Inference for restricted mean survival time as a function of restriction time under length-biased sampling.

Stat Methods Med Res

Department of Biostatistics and Bioinformatics, Duke University, Durham, NC, USA.

Published: September 2024

The restricted mean survival time (RMST) is often of direct interest in clinical studies involving censored survival outcomes. It describes the area under the survival curve from time zero to a specified time point. When data are subject to length-biased sampling, as is frequently encountered in observational cohort studies, existing methods cannot estimate the RMST for various restriction times through a single model. In this article, we model the RMST as a continuous function of the restriction time under the setting of length-biased sampling. Two approaches based on estimating equations are proposed to estimate the time-varying effects of covariates. Finally, we establish the asymptotic properties for the proposed estimators. Simulation studies are performed to demonstrate the finite sample performance. Two real-data examples are analyzed by our procedures.

Download full-text PDF

Source
http://dx.doi.org/10.1177/09622802241267812DOI Listing

Publication Analysis

Top Keywords

length-biased sampling
12
restricted survival
8
survival time
8
function restriction
8
restriction time
8
time
6
inference restricted
4
survival
4
time function
4
time length-biased
4

Similar Publications

The motivation for this paper is to determine factors associated with time-to-fertility treatment (TTFT) among women currently attempting pregnancy in a cross-sectional sample. Challenges arise due to dependence between time-to-pregnancy (TTP) and TTFT. We propose appending a marginal accelerated failure time model to identify risk factors of TTFT with a model for TTP where fertility treatment is included as a time-varying treatment to account for their dependence.

View Article and Find Full Text PDF
Article Synopsis
  • Two main methods for tracking healthcare-associated infections (HAIs) are longitudinal surveillance, which tracks incidence rates over time, and point prevalence surveys (PPSs), which offer quicker snapshots but can be biased.
  • The study compared Ventilator-associated pneumonia (VAP) and central-line-associated bloodstream infections (CLABSI) rates from both a comprehensive surveillance system (GiViTI) and calculated rates from PPS data, using inverse probability weighting to improve accuracy.
  • Findings showed that weighted prevalence estimates better aligned with direct surveillance results, indicating that this method can yield more reliable incidence rates when full surveillance isn't practical.
View Article and Find Full Text PDF

The restricted mean survival time (RMST) is often of direct interest in clinical studies involving censored survival outcomes. It describes the area under the survival curve from time zero to a specified time point. When data are subject to length-biased sampling, as is frequently encountered in observational cohort studies, existing methods cannot estimate the RMST for various restriction times through a single model.

View Article and Find Full Text PDF

Recent innovations have focused on the creation of new families that extend well-known distributions while providing a huge amount of practical flexibility for data modeling. Weighted distributions offer an effective approach for addressing model building and data interpretation problems. The main objective of this work is to provide a novel family based on a weighted generator called the length-biased truncated Lomax-generated (LBTLo-G) family.

View Article and Find Full Text PDF
Article Synopsis
  • The paper introduces a new statistical method for analyzing daily coronavirus case and fatality statistics, resulting in the creation of the Discrete Marshall-Olkin Length Biased Exponential (DMOLBE) distribution.
  • This distribution is adaptable due to its varied probability mass and failure rate functions, and parameters like mean, variance, and hazard functions were calculated to assess its performance.
  • Through simulation studies, it was found that the DMOLBE distribution provides a better fit for COVID-19 mortality data from China and Pakistan compared to other discrete models, demonstrating its effectiveness in this context.
View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!