Loss of participation among evacuees aged 20–37 years in the disaster cohort study after the Great East Japan Earthquake

Abstract

The present study aimed to clarify the characteristics of young evacuees who had missed the Comprehensive Health Check of the Fukushima Health Management Survey (FHMS) after the Great East Japan Earthquake in 2011. The FHMS has been conducted as a prospective cohort study to evaluate the health status of evacuees annually after the great earthquake in 2011. This study focused on the annual participation rate in the Comprehensive Health Check of evacuees aged between 20 and 37 years in 2011 who evacuated due to the Fukushima Daiichi Nuclear Power Plant accident. The characteristics of subjects who did not participate after the second survey year were identified with a multivariate logistic regression model. The participation rate was estimated at 26.6% (9720 among 36,502 residents) and 15.6% (5691 residents) in 2011 and 2012, respectively. The logistic regression model revealed the following characteristics at baseline as independent predictors of non-participation after the second year of the survey: age ≤ 24 years (adjusted odds ratio 2.11, 95% CI 1.84–2.42), 25–29 years of age (1.28, 1.13–1.45), men (1.52, 1.38–1.69), evacuation outside the municipality but within Fukushima prefecture (1.54, 1.40–1.70), evacuation outside the Fukushima prefecture (1.40, 1.21–1.63), anemia (1.23, 1.06–1.43), smoking habit (1.34, 1.21–1.48), and drinking habit (1.20, 1.09–1.32). A medical history of heart disease showed opposite odds ratios, which indicate the association with continuous participation (0.43, 0.26–0.72, respectively). We observed deteriorated participation in the prospective study of the Comprehensive Health Check of the FHMS among evacuees of a younger age group, men, those evacuated outside their municipalities, and those with history of anemia, smoking and drinking habits. Hence, the cohort study may have missed certain population groups with worse health behaviors. Thus, it is necessary to consider various measures to increase the participation rate in the disaster cohort study to understand the long-term health effects of disasters on younger residents in evacuation zones.

Categorization of disaster-related deaths in Minamisoma city after the Fukushima nuclear disaster using clustering analysis

“…However, in non-standard cohort studies, such as disaster or pandemic studies, the methods of participant enrollment and data collection may differ from those of regular cohorts, thereby making it difficult to directly determine the overall characteristics from the collected data 32 – 34 . Therefore, to understand the overall characteristics of disaster-related deaths, it is important to develop a typology of these types of death cases and categorize each death into a group of similar type of cases.…”

Strategies and costs of building and maintaining a longitudinal disaster cohort

“…Few longitudinal studies have methodically traced disaster restoration processes beyond a year or two (McLeod et al 2021). While there are many longitudinal studies that discuss the importance of maintaining cohort retention over time (Abshire et al 2017; Booker, Harding and Benzeval 2011; Byrd-Bredbenner et al 2017; Cotter et al 2005; Fitzgerald et al 2019; Hanna, Scott, Schmidt 2014; Marmor et al 1991; Meyers et al 2003; Scott 2004; Sowmya Rao et al 2005; Teague et al 2018; Vincent et al 2012; Wenemark et al 2010; Yamamoto et al 2022) and the pitfalls of attrition bias (Odierna and Schmidt 2009), there are few studies that specifically address the recruitment and retention strategies in longitudinal studies after a natural disaster (McLeod et al 2021). Pathways to recovery vary greatly among disaster-exposed individuals (Masozera, Bailey and Kerchner, 2007), and the most vulnerable populations are typically the most difficult to reach in longitudinal follow-ups—which can lead to low retention rates and introduce attrition biases to the data (Odierna and Schmidt 2009).…”