Dependence between mortality in regions and prevalence of active SARS-COV2 carriers and resources available to public healthcare organizations

Stepanov, V.S.

Автор: Stepanov V.S.
Журнал: Анализ риска здоровью
Номер: 4
Год издания: 2020
Издательство: Федеральное бюджетное учреждение науки Федеральный научный центр медико-профилактических технологий управления рисками здоровью населения
Местоположение издательства: Пермь
Первая страница: 12
Последняя страница: 22
Номер статьи: 2
DOI: 10.21668/health.risk/2020.4.02.eng
Аннотация: The paper dwells on certain mathematical models showing how epidemics develop, namely, logistic ones, SIR-model, and some others. There is also made a small review of articles that focus on such models showing dynamics of incidence with COVID-19 infection. These models are often successfully applied for data collected in a whole country but on a regional level there are difficulties due to peculiarities of mortality accounting in Russia. In this case regression models can be useful with their advantage at the initial stage in an epidemic process. They also include exogenous variables that influence mortality, for example, a number of doctors and nurses per 100 thousand population, how well hospitals are equipped with ALV devices, and a number of available beds in them. Our research goal was to build up a linear regression model that could be used as a basis for estimating regional mortality caused by COVID-19 as well as for more efficient distribution of the resources mentioned above.The model is based on a set of indicators of resource provision as well data with «active cases». Preliminary, three variables that showed data on resources available to communicable diseases departments in hospitals were transformed into a new single one. Then the model was tested on a training sample containing an endogenous variable on mortality and four factor ones including prevalence of active virus carriers. Regions were included into the training data with different lags; they were included for such days when death cases were registered rarely. Then the estimated model was applied with other values of exogenous variables. It turned out to be quite efficient in estimating COVID-induced mortality for regions from the training sample as well as for several others (for certain time intervals). As a result, we built a regression model and estimated its precision; the model showed a relation between mortality in a region and prevalence of active SARS-CoV-2 carriers and availability of resources to hospitals in it. It can be useful when these resources will distribute. It can also be used to build SIRD, SEIR, and SEIRF models at a regional level when choosing parameters in them related to mortality. A methodology itself that can be similarly applied for other epidemic processes also deserves certain attention. Key words: regression model, mortality estimation, COVID-19, coronavirus infection, logistic equation, SEIR, SIR, ALV
Добавил в систему: Степанов Владимир Сергеевич

	ИСТИНА	Войти в систему Регистрация
	Интеллектуальная Система Тематического Исследования НАукометрических данных
	Главная Поиск Статистика О проекте Помощь

ИСТИНА

Интеллектуальная Система Тематического Исследования НАукометрических данных

Dependence between mortality in regions and prevalence of active SARS-COV2 carriers and resources available to public healthcare organizationsстатья Исследовательская статья Электронная публикация

Прикрепленные файлы