Access to radiotherapy in improving gastric cancer care quality and equality

Data sources

The Global Burden of Disease (GBD) study assessed the disease burden of 369 different diseases and 87 risk factors in 204 countries, 21 regions, and seven super-regions, aiming to provide a comprehensive and comparable global health assessment^12,13. GBD methodology has been described previously. Briefly, the GBD estimation process is based on identifying multiple relevant data sources for each disease including censuses, household surveys, civil registration and vital statistics, disease registries, health service use, air pollution monitors, satellite imaging, disease notifications, and other sources. Cause-specific death rates and cause fractions were calculated using the Cause of Death Ensemble model and spatiotemporal Gaussian process regression. Cause-specific deaths were adjusted to match the total all-cause deaths calculated as part of the GBD population, fertility, and mortality estimates. Deaths were multiplied by standard life expectancy at each age to calculate years of life lost (YLLs). A Bayesian meta-regression modeling tool, DisMod-MR 2.1, was used to ensure consistency between incidence, prevalence, remission, excess mortality, and cause-specific mortality for most causes. Prevalence estimates were multiplied by disability weights for mutually exclusive sequelae of diseases and injuries to calculate years lived with disability (YLDs). In this study, we searched the GBD 2019 database and collected data on gastric cancer incidence, prevalence, mortality, YLLs, YLDs, and disability-adjusted life-years (DALYs) from 1990 to 2019. Gastric cancer was defined according to the International Classification of Diseases Tenth Revision (ICD-10); codes C16–C16.9, D00.2, D13.1, and D37.1 were recorded as new cases of gastric cancer in the GBD dataset¹⁴.

Country-level health technologies and health workforce data was retrieved from WHO Global Health Observatory (GHO, The GHO data repository provided WHO’s statistics on priority health topics including mortality and burden of diseases, noncommunicable diseases and risk factors, health systems, environmental health, violence and injuries and others in 194 Member States. The total density of radiotherapy units per million population and number of medical doctors per 10,000 population were used to represent the health technologies and health workforce, separately. In 2010, WHO launched a country survey on medical devices that allowed to identify the status of high cost medical devices in the Member States, including radiotherapy equipment, both linear accelerators and Cobalt-60. Similar survey was conducted in 2020–2021 update by collecting information directly from country focal points from ministries of health.

Quality of Care Index

First, four secondary indicators were constructed based on the age-standardized rate retrieved from the GBD 2019 dataset, namely¹ the ratio of YLLs to YLDs², the ratio of DALYs to prevalence³, mortality-to-incidence ratio, and⁴ prevalence-to-incidence ratio.

Ratio of YLLs and YLDs = YLLs/YLDs¹

Ratio of DALYs to prevalence = DALYs/Prevalence²

Mortality-to-incidence ratio = Mortality/Incidence³

Prevalence-to-incidence ratio = Prevalence/Incidence⁴

Second, the quality of care index (QCI) was constructed using principal component analysis (PCA) techniques based on the above four secondary indicators. PCA involves mathematical multivariate analysis to extract linear combinations as orthogonal components of specific indicators. Then the component that best describes the variance and variability in the data is denoted the QCI and allocated a score of 0–100. Higher scores on the QCI indicate a high quality of gastric cancer care^15,16.

In this study, we calculated the QCI and analyzed the changing trends in the QCI from 1990 to 2019 for males and females. The estimated annual percentage change (EAPC) and 95% confidence intervals (CI) were calculated using a linear regression model¹⁷: y = α + βx + ε, where y = ln(QCI), x = calendar year, and ε = error term. The value of EAPC equals 100 × (exp(β) − 1) and its 95% CI is attainable in the regression model.

A gender difference ratio (GDR) was calculated to explore the disparity between men and women¹⁸. The GDR was defined as the ratio of the QCI score in women divided by that in men, with a GDR > 1 indicating better gastric cancer care in women compared with men.

The socio-demographic index (SDI) is a comprehensive indicator based on the education level, per capita income, and total fertility rate of individuals under the age of 25 years, which measures the overall development scale of a country. The SDI indicator was also extracted from the GBD 2019 dataset ( In 2019, different countries were classified into five development levels according to the SDI, namely: low (<0.46), low-middle (0.46–0.60), middle (0.61–0.69), high-middle (0.70–0.81), and high (>0.81)¹⁹.

Statistical analysis

Association analysis was conducted to identify the role of radiotherapy units in improving quality of gastric cancer care. Detailed information of the variable definition and coding forms were displayed in Supplementary Table 1. Univariable regression model was applied with the log-transformed gender-specific QCI as the outcome variable and the total density of radiotherapy units per million population was used as independent variable, separately. Multivariable models added the universal health coverage and social development levels defined using SDI as adjusting variables¹⁹. Model 3 further added health workforce (number of medical doctors per 100,000 population) and infrastructure (hospital beds per 10,000 population) as covariates. The effect size was estimated with 95% confidence interval (CI). Then parallel association analysis was conducted with the absolute value of log-transformed GDR was the outcome variable to identify the role of health technologies and health workforce in reducing gender inequality.

All statistical analyses for this study were created using R v4.1.3 software ( All tests were two-sided, and P values < 0.05 were considered statistically significant.

Ethics approval and consent to participate

Patient consent was not required as we only utilized de-identified, publicly available datasets in our analysis, Likewise, we did not seek ethics approval from any ethics committee or authoritative body for the use of this data in the study objectives as it was not required.

Inclusion & ethics statement

All collaborators of this study have fulfilled the criteria for authorship required by Nature Portfolio journals have been included as authors, as their participation was essential for the design and implementation of the study. Roles and responsibilities were agreed among collaborators ahead of the research. This research was not severely restricted or prohibited in the setting of the researchers, and does not result in stigmatization, incrimination, discrimination or personal risk to participants. Local and regional research relevant to our study was taken into account in citations.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.