Ex ante prediction of NCT03914651 (300 IU vs 150 IU rFSH starting dose in women aged 35–42 with poor ovarian reserve) from a dose–oocyte–cumulative live birth model
Time-stamped, ex ante prediction of the randomised trial NCT03914651 (Reproductive & Genetic Hospital of CITIC-Xiangya; 300 IU versus 150 IU follitropin alfa starting dose, GnRH antagonist, women aged 35–42 years with AFC ≤ 5 or AMH ≤ 1.2 ng/mL; primary outcome cumulative live birth from the first stimulation). The prediction comes from a dose–oocyte–cumulative live birth model calibrated only on published data. No parameter was fitted to the outcomes of the trial, whose results were not public at deposit. The trial population was anchored on published data from the same centre. Primary prediction (same-centre anchor, γ = 0.09): CLBR 9.3% (150 IU) versus 9.5% (300 IU), a difference of +0.2 points (95% prediction interval for a trial of 130 women per arm: −6.9 to +7.3). Oocytes: 1.66 versus 1.91. A higher starting dose is predicted not to increase cumulative live birth. The record contains the prediction statement, the complete code, a one-command reproduction (reproduce_all.py) and SHA-256 hashes of every file (MANIFEST.json).
Authors
- Philippe Pinton (ORCID: https://orcid.org/0000-0002-4227-7687)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24411685
- Primary Topic
- Reproductive Health and Technologies
- Type
- article
- Field-Weighted Citation Impact
- 0.00