Publication Characteristics and Reporting Quality of GBD‐Derived Burden of Disease Studies: A Cross‐Sectional Analysis of Lung Cancer

ABSTRACT

Aim

Using lung cancer as a model disease, we systematically analyzed the publication characteristics and reporting quality of burden of disease (BoD) studies derived from the Global Burden of Disease (GBD) Database, to provide evidence for standardizing high‐quality BoD research.

Methods

A cross‐sectional study was conducted. We included GBD‐derived lung cancer BoD studies and evaluated reporting quality using 17 core items of the STROBOD Statement. Univariable and multivariable linear regression models were used to explore factors associated with reporting completeness.

Results

A total of 32 studies showed that the number of publications increased significantly since 2025, with repetitive research topics. Evaluation of the 17 core items in the STROBOD Statement revealed that 88.2% (15/17) of the items had reporting deficiencies. The number of reported core items was 11.19 ± 1.67, and none of the included studies reported all of them. The reporting quality of English studies was significantly higher than that of Chinese studies (p < 0.01).

Conclusions

GBD‐derived lung cancer BoD studies present severe topic homogeneity and suboptimal reporting quality. Developing tailored reporting guidelines is urgently needed to improve the transparency and reliability of BoD research.

1. Introduction

The latest global cancer burden study indicates that lung cancer ranks first among malignant tumors in terms of incidence and mortality [1], making it the most prevalent cancer worldwide. It is estimated that there were approximately 4.8247 million cases and 2.5742 million deaths in China in 2022 [2], imposing a substantial socioeconomic burden [3]. Burden of disease (BoD) studies on lung cancer help formulate prevention strategies and optimize the allocation of medical resources, which is crucial for reducing the social disease burden.

The Global Burden of Disease (GBD) Database is a comprehensive research system for BoD studies, covering 204 countries and regions, including 463 health outcomes and risk factors [4]. The GBD Database has been widely applied to BoD studies, including cancer [5], cardiovascular diseases [6], and infectious diseases [7], serving as a critical data resource for global public health decision‐making and academic research. However, the number of publications derived from the GBD Database has increased rapidly in recent years. Studies with repetitive topics and uneven quality impair the reliability and application of their results.

The STROBOD Statement is a framework for standardizing BoD studies' reporting quality [8]. Taking lung cancer BoD studies derived from the GBD Database as an example, this study conducts a cross‐sectional analysis to clarify the publication characteristics, quantify the reporting quality, and explore the impact of publication factors on reporting quality. This work is expected to provide a reference for advancing the standardized development of BoD studies.

2. Methods

2.1. Study Design

We conducted a cross‐sectional analysis of the lung cancer BoD studies derived from the GBD database. This study was conducted and reported based on the STROBE Statement.

2.2. Search Strategy

We searched PubMed, Embase, Web of Science, Cochrane Library, CBM, CNKI, and WanFang Databases, and the search time ranges from database establishment to September 29, 2025. The search strategy included both themes and free terms and is adapted to the characteristics of each database. The search terms included “Global Burden of Disease” and “pulmonary carcinoma”. Search strategies are provided in File S1.

2.3. Eligibility Criteria

We included lung cancer BoD Studies derived from the GBD Database. The exclusion criteria were as follows: (1) duplicate publications; (2) BoD studies not specifically focusing on lung cancer; (3) BoD studies using additional data sources other than the GBD Database; and (4) literature types including conference abstracts, review articles, letters, and commentaries.

2.4. Selection and Data Extraction

Two reviewers independently conducted studies screening and data extraction. Disagreements were resolved through consultation with a third reviewer. The titles and abstracts of the identified records were screened, followed by a full‐text review of potential studies to determine their eligibility. Data extraction included: (1) publication characteristics (year, authors, language); (2) study characteristics (population age, gender, region, GBD version, prediction model); (3) outcome indicators (age‐standardized incidence rate, ASIR; age‐standardized prevalence rate, ASPR; age‐standardized mortality rate, ASMR; age‐standardized disability‐adjusted life years, AS‐DALYs); and (4) STROBOD reporting quality results.

2.5. Reporting Quality Assessment

The reporting quality of the included studies was assessed by two independent reviewers. During quality assessment, Items 12–21 and 23 of the STROBOD Statement [8, 9] were excluded as inapplicable. The GBD database provides a pre‐standardized modeling framework; researchers use validated estimates directly without designing disease models, reporting raw primary data, or performing independent scenario analyses.

A total of 17 items were included for assessment: Items 1–11, Item 22, and Items 24–28. The item was evaluated by “Reported” (if all required information has been fully reported), “Partially reported” (if part of the required information is missing), “Not reported” (if all required information is missing), or “Not applicable” (if the item is unsuitable for the study).

2.6. Statistical Analysis

Descriptive analysis was performed for frequency and constituent analysis using Excel 2019. The results of studies with repetitive topics (studies with consistent database version, age, gender, and region of the study population) were compared. Sankey diagrams were plotted to analyze the risk factors of the included studies. Univariable and multivariable linear regression models were applied to explore factors associated with the number of fully reported items, including publication year, language, number of authors, and statisticians' participation. Regression coefficients (β) with 95% confidence intervals (CI) were calculated and presented as effect sizes. Chi‐square tests were performed to analyze the impact of each factor on the reporting quality of each item. False discovery rate (FDR) correction was applied to account for multiple testing.

3. Results

3.1. Search Results

We identified 1399 records and removed 451 duplicate publications. A total of 329 irrelevant publications were excluded after title and abstract screening. After excluding 587 studies failing to meet the inclusion criteria, 32 publications were included. The results of the literature screening are presented in Figure 1.

FIGURE 1.

Flow chart of study screening.

3.2. Study Characteristics

Among the 32 included studies [10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41], only 1 was published in 2016, none from 2017 to 2021, and 3–4 publications annually from 2022 to 2024, and 21 in 2025. A total of 90.6% (29/32) of the first authors were Chinese, while 3 were from Iran. The population was mainly the general population (22/32, 68.8%), with only 10 studies focusing on specific age/gender groups; no studies targeted children/adolescents. Geographically, 43.8% (14/32) focused on global data, among which 6 studies compared the Chinese population with the global population. Another 15.6% (5/32) of studies focused on regions or alliances such as the BRICS Economies (including Brazil, Russia, India, China, and South Africa). Additionally, 28.1% (9/32) targeted the Chinese population, among which 1 study compared the differences between China and Australia. A further 9.4% (3/32) of studies refined to provincial‐level data, while 1 study focused on the BoD in Iran. The main outcomes included incidence, prevalence, mortality, and disability‐adjusted life years (DALYs) rate, with 33.3% (11/32) including all of them. Some studies focused on years of life lost (YLL) and years lived with disability (YLD). Among the 32 included studies, only 50.0% (16/32) used predictive models to analyze future BoD trends. Among them, 62.5% (10/16) used the Bayesian Age‐Period‐Cohort (BAPC) model, 18.8% used the Autoregressive Integrated Moving Average (ARIMA) model, 2 studies used both, and 1 study used the Nordpred model (Table 1). Only 3 study reports followed the GATHER Statement [42]

TABLE 1.

Characteristics of the studies.

Study	Age/gender	Region	Database	Outcomes	Prediction model
Mazidimoradi 2025 [10]	≥55/all	Global	GBD2021	①②③	/
Zhang 2025 [11]	≥15/all	China	GBD2021	①②③	ARIMA
Li 2025 [12]	All/female	Global, China	GBD2021	①②③④	ARIMA
Wang 2023 [13]	All/all	Global	GBD2019	①②③	/
Zhao 2025 [14]	All/all	China, Australia	GBD2019	①②③	BAPC
Fu 2025 [15]	≤54/all	Global	GBD2021	①②③④	/
Liu 2025 [16]	All/all	China	GBD2021	①②③④⑤⑥	BAPC
Sun 2023 [17]	All/all	China	GBD2019	①②④	/
Xing 2025 [18]	≥70/all	Global	GBD2021	①②③	/
He 2025 [19]	All/all	China	GBD2021	①	ARIMA
Long 2023 [20]	≥5/all	G20, China	GBD2019	①②③	BAPC
Wu 2025 [21]	≥15/all	Asia Pacific, Africa	GBD2021	①②③④	BAPC + ARIMA
Wang 2024 [22]	≥15/all	Zhejiang (China)	GBD2019	③⑤⑥	/
Bai 2024 [23]	All/all	BRICS	GBD2019	②	/
Khanmohammadi 2022 [24]	All/all	North Africa, Middle East	GBD2019	①②③④	BAPC
Li 2025 [25]	15–49/all	China	GBD2021	①②	/
Rao 2025 [26]	All/all	India	GBD2021	②③	/
Xu 2025 [27]	All/all	China	GBD2021	①②③④	BAPC
Li 2025 [28]	≥25/all	Global	GBD2021	①②③④	/
Xing 2024 [29]	All/all	Global, China	GBD2019	②③⑤⑥	BAPC
Fan 2025 [30]	All/all	Global, China	GBD2021	①②③④	/
Wu 2025 [31]	All/all	Global, China	GBD2021	①②	Nordpred
Zhang 2025 [32]	All/all	Global	GBD2021	①②③④	BAPC + ARIMA
Chen 2025 [33]	All/all	Global, China	GBD2021	①②③④	/
Fang 2023 [34]	All/all	China	GBD2019	①②③	BAPC
Jiang 2022 [35]	All/all	Jiangsu (China)	GBD2019	①②③⑤⑥	BAPC
Liu 2016 [36]	All/all	China	GBD2013	②③⑤⑥	/
Wang 2024 [37]	All/all	BRICS	GBD2021	①②③④	BAPC
Han 2025 [38]	15–45/all	Global	GBD2021	②③④	/
Li 2025 [39]	≥15/all	Global	GBD2021	①②③	/
Shokri Varniab 2022 [40]	All/all	Iran	GBD2019	①②③④	/
Tian 2025 [41]	15–39/all	Global, China	GBD2021	①②③	BAPC

① ASIR; ② ASMR; ③ AS‐DALYs; ④ ASPR; ⑤ YLL; ⑥ YLD. “/” means no prediction model was used.

Abbreviations: ARIMA, Autoregressive Integrated Moving Average model; BAPC, Bayesian Age‐Period‐Cohort model.

3.3. Research Topics and Predictive Results

Studies with repetitive topics were observed among the 32 included studies. Specifically, 2 studies analyzed the burden of lung cancer in the Chinese population based on GBD2019 Database [17, 34]; 3 studies analyzed lung cancer burden between China and global population using GBD2021 Database [30, 31, 33]; and another 3 studies focused on the analysis of lung cancer burden in the Chinese population relying on GBD2021 Database [16, 19, 27].

To further explore the consistency of predictive results, this study selected three groups of publications targeting the same population, using the same database, and predicting the BoD trends at the same time for comparison. All six studies adopted the BAPC model for prediction, with a consistent predicted trend across different publications (Table 2). However, both Xu 2025 and Wang 2024 used data from the GBD2021 Database and applied the BAPC model to predict the lung cancer burden in China in 2035. Notably, their predicted results for the ASPR showed discrepancies.

TABLE 2.

Predictive results.

Number	Study	Age/gender/region	Time	Model	ASIR	ASPR	ASMR	AS‐DALYs
1	Wang 2024 [37]	All/all/China	Data: 1990–2021; projection: 2035	BAPC	50.96	63.61	36.53	916.28
	Xu 2025 [27]	All/all/China	Data: 1990–2021; projection: 2035	BAPC	50.32	73.49	36.53	916.38
2	Fang 2023 [34]	All/all/China	Data: 1990–2019; projection: 2030	BAPC	Male: EAPC = 0.101 Female: EAPC = 2.236	/	Male: EAPC = –0.537 Female: EAPC = 1.542	/
	Zhao 2025 [14]	All/all/China	Data: 1990–2019; projection: 2030	BAPC	Male: 64.44, PC = 4.37% Female: 31.99, PC = 29.2%	/	Male: 56.54, PC = –2.6% Female: 27.5, PC = 20.3%	Male: 1176.74, PC = –2.25% Female: 584.9, PC = 18.84%
3	He 2025 [19]	All/all/China	Data: 1990–2021; projection: 2036	BAPC	Male: increase by 4.8% Female: increase by 35.7%,	/	/	/
	Chen 2025 [33]	All/all/China	Data: 1990–2021; projection: 2036	BAPC	Male: stop increasing and tend to stabilize Female: keep increasing	/	Male: stop increasing and tend to stabilize Female: Keep increasing	/

Abbreviations: AS‐DALYs, age‑standardized disability‑adjusted life years; ASIR, age‐standardized incidence rate; ASMR, age‐standardized mortality rate; ASPR, age‐standardized prevalence rate; BAPC, Bayesian Age‐Period‐Cohort model; EAPC, estimated annual percentage change; PC, percentage change.

“/” means not reported.

3.4. Risk Factor Analysis

Among the 32 included studies, 37.5% (12/32) did not analyze risk factors, while 65.6% (21/32) reported three categories of risk factors: behavioral risks, environmental/occupational risks, and metabolic risks. Specifically, 50.0% (16/32) reported environmental/occupational risks, mainly including air pollution and occupational risks; 59.4% (19/32) reported behavioral risks, mainly including tobacco smoke and dietary risks; 9.4% (3/32) reported metabolic risks, mainly including high fasting plasma glucose (Figure 2).

FIGURE 2.

Risk factors of lung cancer.

3.5. Reporting Quality Analysis

Reporting quality varied significantly across items, with 88.2% (15/17) of items having reporting deficiencies (Figure 3). Among the 32 studies, reported items were 11.19 ± 1.67. Specifically, 19 studies were below the average level, and none met all 17 core STROBOD Statement items.

FIGURE 3.

Reporting quality of included studies.

High reporting rate items (>95%) were as follows: Item 3 (Introduction) achieved a reporting rate of 100.0% (32/32); Item 5 (Reference population) achieved a reporting rate of 96.9% (31/32); Item 6 (Reference period) achieved a reporting rate of 100% (32/32); Item 10 (Age‐conditional life expectancy used for calculating) achieved a reporting rate of 96.9% (31/32); Item 22 (Results) achieved a reporting rate of 96.9% (31/32); Item 24 (Summarize the key study findings) achieved a reporting rate of 96.9% (31/32).

Low reporting rate items (<15%) were as follows: Item 7 (Source and distribution of input parameters) showed a partial reporting rate of 100% (32/32); Item 26 (Strengths and limitations) showed a reporting rate of 12.5% (4/32); and Item 28 (Funding and Conflicts of interest) showed a reporting rate of 51.5% (4/32).

3.6. Publication Factors

This study conducted a stratified analysis of differences in reporting quality from four publication factors: publication year, language, number of authors, and participation of statisticians (File S2). In univariable linear regression, studies published before 2025 had more reported items than those after 2025, β = –0.50, 95% CI: –1.77 to 0.77. Specifically, the reporting rate of Item 26 was higher in studies published before 2025 (p < 0.05), not significant after FDR correction. English studies had more reported items than Chinese studies, β = 1.90, 95% CI: 0.83 to 2.97. English studies had higher reporting rates for Item 1 and Item 27 (p < 0.05), not significant after FDR correction. Studies with more than 6 authors had more reported items than those with ≤ 6 authors, β = 1.15, 95% CI: –0.02 to 2.32. Studies without statisticians had more reported items, β = –0.32, 95% CI: –1.58 to 0.94.

In multivariable linear regression adjusting for all above factors (Figure 4), only English language remained independently associated with a greater number of fully reported items, β = 1.70, 95% CI: 0.61 to 2.80, (p < 0.05). No independent associations were observed for publication year, number of authors, or statistician involvement (p > 0.05).

FIGURE 4.

Impact of publication factors on reporting quality.

4. Discussion

This study included 32 studies of lung cancer BoD derived from the GBD Database and conducted a cross‐sectional analysis to characterize publication trends and research features, while assessing reporting quality. Results showed a rapid growth in publications since 2025. Most studies focused on the global lung cancer burden, with core outcomes including ASIR, ASMR, ASPR, and AS‐DALYs. Behavioral risk was identified as the predominant contributor. Using core items of the STROBOD Statement, we found that 88.2% of the items were incompletely reported, and none of the studies fully reported all items. Further analysis revealed that publication language was significantly associated with reporting quality, with English studies showing higher reporting completeness than those in Chinese.

Publications on lung cancer BoD derived from the GBD Database have increased rapidly since 2025, accompanied by repetitive research topics. In our study, 90.6% of the included studies had Chinese first authors, consistent with a survey in Science [43]. This phenomenon may be partly driven by academic evaluation pressures and the accessibility of public databases, leading to research homogenization. Such repetitive studies consume limited research resources and provide limited value for policy‐making. In response, several journals, including PLOS and Frontiers, have begun rejecting manuscripts solely derived from secondary analyses of public databases. Correspondingly, the GBD Collaborators are also updating their policies to restrict publications and protect scientific credibility. Specifically, national and regional GBD analyses require notification to the GBD Secretariat via email upon submission to a journal, while global and subnational studies face stricter authorization rules.

Among the included studies, only 50% adopted predictive models to analyze BoD trends, of which 62.5% utilized the BAPC model. The consistency of results across studies using BAPC supports its stability for BoD prediction [44]. Some studies combined BAPC with ARIMA, which improves robustness and compensates for the limitations of single models [21]. However, inconsistencies were still observed in certain indicators, such as ASPR [27]. Even with identical models, variations in modeled data scope, parameter settings, and calibration iterations may lead to divergent results [45]. Such inconsistencies reduce the credibility of findings and limit their value for public health decision‐making. Notably, most studies only reported the model name without detailed settings, which compromises study reproducibility and reliability. Among the 16 predictive studies, only 3 declared adherence to the GATHER Statement, a reporting guideline for disease forecasting and health impact modeling [42], indicating insufficient methodological transparency. Standardized reporting of model specifications and stronger implementation of the GATHER Statement are needed to improve reproducibility and reliability.

The STROBOD Statement provides an authoritative checklist for BoD studies [8]. However, there are certain differences between the BoD studies derived from the GBD database and the conventional research. Unlike conventional studies that rely on primary data collection, the GBD database provides standardized and centrally processed data. Therefore, some STROBOD items, such as disease models, multimorbidity adjustments, and scenario analysis, are not applicable, as these processes are already standardized in the GBD Database. To ensure an appropriate assessment, we selected 17 core items based on the characteristics of the GBD Database. The results showed that an average number of reported items was 11.19 ± 1.67, and none of the studies met all the item requirements. This may reflect limited awareness and insufficient implementation of reporting guidelines, leading to incomplete and nonstandard reporting.

Reporting quality varied across items. Items 3 and 6 achieved 100% reporting rates, with low reporting difficulty and high necessity. In contrast, the reporting rates of Items 7, 8, 9, 26, and 28 were ≤ 50%. Since the GBD Database provides standardized data, researchers may overlook detailed reporting of epidemiological and demographic inputs, reducing the traceability and comparability of estimates. Reporting of Item 26 is related to the requirement differences, which limit the assessment of applicability and reliability. Both Chinese and English studies both neglected generalizability discussions, and Chinese journals do not require reporting of strengths and limitations. Insufficient reporting of Item 28 may relate to inadequate disclosure of conflicts of interest [46], which may introduce bias and weaken study objectivity. Most of these items are related to the reproducibility and transparency, with insufficient reporting undermining research credibility, which has been reported in systematic reviews and meta‐analyses [47].

To clarify the effect of publication factors on reporting quality, we conducted a stratified analysis. Publication language was the only factor significantly associated with reporting quality, with English studies showing higher completeness. This finding is consistent with other studies on systematic reviews and meta‐analyses [48], likely due to insufficient implementation of reporting guidelines by Chinese journals [49]. Specifically, language significantly affected Items 1 and 27, with better reporting in English studies. This is associated with the English journals’ manuscript standardized requirements and the promotion of open science [50]. No statistical significances for publication year, number of authors, and statisticians' participation, although some trends were observed. For example, studies published before 2025 reported more items than those published in 2025, possibly because the hastier submissions resulted in lower reporting completeness [51]. Studies with more than 6 authors had more reported items, as more collaborators tend to improve research and reporting quality [52]. Unlike other evidence from systematic reviews [53], the absence of statisticians did not reduce reporting quality, which may reflect the standardized analytical framework in the GBD Database.

This study conducts a cross‐sectional analysis of lung cancer BoD derived from the GBD database, revealing an abnormal publication growth since 2025, suggesting a potential academic resource waste. In addition, this study is the first reporting quality assessment of GBD‐derived BoD studies, identifying commonly underreported items, which provides a reference for reporting guidelines development. This study has limitations. To reduce heterogeneity, only studies derived from the GBD Database were included, which limits the generalizability of our findings to other BoD studies. Future work will analyze BoD studies from multiple data sources, construct a generalizable strategy for improving reporting quality, and develop the standardization of BoD studies.

In summary, the STROBOD Statement is not fully applicable to BoD studies derived from the GBD Database. A dedicated reporting guideline should be developed and incorporated into peer review [54]. Academic societies should promote guideline dissemination and implementation, encouraging researchers' proactive adherence to the reporting guideline [55]. Collaboration among researchers, journals, and the EQUATOR Network is essential to enhance research credibility. In addition, exclusive reliance on the GBD Database may limit data representativeness [56]. Future studies should integrate real‐world data or conduct cross‐database validation to improve the clinical and policy reference value of BoD studies [57, 58].

In conclusion, using lung cancer as a model disease, this study analyzed the BoD studies derived from the GBD Database and evaluated the reporting quality by the core items of the STROBOD Statement. It was found that BoD studies derived from the GBD Database are growing rapidly, with problems of repetitive topics and deficiencies in reporting quality. In the future, we will develop an exclusive reporting guideline for BoD studies derived from the GBD Database, aiming to improve the reporting quality of studies and reduce the waste of academic resources.

Funding

This work was supported by the National Natural Science Foundation of China (82360498); Science and Technology Department of Gansu Province under Grant (23JRRA1537); and 2026 Key Talent Project of Gansu Provincial Health Commission (2026SWJWRC001).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: Search Strategy (PubMed as an Example).

Supporting File 2: Classification of Publication Factors.