Improving estimation efficiency of case-cohort studies with interval-censored failure time data

Area of Science:

• Biostatistics
• Epidemiology
• Survival Analysis

Background:

• Case-cohort studies are cost-effective for large cohorts with expensive covariates.
• Interval-censored failure time data requires specialized analysis methods.
• Standard inverse probability weighting in case-cohort studies can be inefficient.

Purpose of the Study:

• To develop an efficient regression analysis for case-cohort studies with interval-censored failure time data.
• To improve upon existing inverse probability weighting methods by incorporating full cohort information.
• To provide a statistically robust method for analyzing complex survival data.

Main Methods:

• Developed a sieve maximum weighted likelihood estimator under the Cox model.
• Proposed an update procedure using information from the full cohort to enhance the initial estimator.
• Employed a weighted bootstrap procedure for variance estimation.

Main Results:

• The proposed updated estimator is consistent and asymptotically normal.
• The updated estimator is at least as efficient as the original estimator.
• The method effectively incorporates auxiliary variables for improved estimation efficiency.

Conclusions:

• The novel method offers a more efficient and accurate approach to regression analysis in case-cohort studies with interval-censored data.
• The proposed technique enhances survival data analysis by leveraging full cohort information.
• Simulation results and a real-world trial application demonstrate the method's practical utility and effectiveness.