Mapping tuberculosis fatalities by region and age group in South Korea: A dataset for targeted health policy optimization

TL;DR

It is demonstrated that incorporating age structure can give rise to distinct optimized hospital allocation patterns, even when the total number of minimized fatalities is similar, revealing trade-offs between efficiency and demographic targeting.

Abstract

In South Korea, age-disaggregated tuberculosis (TB) data at the district level are not publicly available due to privacy constraints, limiting fine-scale analyses of healthcare accessibility. To address this limitation, we present a high-resolution, district-level dataset on tuberculosis (TB) fatalities and hospital accessibility in South Korea, covering the years 2014 to 2022 across 228 districts. The dataset is constructed using a reconstruction method that infers age-disaggregated TB cases and fatalities at the district level by integrating province-level age-specific statistics with district-level spatial and demographic data, enabling analyses that account for both spatial heterogeneity and age structure. Building on an existing hospital allocation framework, we extend the objective function to an age-weighted formulation and apply it to the reconstructed dataset to minimize TB fatalities under different age-weighting schemes. We demonstrate that incorporating age structure can give rise to distinct optimized hospital allocation patterns, even when the total number of minimized fatalities is similar, revealing trade-offs between efficiency and demographic targeting. In addition, the dataset supports temporal analyses of TB burden, hospital availability, and demographic variation over time, and provides a testbed for spatial epidemiology and optimization studies that require high-resolution demographic and healthcare data.

Figures (9)

• Figure 1: (a) Quartile plots of the fatality rate $\phi_t$ by age group $t$ across 16 provinces in 2022 (excluding Sejong-si due to the absence of reported fatalities). (b) Average age of the general population across the districts. Metropolitan districts tend to have younger populations, while non-metropolitan districts are characterized by older populations. (c) The distribution of average age displays a heterogeneous pattern in South Korea, indicating an overconcentration of younger individuals in metropolitan areas.
• Figure 2: Temporal patterns of the raw data for (a) the number $N$ of newly reported TB cases, (b) the number $D$ of TB-related deaths, and (c) the number $H$ of hospitals. The number of new TB cases and deaths is showing a declining trend. The number of secondary care hospitals is slightly increasing in the Gyeonggi-do region. The 16 provinces are represented by distinct symbols.
• Figure 3: Regional and age-grouped tendency of (a) TB patients fraction $n$, (b) TB deaths fraction $d$, and, as a comparison, (c) the general population demographic fraction $g$ for the respective province in 2022. Both the proportion of new TB cases and TB-related deaths tend to increase with age. The 16 provinces are represented by distinct symbols.
• Figure 4: (a) District-level values of the difference ${(\eta^{\rm (opt)}_{\rm age} - \eta^{\rm (opt)}_{\times})}/{\eta}$ across South Korea. The inset shows a magnified view of Seoul. Uncolored districts indicate no data satisfying the criteria. (b-c) Comparison of the difference ${(\eta^{\rm (opt)}_{\rm age} - \eta^{\rm (opt)}_{\times})}/{\eta}$ for $\rho_{s,5}$ (b) and $\rho_{s,5}/\tilde{\eta}_{s,5}$ (c). No distinct pattern observed along $\rho_{s, 5}$, while a steady, roughly linear increase is seen along $\rho_{s,5}/\tilde{\eta}_{s,5}$ despite the log-scaled x-axis. (d) The minimum fatalities $E^{\rm (min)}_{\rm{fatalities}}$ as a function of the age weighting parameter $a$ in 2022 is shown representatively, with $a_{\rm min}=-1.4$ and $a_{\rm max}=0.3$. The purple hexagon and green circle represent the cases of $a=a_{\rm min}$ and $a=0$, respectively. The inset figure shows the minimized number of fatalities for each age group $t$ as a function of $a$. (e) The yearly trend of the deviation of $E^{\rm (min)}_{\rm{fatalities}}$ from the original deaths $D$ when $a=a_{\rm min}$ and $a=0$. A bigger value indicates a greater reduction in fatalities compared to the original deaths $D$. Across all analyzed years, the cases with $a=a_{\rm min}$ achieve a better reduction in fatalities than those of $a=0$.
• Figure 5: (a) Relationship between relative changes in optimal hospital allocation $\Delta \eta^{\rm (opt)}_{s}$ versus fatality $\Delta \epsilon^{\rm (opt)}_{s}$ across districts under the $a\simeq a_\mathrm{min}$ case compared to the baseline case ($a=0$). (b) A distribution of fatality change $\Delta \epsilon^{\rm (opt)}_{s}$ for age group $t$. $\Delta \epsilon^{\rm (opt)}_{s}$ for $t=5$ (ages 80 and above) concentrated in the negative range, unlike other age groups.