Natural conception: repeated predictions over time.

Area of Science:

• Reproductive Medicine
• Clinical Epidemiology
• Biostatistics

Background:

• Unexplained subfertility affects many couples seeking fertility treatment.
• Existing prediction models offer a one-time probability assessment, limiting their utility for ongoing management.
• There is a need for dynamic models that can reassess conception chances over time.

Purpose of the Study:

• To develop and evaluate a dynamic prediction model for natural conception in couples with unexplained subfertility.
• To provide repeated, updated probability estimates for conception at various time points.
• To enhance counseling for couples undergoing fertility workups.

Main Methods:

• Prospective cohort study of 4999 couples undergoing fertility workup.
• Development of a dynamic prediction model using predictors like female age, duration of subfertility, and sperm motility.
• Internal validation of the model's calibration, discrimination, and prediction range utility.

Main Results:

• The dynamic model estimated a 27% median probability of conception in the first year.
• For couples not pregnant after 6, 12, and 18 months, the model estimated subsequent yearly conception probabilities of 20%, 15%, and 13%, respectively.
• The model's predictions were sufficiently broad for counseling up to two years post-workup.

Conclusions:

• The developed dynamic prediction model offers a valuable tool for reassessing natural conception chances in unexplained subfertility.
• This model provides crucial information for counseling couples, especially after periods of unsuccessful expectant management.
• Further external validation is required to confirm the model's generalizability.