The performance of a fertility tracking device

Area of Science:

• Reproductive endocrinology and technology
• Biostatistics and data analysis

Background:

• Fertility tracking devices provide direct-to-user insights into a woman's reproductive cycle.
• Understanding algorithm adaptability is crucial for accurate fertility prediction.

Purpose of the Study:

• To evaluate how a fertility tracking device algorithm adapts to individual menstrual cycle variations.
• To assess algorithm performance under diverse conditions and data inputs.

Main Methods:

• Retrospective analysis of 107,020 menstrual cycles from 5328 women using a fertility tracking device (January 2004-November 2014).
• Processing of basal body temperature and menstruation data through Daysy 1.0.7 firmware.
• Sensitivity analyses were performed on temperature noise, skipped measurements, and user characteristics.

Main Results:

• The algorithm successfully distinguished biphasic cycles and provided personalized fertility status.
• A direct linear relationship was observed between the number of measurements and the fertility tracker's output.
• With 80-100% usage, the device identified 41.4% fertile, 42.4% infertile, and 15.9% undefined days, with infertile days decreasing and undefined days increasing proportionally to measurements.

Conclusions:

• Fertility tracker algorithms can accurately personalize fertility status based on daily basal body temperature and menstruation data.
• The number of daily measurements directly correlates with the precision of the fertility tracker's output.
• The device demonstrates efficacy in distinguishing biphasic cycles, offering valuable reproductive health information.