The COVID-19 pandemic presented unprecedented challenges globally. While many past pandemics (Spanish Flu, Ebola, SARS, etc.) offer historical context, the unique nature of COVID-19 and societal changes necessitate new approaches to understanding and mitigating its spread. This study aimed to address the critical question: Why did the COVID-19 outbreak spread so rapidly in its early stages? It is hypothesized that inadequate public health communication and response contributed significantly to the rapid spread. The study focuses on Japan, analyzing a large dataset to provide policymakers with scientific evidence for improved public information strategies. The study adopts a life-oriented approach, examining a broad range of behaviors and attitudes to gain a comprehensive understanding of individual responses to the pandemic.

Existing research on past pandemics and outbreaks (e.g., the 1918 influenza pandemic, Ebola outbreaks, SARS) highlights factors like risk perception, information reliability, and public health interventions. However, these studies often lack comprehensive modeling analyses or fail to capture the full spectrum of behavioral changes in modern society. Cowling et al. (2020) conducted a study in Hong Kong, examining correlations between non-pharmaceutical interventions and COVID-19 spread, but relationships between factors remained unclear. This study seeks to address this gap by employing a broader, more detailed approach that incorporates a comprehensive set of behavioural and psychological factors and more advanced analytical methods.

A nationwide retrospective postal survey was conducted in Japan at the end of March 2020, when the country had 1953 confirmed cases. 1052 participants provided valid data, ensuring a balanced distribution across age, gender, and region. The survey employed a life-oriented approach, capturing changes in daily activities (trips, work/school, shopping, social activities, etc.) and attitudes towards the pandemic, government responses, and information sources. Data included changes in various activities from December 2019 to March 2020. Potential factors influencing behavior were analyzed, including direct triggers (government requests, personal understanding), information reliability, risk perception (personal and perceived risk for others), and attitudes towards government policies and communication. Aggregate analysis and structural equation modeling (SEM) were used to assess the relationships between these factors and behavioral changes. A data mining approach was also employed.

Significant changes in daily life were observed, with a substantial decrease in activities involving close contact and travel (e.g., avoidance of crowded places, reduction in daily trips and public transport usage). The government's February 28th announcement of school closures acted as a major trigger for behavioral change. However, a substantial proportion of respondents showed concerns about government communication, information reliability, and the effectiveness of policies. Analysis revealed that communication from close social network members (family, friends, colleagues) had the most influence on behavioral changes. Risk perception and information reliability were also significant factors. The SEM analysis indicated that 'triggers of behavioral changes' were the most influential factor on changes in daily life, followed by risk perceptions. Information reliability and attitudes, however, did not show a significant direct effect on changes in daily life, though information reliability showed significant effects on attitudes. The data mining approach further highlighted the need for differentiated communications tailored to different behavioral changes. For example, changes in daily trips were influenced by self-understanding and information from health institutes, while avoiding crowded places was influenced by government requests and risk perception.

The findings demonstrate that improving public communication is crucial for mitigating pandemic spread. The emphasis on social networks suggests a need for interventions that leverage social influence. The identified concerns about government communication highlight the need for transparency, clear and consistent messaging, and building public trust. Furthermore, the study underscores the need for a multi-faceted approach, tailoring communications to different behaviors and targeting specific groups within the population. The LASTING approach is proposed for better implementation of physical distancing and mitigation of infection risks.

This study provides valuable insights into individual responses to the early stages of the COVID-19 pandemic in Japan. It highlights the crucial role of effective public communication, tailored strategies targeting social networks, and addressing the specific needs and concerns of the population. Future research should focus on refining communication strategies, exploring long-term behavioral impacts of the pandemic, and developing more robust models that incorporate human behavior to improve pandemic preparedness and response.

The study is limited to the Japanese context and the early stages of the pandemic. The retrospective nature of the survey may introduce recall bias. The sample size, while large, might not fully capture the diversity of responses across all population segments in Japan. Future research might consider longitudinal studies to assess long-term behavioral changes and the evolution of public attitudes.