The purpose of this study is to grasp the current situation of disaster prevention measures in special elderly nursing homes. A questionnaire survey was conducted in special nursing homes nationwide, and received responses from 1685 facilities. All facilities were classified into four types, based on the number of disasters in each of three phases, then further subcategorized into ten types. In addition, the whole country was divided into ten regions, then further classified into four groups following the conduct of the principal component analysis and cluster analysis for eleven disaster prevention response items for all facilities. Next, all facilities were analyzed from the mutual relationship of “Disaster Prevention Response”, “Type”, and “Group”. The main results revealed are as follows.
1) An overview of disaster prevention measures at facilities nationwide shows that “Understanding hazard maps in the preparatory stage before a disaster occurs”, “Formulating a disaster prevention plan at night in the planning stage”, “Fire prevention training”, “Practical training” were carried out at a high rate. On the other hand, the measures were not sufficient toward “Formulation of BCP”, “Conclusion of evacuation support agreement”, and “Securing of emergency rations for more than four days”.
2) It turns out that there is a difference in the number of disasters for each facility at three phases: “Disasters that may occur”, “Disasters including in planning”, and “Disasters assumed in disaster prevention drills”. Excluding fires, many of the facilities toward the disaster prevention measures, gradually reduced the number of target disasters in the three phases.
3) When all facilities were classified by type, based on the degree of change in the number of disasters by phase,
it was possible to classify into four types: “Type A: Reality-based”, “Type B: Possibility-based”, “Type C:
Priority-based”, “Type D: Irregular-based”. “Type A: Reality-based” type, in which the number of disasters does not change with respect to the number of possible disasters, is the most common, then followed by “Type-C:
Priority-based”, in which the number of target disasters decreases, accounts for almost all. Besides, “Type A: Reality-based” and “Type C: Priority-based” have less “Disasters assumed in disaster prevention drills” than “Disasters included in planning” occupy roughly a half of the total.
4) The whole country was divided into ten regions, and the principal component analysis was performed on eleven disaster response items, and two factors were extracted per region. Furthermore, each facility was classified into four groups, based on the cluster analysis. As a result, the four regions of Kyushu, Kanto, Hokuriku, and Kansai were related to “Type”, “Group”, and “Disaster response items”.
5) Taking the Kanto region as an example, the seven prefectures in the Kanto regions were divided into two main groups, based on the content of disaster response. Tokyo and Kanagawa prefectures have high levels of “Factor 1: Plan/Training”, and “Factor 2: Practice/Response”. In the other five prefectures, the level is low on “Factor 2: Practice/Response”, and the disaster prevention drills are conducted by reducing the number of disasters per phase.
6) Similar to the Kanto region, due to the influence of local government conditions, as well as the conditions of the climate and the location, the regional characteristics are likely to be created, by not only the regional characteristics, but also at prefectural level.
