Казахский Национальный Университет им. аль-Фараби, г. Алматы, Казахстан; 5 ФГАОУ ВО «Северо-Восточный федеральный университет им. М. К. Аммосова», г. Якутск В русскоязычной биомедицинской научной литературе одной из широко распространенных, но редко обсуждаемых проблем статистического анализа является проблема множественных сравнений. Она заключается в том, что увеличение числа проверяемых в процессе исследования статистических гипотез приводит к росту вероятности возникновения ошибок первого типа и ложных выводов о наличии различий там, где их нет, что связано с риском проведения необоснованных медицинских вмешательств, профилактических мероприятий и сопряжено с необоснованными расходами. В зарубежной научной литературе этой проблеме посвящено множество работ, как предлагающих новые подходы к решению проблемы множественных сравнений, так и описывающих применение уже общепризнанных методов. Однако в русскоязычной литературе такие работы встречаются редко. Целью данной статьи является восполнение имеющегося пробела путём представления способов решения проблемы множественных сравнений в медико-биологических исследованиях. Представлены методы решения проблемы на этапах планирования, статистического анализа и интерпретации результатов. Рассмотрены одношаговые методы, такие как поправки Бонферрони и Сидака (Шидака), и многошаговые: Холма-Бонферрони, Холма-Сидака (Шидака), Бенджамини-Хохберга и другие. Приведены примеры их использования, а также алгоритм их реализации как вручную, так и с помощью пакета статистических программ SPSS.

In this paper we present definition, classification and main characteristics of qualitative research as well as its advantages and disadvantages. We also compare the main features of qualitative and quantitative study designs. Examples are given to demonstrate practical steps in qualitative inquiry to demonstrate that qualitative research can sufficiently contribute to evidence generation in medicine and public health complementing quantitative research with answers on questions "why" and "how", which cannot be answered by quantitative research. The paper presents basic information on qualitative studies and does not substitute specialized literature.

In this article we present the main methodological principles of planning and performing ecological (correlation) studies as well as the principles of statistical analysis of data obtained in ecological studies. The theoretical background of this study design, its main advantages and disadvantages are presented. We also describe research questions that can be studied using ecological study design. Step by step instructions for statistical data analysis using free online calculators are presented. Correlation coefficients and their interpretation are described using example from real studies. Examples of ecological studies from the literature including examples from the Arkhangelsk region are presented.

We studied the prevalence and determinants of smoking and desire to stop smoking in a cross-sectional study among 1174 randomly selected adults aged 45+ years in Almaty, Kazakhstan. Associations between smoking and its correlates were studied by multivariable Poisson regression. Prevalence ratios (PR) with 95 % confidence intervals (CI) were calculated. Among current smokers we also studied factors associated with their desire to quit. Altogether, 40.7 % of men were current smokers and 63.1 % of them desired to stop smoking. The corresponding numbers for women were 10.0 % and 72.1 %. Male gender (PR = 4.14; 95 % CI: 3.18-5.40), Russian ethnicity (PR = 1.56; 95 % CI: 1.23-1.97), secondary or less education (PR = 1.37; 95 % CI: 1.09-1.73) and having satisfactory or worse psychological family climate (PR = 1.84; 95 % CI: 1.26-2.67) were positively associated with smoking. Men who reported poor (PR = 1.24; 95 % CI: 1.11-1.38) or satisfactory (PR = 1.17; 95 % CI: 1.08-1.27) health, had very good psychological family climate (PR = 1.20; 95 % CI: 1.03-1.41), and smoked 10-19 cigarettes a day (PR = 1.29; 95 % CI: 1.17-1.41) were more likely to report a desire to quit smoking. Number of daily smoked cigarettes, self-rated health, and psychological family climate were associated with the desire to quit.

An increase in life expectancy is one of the main strategic objectives declared by the Russian Federation. Thus, an understanding of how this objective can be achieved with available recourses in the most efficient way is warranted. We propose an automated method for estimating the contribution of cause-specific mortaLity to Life expectancy. To iLLustrate the proposed method, we used the data from primary mortality databases in the Krasnoyarsk region - one of the largest federal subjects of the Russian Federation - and the data on the average population of the of the region from 1999 to 2018 from the Federal state statistics office in Krasnoyarsk, Khakassia Republic and Tyva Republic. A computer program "DeathAnaLytics" has been developed by the authors for automated calculation of the contribution of cause-specific mortality to Life expectancy. The main idea behind is to calculate an integral indicator that takes into account both the contribution of deaths from various causes and the absolute number of these deaths. The paper presents the stages of calculation, interpretation and a practical example. The use of the methodology presented in the article aLLows to identify the causes of death that have the greatest impact contribution to reduction of Life expectancy, which in turn aLLows to identify targets for pubLic heaLth measures that wiLL most effectiveLy increase Life expectancy of the popuLation.