Оптимальный подбор пар субъектов в исследованиях дизайна «случай-контроль»: демонстрация метода сопоставления на примере анализа влияния пандемии COVID-19 на персонал медицинских организаций

Abstract

Case-control studies are retrospective observational studies designed to determine the relationship between the outcome of interest and its potential cause. This study design has known disadvantages: possible bias in the results, selection bias in the control group, etc. In order to avoid inconsistency between the study and control groups and thus underestimating of the possible influence of intervening factors (confounder), a whole class of methods is used to ensure that case-control matching pairs are selected and matched. Objective. To discuss and demonstrate the possibility of forming groups of research subjects (impact and control) on the basis of matching variables using the optimal matching procedure and compare the results of paired and unpaired analysis. Materials and methods. The baseline data were the results of a cross-sectional study of the impact of pandemic COVID-19 on health worker behavior. An example algorithm was used to assess the impact of awareness of different aspects of SARS-CoV-2 on health professionals. The processing was done in the R software environment. Discussion. The present study had two objectives: to demonstrate a method for optimal matching of case-control pairs for specific data and to estimate the effect of certain factors on the outcome of interest. Using this method allows for maximum consistency between study and control groups on selected matching variables. This tool simplifies and automates data preparation for subsequent analysis when case and potential control data are presented in a single dataset. Thus, the example shows that with different values of the number of controls per case indicator the point and interval estimate of the odds ratio varies slightly and indicates a statistically insignificant association between the need for more information and the forced use of leave by the health professional due to various concerns about the impact of COVID-19.