基于对数累积风险比轨迹的获益时间估计方法研究*

樊静 ,  蔡嘉鑫 ,  索珊珊 ,  查景丽 ,  李超 ,  邵方

中国卫生统计 ›› 2026, Vol. 43 ›› Issue (5) : 660 -668.

PDF (1561KB)
中国卫生统计 ›› 2026, Vol. 43 ›› Issue (5) : 660 -668. DOI: 10.11783/j.issn.1002-3674.2026.05.004
论著

基于对数累积风险比轨迹的获益时间估计方法研究*

作者信息 +

Study on Time to Benefit Estimation Method based on Log Cumulative Hazard Ratio Trajectory

Author information +
文章历史 +
PDF (1597K)

摘要

目的 传统获益时间(time to benefit, TTB)估计多基于序贯截断(sequential truncation)的Cox比例风险模 型风险比(hazard ratio, HR),但存在反复截断数据导致统计性质不明确、结果易被误解为时变效应及计算量大等问题。 本研究旨在建立一种方法学更严谨、计算效率更高的TTB估计框架。方法 提出基于对数累积风险比时间轨迹的TTB估 计新框架。整合Weibull回归模型与Kaplan-Meier (KM) 估计,并通过模拟研究及真实临床试验数据,系统比较新框架与 传统序贯截断Cox模型在TTB 估计中的性能。结果 模拟研究表明,新框架中Weibull回归模型综合性能最优,估计精度 高且稳定,均方根误差(root mean square error,RMSE)始终小于1。相比之下,序贯截断Cox 模型在部分情景中对TTB 持 续显著时间点估计的RMSE 超过10。此外,Weibull 回归模型计算速度较序贯截断Cox 模型提升约35~250 倍。结论 基 于对数累积风险比时间轨迹的TTB估计框架在避免序贯截断方法缺陷的同时,为治疗效应时间演变的稳健评估提供了统 一且可靠的方法。

Abstract

Objective Conventional estimation of the Time to Benefit (TTB) typically relies on Hazard Ratios (HR) derived from sequentially truncated Cox proportional hazards models. However, this approach is limited by poorly defined statistical properties due to repetitive data truncation, potential misinterpretation of results as time-varying effects, and high computational intensity. This study aims to establish a more methodologically rigorous and computationally efficient framework for TTB estimation.Methods We propose a novel TTB estimation framework based on the temporal trajectory of the log-cumulative hazard ratio. By integrating Weibull regression models with Kaplan-Meier (KM) estimation, we systematically compared the performance of the proposed framework against the traditional sequentially truncated Cox model using both simulation studies and real-world clinical trial data.Results Simulation results demonstrate that the Weibull regression model within the new framework yields superior overall performance, characterized by high estimation precision and stability; its root mean square error (RMSE) remained consistently below 1. In contrast, the RMSE of the sequentially truncated Cox model exceeded 10 in certain scenarios when estimating the time point of sustained significance for TTB. Furthermore, the computational speed of the Weibull regression model was approximately 35 to 250 times faster than that of the sequentially truncated Cox model.Conclusion The TTB estimation framework based on the temporal trajectory of the log- cumulative hazard ratio effectively addresses the inherent flaws of sequential truncation methods. It provides a unified and reliable methodology for the robust assessment of the temporal evolution of treatment effects.

关键词

获益时间 / 生存分析 / Weibull回归模型 / Kaplan-Meier估计 / Cox比例风险模型

Key words

Time to benefit / Survival analysis / Weibull regression model / Kaplan-Meier estimation / Cox proportional hazards model

引用本文

引用格式 ▾
樊静,蔡嘉鑫,索珊珊,查景丽,李超,邵方. 基于对数累积风险比轨迹的获益时间估计方法研究*[J]. 中国卫生统计, 2026, 43(5): 660-668 DOI:10.11783/j.issn.1002-3674.2026.05.004

登录浏览全文

4963

注册一个新账户 忘记密码

参考文献

[1]

Lee SJ, Leipzig RM, Walter LC. “When will it help?” incorporating lagtime to benefit into prevention decisions for older adults[J].JAMA, 2013, 310(24): 2609.

[2]

Krishnaswami A, Peterson ED, Goyal P, et al. Time to benefit and harm of intensive blood pressure treatment: insights from SPRINT [J].European Heart Journal—Quality of Care and Clinical Outcomes, 2021, 7(4): e1-e2.

[3]

Yourman LC, Cenzer IS, Boscardin WJ, et al. Evaluation of time to benefit of statins for the primary prevention of cardiovascular events in adults aged 50 to 75 years: a meta—analysis [J].JAMA Internal Medicine, 2021, 181 (2): 179-185.

[4]

Lu M, Wu Z, Zhao Y, et al. Time to benefit estimation in multicenter studies using flexible hazard shared frailty models [J].BMC Medical Research Methodology, 2026, 26 (1): 84.

[5]

Chen K, Nie Z, Shi R, et al. Time to benefit of sodium—glucose cotransporter—2 inhibitors among patients with heart failure[J].JAMA Network Open, 2023, 6(8): e2330754—e2330754.

[6]

Ho VS, Cenzer IS, Nguyen BT, et al. Time to benefit for stroke reduction after blood pressure treatment in older adults: a meta —analysis[J].Journal of the American Geriatrics Society, 2022, 70 (5): 1558-1568.

[7]

Horiguchi M, Tian L, Kehl KL, et al. Assessing delayed treatment benefits of immunotherapy using long—term average hazard: a novel test/estimation approach [J].Lifetime Data Analysis, 2025, 31 (4): 784-809.

[8]

Bellera CA, MacGrogan G, Debled M, et al. Variables with time—varying effects and the Cox model: some statistical concepts illustrated with a prognostic factor study in breast cancer[J].BMC Medical Research Methodology, 2010, 10(1): 20.

[9]

Carroll KJ. On the use and utility of the Weibull model in the analysis of survival data[J].Controlled Clinical Trials, 2003, 24(6): 682-701.

[10]

Goel MK, Khanna P, Kishore J. Understanding survival analysis: Kaplan—Meier estimate[J].International Journal of Ayurveda Research, 2010, 1(4): 274.

[11]

Zhang Z. Parametric regression model for survival data: Weibull regression model as an example[J].Annals of Translational Medicine, 2016, 4(24): 484.

[12]

Odell PM, Anderson KM, D’Agostino RB. Maximum likelihood estimation for interval—censored data using a Weibull—based accelerated failure time model[J].Biometrics, 1992, 48(3): 951-959.

[13]

Pawitan Y. In all likelihood: statistical modelling and inference using likelihood[M]. Oxford: Oxford University Press,2001.

[14]

Rodenkirchen J, Hoyer A. An Extension of Greenwood’s Formula Variances [PP/OL]. (2024—06—12) [2025—12—05].https://doi.org/10. 48550/arXiv. 2406. 07994.

[15]

SPRINT Research Group. A randomized trial of intensive versus standard blood—pressure control[J].New England Journal of Medicine, 2015, 373(22): 2103-2116.

[16]

Bhatt DL, Szarek M, Steg PG, et al. Sotagliflozin in patients with diabetes and recent worsening heart failure[J].New England journal of medicine, 2021, 384(2): 117-128.

[17]

Greenwood M. A report on the natural duration of cancer [R].Reports on Public Health and Medical Subjects, No. 33. London: His Majesty’s Stationery Office, 1926.

[18]

Neyman J. Outline of a theory of statistical estimation based on the classical theory of probability[J].Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences, 1937, 236(767): 333-380.

[19]

Hernán MA. The hazards of hazard ratios[J].Epidemiology, 2010, 21(1): 13-15.

基金资助

国家自然科学基金(82473732;82322060)

四大慢病重大专项(2024ZD0522700)

南京医科大学南京公共卫生研究院科研创新项目(NCX2405)

煤炭环境致病与防治教育部重点实验室开放课题基金项目(MEKLCEPP/SXMU-2025MEKLCEPP/SX MU-202501)

AI Summary AI Mindmap
PDF (1561KB)

0

访问

0

被引

详细

导航
相关文章

AI思维导图

/

〈 〉