Hepatitis B eligibility criteria in nasopharyngeal carcinoma trials of immune checkpoint inhibitors: a cross-sectional registry analysis — supporting registry data and analysis code

Published: 17 November 2026| Version 3 | DOI: 10.17632/jr58x64ytc.3
Contributors:
Zixuan Wang,
,
,

Description

Frozen public ClinicalTrials.gov records supporting a cross-sectional analysis of hepatitis B eligibility wording in nasopharyngeal carcinoma trials of immune checkpoint inhibitors. The primary cohort comprises 211 records first posted from 1 January 2015 through 17 October 2026, using the 17 October 2026 snapshot. The 827 screened candidates comprise 211 main-analysis records, 16 sensitivity-only records, 596 exclusions and four unresolved records. Explicit conditional permission occurs in 31/211 trials (14.7%; Wilson 95% confidence interval 10.5%-20.1%). Version 3 adds agreement analyses and a dated clarification of the independently recorded author decisions for the 2026 extension. Wang and Song independently assessed 54 candidates and the HBV categories of the 32 included records before discussion or reconciliation. Screening agreement is 54/54 (100%; unweighted Cohen's kappa 1.000); included-record HBV-category agreement is 32/32 (100%; kappa 1.000). These are batch-specific estimates. The worksheets retain preliminary proposals; independence between reviewers does not establish blinding to them. No scientific decisions changed. The primary time window was extended on 12 November 2026 after both results had been inspected. The original 2015-2025 cohort remains a sensitivity analysis: 27/179 (15.1%). The original protocol and dated amendment chronology are retained. Use Supplementary_Data_S2_211records_20261117.zip for the current data, evidence, tables, figures, agreement matrices and reproducible code. Supplementary_Data_S1_MendeleyData_v3.zip and Version_1_documentation preserve the historical analysis; its agreement results concern the original screening frame. README_for_repository.md and FILE_DESCRIPTIONS.md describe this release. Hong Kong, Macao and Taiwan are included in the China group, with trial deduplication. No individual participant dataset is included. Registry text is attributed to ClinicalTrials.gov and retains its source terms (https://clinicaltrials.gov/about-site/terms-conditions). Authors' original contributions are offered under CC BY 4.0. Computational reproduction verifies arithmetic, not the correctness of textual judgments.

Files

Steps to reproduce

Extract Supplementary_Data_S2_211records_20261117.zip, preserving UTF-8 filenames and folders. From the extracted Supplementary_Data_S2 directory, run python code/analyze.py with Python 3.12 or compatible Python 3. This standard-library program makes no network requests and regenerates derived_main211.json, analysis.json and cohort tables from archived inputs. Expected results: primary 31/211 (14.7%); original-window sensitivity 27/179 (15.1%). Run python code/analyze_author_agreement.py to reproduce agreement from the preserved Wang and Song decision columns. Outputs include author_review/agreement_2026_20261117.json, tables/S8_2026_agreement_summary.csv and tables/S8_2026_agreement_contingencies.csv. Expected screening agreement: 54/54, kappa 1.000. Expected included-record HBV-category agreement: 32/32, kappa 1.000. Full margins and zero cells are included; excluded records with blank categories are not counted as category agreements. For Figure S1, run python code/plot_screening_flow.py with matplotlib (3.11.2 used). PDF, PNG and SVG outputs are in figures/. For the historical analysis, extract Supplementary_Data_S1_MendeleyData_v3.zip separately and run its own code/analyze.py. Do not mix S1 and S2 directories. Read AMENDMENT_20261112.md, provenance.json and author_review/independence_clarification_20261117.json for the amendment and assessment chronology. Check file integrity against manifest.json SHA-256 hashes. Source records and author decisions are not overwritten.

Categories

Clinical Trial, Hepatitis B, Nasopharyngeal Carcinoma

Licence