Between March 27 and 28, 2020, a total of 4,000 participants in Japan stratified by age (20s, 30s, 40s, 50s, and ≥60 years) and gender (400 in each group) took part in this study. Between April 17 and 21, 2020 a total of 2,000 participants in the UK and 2,000 participants in Spain stratified by age (20s, 30s, 40s, 50s, and ≥60 years) and gender (200 in each group) took part in the study. The concurrent situations in each country are presented in Table 1.
We first excluded invalid answers from the analysis because there might be some participants who gave non-serious answers considering that participating in the survey would be rewarded. Participants whose answer to all 30 of the items in section C, D, E, and F excepting number of hand washing was '1' were excluded. As a result, 19 participants in Japan, 13 in the UK, and 5 in the Spain were excluded from the analysis. The remaining 7,963 participants were subjected to the analyses.
The participants' demographic data are presented in Table 2. Regarding educational background, the participants in Japan had a higher percentage of university degrees than the participants in the UK and Spain (χ2-test, Pearson χ2 = 674.390, df = 6, p<0.001, adjusted residual for university degree or above of the participants in Japan = 2.4). Compared to the participants in the UK and Spain, those in Japan had a lower percentage of having infected people among their family or relatives (χ2-test, Pearson χ2 = 2055.999, df = 4, p<0.001, adjusted residual for participants in Japan for whom this answer was 'yes' = −38.3), or at their workplace (χ2-test, Pearson χ2 = 1139.305, df = 4, p<0.001, adjusted residual for participants the answer 'yes' in Japan = −18.6).
Regarding the level of knowledge about COVID-19, the results of this series of surveys revealed definite and significant differences among the three countries, as illustrated in Figure 1. The participants in the UK and Spain had a deeper understanding of COVID-19 than the participants in Japan (ANOVA with Games-Howell test) at least in their subjective thought.
Regarding anxiety about COVID-19, there were also significant differences among the countries (Fig. 1). The participants in Spain were more anxious about being infected with COVID-19 than those in Japan, but less anxious than those in the UK. The participants in Japan were less afraid of infecting others with COVID-19 than those in the UK and Spain. Regarding their anxiety about the severity of the disease once infected, the participants in Japan were less concerned than those in the UK and Spain. Concerning the spread of COVID-19, the participants in Spain were the most anxious, followed by those in the UK and then Japan.
We examined the frequency of access to and the credibility of a variety of sources of information about the virus and pandemic, and the participants' survey responses revealed some country-specific characteristics (Fig. 2). The participants in Japan reported having significantly less frequent access to any form of information compared to those in the UK and Spain. Remarkably, the proportion of participants in Japan who said that they had never had any access to official announcements, radio, or specialists was twice as high as that of the other two countries. The participants in Spain had more frequent access than those in the UK to information sources such as the government, a social network, radio, friends and neighbors, and specialists.
Regarding the credibility of the COVID-19 information sources, the participants in Japan were unlikely to trust most of the types of information sources compared to those in the UK and Spain. The participants in Spain had less trust in the information from the government compared to the participants in Japan and the UK. Social networks and friends and neighbors were not deemed a credible information source in any of the three countries, but the participants in Spain reported slightly more trust in these sources compared to those in Japan and the UK.
We compared the number of daily handwashing among countries. The range of handwashing was 4–5X/day in all three countries (data not shown). A Kruskal-Wallis test with Dunn-Bonferroni's correction revealed that the participants in Spain washed their hands significantly fewer times over the course of a day than the participants in the UK (p<0.001), but more times than those in Japan reported in the past survey (p<0.001).
The questionnaire answers regarding precautionary behaviors are illustrated in Figure 3. Among the active behaviors, the patterns of handwashing and using disinfectant showed significant differences among the countries, but the elements of these differences were complex. For example, some UK respondents washed their hands very frequently, although there were also more respondents in the UK who never washed their hands compared to Spain and Japan. We also observed that 57.7% of the participants in the UK never wore a medical mask whereas 16.1% and 9.5% of those in Spain and Japan did, respectively. The participants in Japan were less likely to engage in avoidance behaviors compared to those in the UK and Spain. Notably, the participants in Japan were far less likely to avoid school or work and less likely to avoid using public transport than the other countries' respondents. The majority of the participants in the UK and Spain reported that they rarely or never went to school or work, or used public transport.
As an additional analysis, we compared the data of the degree of knowledge, anxiety, frequency and credibility of virus information, and precautionary behaviors between the participants who had acquaintances who were affected by the virus and those who did not. Almost every item in the questionnaire showed a significant difference between these two groups.
We reanalyzed the differences in the items presented above between countries by examining only the participants who did not have an infected acquaintance. The influence of eliminating the answers of the participants with an infected acquaintance was limited.