Abstract
Study question
Are period trackers giving women accurate information about their periods and ovulation?
Summary answer
The top 10 period trackers gave conflicting information on period dates, ovulation day and the fertile window.
What is known already
Period tracking applications allow women to track their menstrual cycles and receive a prediction for their periods. The majority of applications also provide predictions of day of ovulation and the fertile window. Previous research indicates applications are basing predictions on assuming women undergo a textbook 28-day cycle with ovulation occurring on day 14 and a fertile window between days 10 and 17.
Study design, size, duration
An audit of menstrual cycle apps was conducted on the Apple app store using menstrual cycle tracker/period tracker as the search terms. The top ten apps that followed the inclusion and exclusion criteria were analysed and used for this study. All apps had the ability to allow retrospective data entry giving future cycle predictions and fertile window, and nine of the apps predicted a day of ovulation.
Participants/materials, setting, methods
Five women’s profiles for 6 menstrual cycles were created and entered into each app. Cycle length (CL) and ovulation day (OD) for the 6th cycle were; Woman 1 – Constant 28 day CL, 0D 16, Woman 2 – Average 23 day CL, OD 13, Woman 3 – Average 28 day CL, OD 17, Woman 4 – Average 33 day CL, OD 20 and Woman 5 – Irregular, average 31 day CL, OD 14.
Main results and the role of chance
For cycle length, the apps all predicted woman 1’s cycles correctly but for women 2–5, the apps predicted 0 to 8 days shorter or longer than expected. For day of ovulation; for woman 1, no apps predicted this correctly; the apps ranged from day 13–15. For woman 2, 1 app was correct and overall the apps showed a lot of variation from day 8 to 13. For woman 3, no apps got it right, with a range of day 13–16. For woman 4, 2 apps got it right, but the apps ranged from day 13–20. For woman 5, no apps got right; the apps ranged from day 13–21. Irrespective of cycle length, 7 apps predicted a fertile window of 7 days in almost all cases; except 1 app that predicted 6 days for woman 2 and a different app which predicted 8 days for woman 4. For the remaining 3 apps, one always predicted a 10 day fertile window. One app predicted an 11 day fertile window in 4/5 women. One app predicted a 12 day fertile window in 4/5 women.
Limitations, reasons for caution
The five profiles created spanned a range of observed cycle characteristics, but many permutations are possible. A Monte Carlo type analysis could be conducted to examine these possibilities to provide more precise assessment of app performance, but as data had to be added manually into apps, this was not possible.
Wider implications of the findings: The apps do not use the same algorithm and show variation. The information given by these apps is not 100% accurate, especially for the day of ovulation and the fertile window which can only be predicted if using a marker of ovulation, such as basal body temperature or ovulation sticks.
Trial registration number
Not applicable
Periodic Coefficients and Random Fibonacci Sequences
