Background and research objectives:

Demographic and epidemiologic transitions have resulted in important shifts in population health patterns from frequently fatal, infectious, short-course diseases to chronic conditions associated with lifelong disabilities. These changes in population health patterns created the need for new indicators that were able to comprehensively quantify the impact of disease conditions on health. Summary measures of population health capture the overall impact of diseases by combining the effects of mortality and morbidity into a single indicator. Introduced in the first global burden of disease study, the disability-adjusted life year (DALY) has evolved into one of the most prominent summary measures. Despite the success of the burden of disease approach and the DALY, there is still a substantial need for adjustments to better meet the characteristics of infectious conditions and for sub-national estimates of disease burden. This thesis pursues these objectives by introducing a methodology to measure the disease burden of infectious conditions and presenting sub-national burden of disease estimates.

Methodological concepts:

To meet the first objective of the thesis, the incidence- and pathogen-based DALY approach was used to accommodate the specific characteristics of infectious diseases. Acute disease courses, as well as future short- and long-term sequelae were included through the use of outcome trees representing the natural history of infectious diseases. In addition to the calculation of the current disease burden for hepatitis B, influenza, measles and salmonellosis in Germany, projections of future burden were performed for influenza and hepatitis B in the Netherlands. To meet the second objective of the thesis, Standard Expected Years of Life Lost (SEYLL) as a standardized measure of disease burden due to premature mortality was used to estimate the years of healthy life lost at sub-national level for Hong Kong and North Rhine-Westphalia (NRW).

Results:

Introducing the incidence- and pathogen-based DALY approach, the highest disease burden in Germany (2005–2007) was estimated for infections with influenza virus (33,116 (95% UI: 29,504–36,849) DALYs/year), followed by salmonella spp. (19,115 (95% UI: 14,803–24,328) DALYs/year), hepatitis B virus (8,708 (95% UI: 7,335–10,163) DALYs/year) and measles virus (740 (95% UI: 413–1,066) DALYs/year). Infections with hepatitis B virus and salmonella spp. showed the highest burden related to sequelae, with 98% and 56.6% of the overall burden, respectively. Predicting the disease burden from 2000 to 2030 in the Netherlands showed increases of disease burden from 1,196 (95% UI: 1,003–1,328) DALYs to 1,343 (95% UI: 1,194–1,493) DALYs for hepatitis B and from 22,712 (95% UI 21,132–24,290) DALYs to 51,609 (48,212–55,198) DALYs for influenza virus. The greatest reductions in the future disease burden due to hepatitis B infections were calculated for the scenario simulating the uptake of vaccination in all age groups, with reductions in DALYs of 32%. For influenza the greatest reductions were simulated in the scenario with more effective age-targeted vaccination, with reductions in DALYs of 45%.

Using the SEYLL as a measure of premature mortality at sub-national level resulted in 1.75 million SEYLLs in North Rhine-Westphalia in 2005 and 524,707 SEYLLs in Hong Kong in 2010. Non-communicable diseases had the highest shares of SEYLLs, with 89.1% in NRW and 78.8% in Hong Kong. In comparison with prioritization by standard death counts, both studies highlighted self-inflicted injuries for males rising in priority from 13th to 8th rank in NRW and from 9th to 6th rank in Hong Kong, when using SEYLL as a measure of premature death. Scenario analysis identified that using different assumptions about social value choices diminished the SEYLLs by up to 51.6% (scenario I) and had selective impact on the different disease groupings. Changing the standard life expectancy values to the ones observed in Hong Kong, the disease burden increased by 10.8%.

Conclusions:

The results of the thesis showed that going beyond the one-size-fits-all solution as used in the global burden of disease study and drilling down estimates to a sub-national level can provide sound additional information on population health patterns. The use of the incidence- and pathogen-based DALY approach highlighted sequelae of infections with hepatitis B and salmonella spp. in particular as an important component of the overall disease burden. This is of major importance when prevention measures aim at avoiding the initial infection. The results of the thesis also highlighted that the approach can also be used to predict future disease burden by including the impact of population and disease dynamics. The results presented in the thesis further fill existing data gaps for sub-national burden of disease estimates and introduce the SEYLL as a suitable measure for such assessments. The studies also presented the advantage of using the SEYLL measure over standard measures of mortality and highlighted the impact of social value choices. Overall the thesis showed that the DALY and its components can serve as powerful indicators of population health, but there is a need to adjust the measures for specific settings. Further, the thesis also emphasized the critical need for transparency when using the burden of disease approach and population health measures. Finally, it also accentuates the importance of increasing the quality of epidemiological data, because in the end summary measures are only as good as the epidemiological input data.