Estimating time to pregnancy from current durations in a cross-sectional sample.

Area of Science:

• Reproductive Epidemiology
• Biostatistics
• Survival Analysis

Background:

• Estimating time to pregnancy (TTP) is crucial for reproductive health.
• Traditional methods often rely on longitudinal data, which can be resource-intensive.
• Cross-sectional data presents unique statistical challenges for TTP estimation.

Purpose of the Study:

• To propose and investigate a new statistical design for estimating TTP distributions.
• To address the challenges of using cross-sectional data for TTP estimation.
• To explore both non-parametric and parametric estimation approaches.

Main Methods:

• The proposed design involves recording current durations in a cross-sectional sample of women.
• Statistical methods analogous to renewal time distributions from backward recurrence times are employed.
• Non-parametric estimation is studied in detail, with a parametric approach also indicated.

Main Results:

• The study illustrates the proposed method using Monte Carlo simulations.
• The approach is further validated on data from a European collaborative study.
• The performance and applicability of the new estimation design are discussed.

Conclusions:

• The developed statistical design offers a viable method for estimating TTP distributions from cross-sectional data.
• This approach provides a valuable alternative for reproductive health studies, especially when longitudinal data is unavailable.
• Further discussion on the role and applicability of this method is provided for researchers.