Summary_Danilova (1136935), страница 2
Текст из файла (страница 2)
The latter is necessary to assurereliable and correct results. These considerations motivated the direction of our study tofocus on the regional peculiarities of mortality in Russia.6Meslé F., Vallin J. Mortality in Europe: The divergence between East and West // Population (English edition). 2002.
№57 (1). P.157–197.The objective of the study: use statistics on the regional mortality by causes ofdeath to evaluate the inequality of the progress of regions in the epidemiologicaltransition and to estimate the components of persisting large regional disparity in lifeexpectancy.To reach this goal the following tasks had to be completed:1.Evaluate the quality and comparability of the statistical data on mortalitygathered by the regions (all-cause and cause-specific) in Russia at the regional level;2.Track how changes in life expectancy at the regional level followed thechanges in life expectancy at the national level. Determine which regions experiencedthe changes in life expectancy significantly higher or lower of those observed for Russiaas a whole;3.Analyze the changes in interregional life expectancy inequality and itsinterrelation with the changes in life expectancy at the national level in Russia;4.Estimate the contribution of single age groups and causes of death tointerregional life expectancy inequality in Russia and to the changes of the inequalityover time;5.Determine spatial characteristics of mortality differentiation for single agegroups and causes of death and estimate the interregional inequality.Theoretical background of the study-The concept of the epidemiologic transition of A.Omran;-The concept of health transition of J.Frenk et al., who rethought anddeveloped the concept of epidemiologic transition;-Theoretical views on the processes of divergence and convergence ofdemographic processes during the demographic transition and health transition inparticular.Data and methods1.Data provided by the Federal State Statistics Service of the RussianFederation (the Rosstat).
We used population estimates and death counts for the period1989-2015 by age (5-year age groups), sex, regions, and causes of death. The data forRussia as a whole is available for the period 1965-2015;2.Data from the Russian Fertility and Mortality Database of the Center forDemographic Research of the New Economic School. Mortality rates and populationestimates by age (1-year age groups), sex, and region for the period 1989-2015 wereused, as were regional mortality data for the around-census periods 1969-1970 and1979-1980;3.To compare mortality levels across countries, the data from the HumanMortality Database (a joint project of Max-Planck Institute for Demographic Researchand University of Berkley, California) were used;4.To decompose the gap in life expectancy between the groups of Russianregions and Poland, the data of the Human Cause-of-Death Database (a joint project ofMax-Planck Institute for Demographic Research and the French Institute forDemographic Studies of INED) were used;5.The data on the causes of death in the departments of France wereprovided by the French Institute for Demographic Studies (INED).6.To plot the data on the maps, an open dataset of spatial objects wasdownloaded from the GIS-Lab.The following methods were used in the study:-The methods of demographic analysis of mortality (including the methodsof standardization of mortality rates, constructing life tables)-The method of step-wise replacement decomposition of life expectancy byage and causes of death7;-Statistical methods for analyzing distributions;-The methods of spatial analysis (mapping, estimating indices of spatialautocorrelation).The novelty of this study1.A complex analysis of the quality and comparability of the mortality datagathered at the regional level in the Russian Federation was performed.
The leadingproblems with the data were systematized: lowered mortality rates at very old ages,incomparability of regional approaches to the selecting of the underlying cause of death,the influence of “Senility” on the mortality rates from the other causes of death at agesover 80. In the study, a list of causes of death was proposed which made it possible toperform regional mortality analysis by causes under the circumstances of differentapproaches to certification and coding of the underlying cause.2.Regional trajectories of the changes in the life expectancy from the late1960s to 2015 were traced. We defined the regions where the trends of the changes inlife expectancy were unique and distinguishing from those observed for Russia as awhole.
We identified the diverging contribution of the cities of Moscow and SaintPetersburg to the change of inter-regional inequality in life expectancy.3.We analyze the changes of the inter-regional inequality in life expectancyover time and identify the components (by age and causes of death) of this inequalityfrom the late 1960s through the present day. It is shown that today the decrease inmortality from external causes at young and middle ages and the circulatory diseases atmiddle ages is contributing the most to the decrease of inter-regional inequality in life7Andreev E.
M., Shkolnikov V. M. Begun A. Z. Algorithm for decomposition of differences between aggregate demographicmeasures and its application to life expectancies, healthy life expectancies, parity-progression ratios and total fertilityrates // Demographic Research. 2002. № 7(14). P. 499-522.expectancy. This is a contrast to the decrease of mortality from circulatory diseases atolder ages, which is leading to the growing diversity across regions.4.Using the methods of spatial analysis, we estimated spatial dependence inhow mortality indicators (all-cause and cause-specific) are spread across the territory ofRussia and how this dependence tends to change over time.
We show that there issignificant spatial autocorrelation in mortality within Russia: regional with lowermortality levels tend to be found adjacent to the regions in which mortality levels arealso lower.Results of the studyQuality and comparability of the mortality data gathered in Russia at thesubnational levelThe reliability and accuracy of a cause-specific mortality analysis relies on thequality of the data. The checks for the data quality performed in the current studyshowed that there are systematic problems with the quality and comparability ofmortality data gathered at the regional level in Russia. Insufficient quality of the dataposed a severe limitation for the analysis.Demographic analysis at the subnational level is more complicated than at thenational level, with possible biases caused by migration across territories.
In Russia, adeath is statistically assigned to that region where the Civil registration office (ZAGS)registered it8. According to the Federal law “On the registering acts of civil status”,death registration can be performed “according to the most recent address of thedeceased, place of death, place of body discovering…” and some other stipulationslisted9. Consequently, the death counts assigned to the particular region in the statistics8Surinov A.E. The Rosstat data as input information for actuarial calculations in insurance. Completeness and reliability.The presentation on the XIth International Insurance Conference. Moscow, 5-6 June, 2013.9Federal Law of 29 December, 2017 N 472-FZ «On introducing changes to the Federal Law of 15 November, 1997 N 143-FZ«On the Acts of Civil Status» (with changes and additions)may include deaths of residents of other regions, in those instances where their deathswere registered in that particular region.
By the same token, the death counts may notinclude the deaths of all of the region’s residents, should these deaths have beenregistered elsewhere. It is obvious that such a system can cause some numeratordenominator bias in the mortality rates estimation. This problem can be especiallycrucial for the federal cities and surrounding regions.Death registration, either by the place of residence or by the place of death, mayalso result in a so-called statistically “immortal population”, which happens when thedeath of a region’s resident was registered outside the region, but he/she is still beingcounted in the population.
At young ages the death probability is rather low, i.e., thelow number of deaths corresponds to a large number of population, and these problemscannot bias the demographic indicators significantly. However, if the proportion of“immortal population” accumulates while moving to the older ages, the mortality ratesmay be substantially biased.While comparing regional mortality levels for cause-specific mortality, it isessential to take into account the comparability of regional approaches to selecting theunderlying cause of death. In the current study, we rely solely on the secondary data,i.e., on data processed and published by the statistical offices.
Consequently, noadditional data other than death counts by regions, sex, age, and causes of death wereavailable. The conclusions on the comparability of regional approaches to selecting theunderlying cause could be made only using the indirect statistical methods. For 42groups of causes of death, we calculated the coefficient of variation and the ratiobetween the 3rd maximum and the 3rd minimum for the share of these causes in thecause-specific mortality structures of regions for the period 2013-2015.
For comparison,we calculated the similar indicators for the departments of France.The analysis showed that mortality levels from neoplasms have the highestconsistency across all Russian regions. High cross-regional consistency was also foundfor the large groups included in the ICD-chapter of circulatory diseases (ischemic heartdisease and cerebrovascular diseases), diseases of the digestive system, and transportaccidents.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
