Grima Simon, Rupeika-Apoga Ramona, Kizilkaya Murat, Romānova Inna, Dalli Gonzi Rebecca, Jakovljevic Mihajlo
Department of Insurance, Faculty of Economics, Management and Accountancy, University of Malta, Msida, Malta.
Faculty of Business, Management and Economics, University of Latvia, Riga, Latvia.
Risk Manag Healthc Policy. 2021 Nov 26;14:4775-4787. doi: 10.2147/RMHP.S341500. eCollection 2021.
To statistically validate the PREM (Pandemic Risk Exposure Measurement) model devised in a previous paper by the authors and determine the model's relationship with the level of current COVID-19 cases (NLCC) and the level of current deaths related to COVID-19 (NLCD) based on the real country data.
We used perceived variables proposed in a previous study by the same lead authors and applied the latest available real data values for 154 countries. Two endogenous real data variables (NLCC) and (NLCD) were added. Data were transformed to measurable values using a Likert scale of 1 to 5. The resulting data for each variable were entered into SPSS (Statistical Package for the Social Sciences) version 26 and Amos (Analysis of a Moment Structures) version 21 and subjected to statistical analysis, specifically exploratory factor analysis, Cronbach's alpha and confirmatory factor analysis.
The results obtained confirmed a 4-factor structure and that the PREM model using real data is statistically reliable and valid. However, the variable Q14 - hospital beds available per capita (1000 inhabitants) had to be excluded from the analysis because it loaded under more than one factor and the difference between the factor common variance was less than 0.10. Moreover, its Factor 1 and Factor 3 with NLCC and Factor 1 with NLCD showed a statistically significant relationship.
Therefore, the developed PREM model moves from a perception-based model to reality. By proposing a model that allows governments and policymakers to take a proactive approach, the negative impact of a pandemic on the functioning of a country can be reduced. The PREM model is useful for decision-makers to know what factors make the country more vulnerable to a pandemic and, if possible, to manage or set tolerances as part of a preventive measure.
对作者之前论文中设计的大流行风险暴露测量（PREM）模型进行统计验证，并根据真实的国家数据确定该模型与当前新冠病毒病病例数（NLCC）以及当前新冠病毒病相关死亡数（NLCD）水平之间的关系。
我们使用了同一主要作者之前研究中提出的感知变量，并应用了154个国家的最新可用真实数据值。添加了两个内生真实数据变量（NLCC）和（NLCD）。数据使用1至5的李克特量表转换为可测量值。每个变量的所得数据输入到社会科学统计软件包（SPSS）版本26和结构方程模型（Amos）版本21中，并进行统计分析，具体包括探索性因子分析、克朗巴哈系数和验证性因子分析。
所得结果证实了一个四因子结构，并且使用真实数据的PREM模型在统计上是可靠且有效的。然而，变量Q14——每千名居民可用医院病床数必须从分析中排除，因为它在多个因子下有载荷，且因子共同方差之间的差异小于0.10。此外，其与NLCC的因子1和因子3以及与NLCD的因子1显示出统计学上的显著关系。
因此，所开发的PREM模型从基于感知的模型转变为基于现实的模型。通过提出一个使政府和政策制定者能够采取积极主动方法的模型，可以减少大流行对一个国家运作的负面影响。PREM模型有助于决策者了解哪些因素使该国更容易受到大流行的影响，并在可能的情况下作为预防措施的一部分进行管理或设定容忍度。
