Mortality rates have significant differences between regions of the Russian Federation. Standardized mortality rates (SMR) are influenced by many factors, including diagnosis criteria and causes of death.
Analysis of coefficient of variation of SMR from malignant neoplasms compared with other causes of death in the subjects of Russian Federation.
The study formed 3 comparison groups: malignant neoplasms, other causes and causes of death, presented in ICD-10 to indicate conditions that do not have clear diagnostic criteria and rules for its use, as well as unspecified and not recommended for use in mortality statistics. To calculate SMR out of 79 reasons selected from the «Brief nomenclature of causes of death of Russian Statistic Bureau (RosStat)», the European standard was used. The accuracy of differences between the groups was evaluated using one-way analysis of variance with pairwise comparison and Bonferroni correction.
In the group of malignant neoplasms, there were significant differences in the coefficients of SMR variability between the subgroup of neoplasms with an exact indication of localization and the subgroup of malignant neoplasms of other and inaccurately defined localizations. However, even the causes of death from malignant neoplasms that do not have sufficiently clear application criteria have a significantly less pronounced coefficient of SMR variability compared to the other two groups. The causes of death, starting in the RosStat nomenclature with the words «other diseases», and the causes associated with the diseases presented in ICD-10 to indicate conditions that do not have clear diagnostic criteria and rules of its use, as well as unspecified and not recommended for use, had practically the same high level of SMR variability coefficient. A certain contribution to the differences in regional SMR indicators from specific causes can be affected by the inconsistency of the criteria for diagnosing diseases and choosing the initial cause of death.
To obtain the correct indicators of mortality from individual causes / groups of causes, we need unified generally accepted definitions and rules for the application of conditions presented in ICD-10, which do not have clear diagnostic criteria, as well as unspecified; and as well as rejection of statistics not recommended for use because of the initial causes of mortality.