Improving the quality assessment of mortality data in Serbia by determining the vital statistics performance index

[M. R. Anđelković-Apostolović]

142 lista

Bibliografija: listovi 118-132. Beleška o autoru: list 142.

Medical statistics and informatics

The traditional approach to mortality quality assessment was based on evaluating the completeness of vital statistics and the proportion of R diagnoses, while the new principle determines the Vital Statistics Performance Index for Quality-VSPI(Q). The goals of the doctoral dissertation are to improve the assessment of the quality of data on death outcomes by determining the VSPI(Q) index in Serbia from 2005 to 2019 and assessing the level of knowledge and attitudes of physicians when filling out the death certificate (DC). In the first part of the research, the ANACONDA software package was used to assess the quality of death data, which analyses the accuracy and completeness of mortality data with demographic and epidemiological global estimates. In the second part of the research, a modified Mid-America Heart Institute (MAHI) Death Certificate Scoring System determined the assessment of physicians' knowledge. The mortality data in Serbia was assessed as medium quality for the entire fifteenyear period (VSPI(Q) score of 67.2%). A constant increase in the quality of mortality data was based on the increase in the VSPI(Q) score from 55.6 (medium quality) to 70.2 (high quality) in the follow-up period. It was found that 24.2% belonged to garbage codes (GC), while 18.5% belonged to insufficiently specified causes of death. Death due to external factors is represented by 0.8%, below the global average of 5-10%. According to the GC typology, the most common error belongs to category 3 (24.7%), representing the intermediate cause of death. The most common ill-defined diagnosis for both sexes is nonspecific cardiomyopathy (I42.9). As a result of the ANACONDA analysis, 9.2% of GC in females and 9.1% of GC in males were successfully redistributed. A strong positive correlation was found between the VSPI(Q) score and sociodemographic indices for Serbia. The physician's knowledge of coding DC is suboptimal for all three cases. The optimal level of knowledge was achieved by 51.4% of physicians for coding the natural death, 42.3% for coding the violent death, and 19.5% for coding the undetermined death. Predictors of better knowledge are education, frequency of filling out DCs, and the number of completed DCs. The results indicate an increase in the quality of mortality data in Serbia and a large percentage of GC in the fifteen years, reflected in physicians' suboptimal knowledge. For this purpose, it is necessary to adapt education and focus on the critical aspects of each cause of death.

mortalitetna statistika, potvrda o smrti, VSPI(Q), garbage kodovi, Srbija

mortality statistics, death certificate, VSPI(Q), garbage codes, Serbia

616‑036.8:311(497.11)(043.3)

B110

Tekst