across focus groups. For the second part of the discussion, the facilitator prompted participants to discuss the comparison of all three materials (e.g. “Which material was the most and least effective?”, “How does the focus of all materials compare?”). Each focus group discussion lasted approximately 90 minutes. Written individual informed consent was obtained by participants before starting the focus group discussions, and partici- pants were each compensated one hundred dollars for their time. All discussions were audio recorded, transcribed verbatim and checked for accuracy. Analysis We conducted qualitative thematic analyses on the discussion transcripts in a qualitative anal- ysis software (NVivo 11.4). The coding procedure was developed jointly by the four authors and two researchers independently coded all the data, with a third researcher being consulted in case of disagreements. First, we organised the segments into categories according to if they discussed a) the article, b) video, c) map, or d) the comparison of materials. Second, we identi- fied themes describing participants’ responses to each material separately. We coded themes describing participants’ initial reactions to materials (e.g. confusion, skepticism, interest), atti- tude and behaviour change, and their perceptions of the key message of the material. These themes were compared across all three materials to identify similarities and differences in

responses across the materials. Third, we coded the segments in which participants were explicitly asked to compare the materials, by identifying key themes emerging from the discus- sions. To guide our identification and interpretation of key themes, we calculated frequency counts of each theme and compared them across materials and across levels of concern for cli- mate change (high vs. low concern, as indicated on the pre-screening questionnaire). Results Overall, participants critically evaluated the materials and they expressed both positive and negative responses, and generally participants in high climate concern groups discussed more positive responses while those in low climate concern groups were more negative and critical. Themes that emerged to describe participants’ responses are presented first according to each material separately, followed by themes emerging in the discussion of comparison of materials. Differences and similarities between high and low climate concern groups are highlighted throughout. Material perceptions Article. When reading the article, many people across all six groups expressed that the specific statistics and facts about Lyme disease were something that stood out to them, either as new information or something they thought was interesting and compelling, including: the rate of spread of blacklegged ticks per year; the increase in cases of Lyme disease; the fact that blacklegged ticks were not previously in the Prairies; and the definition of Lyme disease. When

asked what the key message of the article was, the majority of people (in 4 out of 6 groups) said it was the connection between Lyme disease and climate change, while a minority (2 out of 6 groups) said it was primarily that Lyme disease and ticks are increasing, and there is a need to adapt. Across all three low climate change concern groups, there were many people in each group who remained skeptical of the link between climate change and Lyme disease after reading the article. Many said that the article did not show proof that the rise in Lyme disease is related to climate change, while others said they simply “don’t buy it”. In one low climate concern PLOS ONE Evidence-based communication on climate change and health PLOS ONE | https://doi.org/10.1371/journal.pone.0252952 June 10, 2021 5 / 17