Understanding cycle dynamics of women who are trying to conceive: a real-world data approach
Primary Research Question: What are the cycle characteristics of women trying to conceive (TTC) using real-world data collected nightly via the Ava Fertility Tracker?
Key Findings: We conducted a retrospective, longitudinal analysis of data from 74,671 Ava users who recorded 483,881 menstrual cycles between December 2016 and September 2019. On average, women were of relatively normal weight and were typically TTC for their first child. Despite typical mean cycle (28.98 days, SD=3.19) and luteal lengths (13.94 days, SD=15.52), an atypical variation in cycle length per user (9 days) was observed. On average, the Ava Fertility Tracker detected peak fertility on cycle day 16.45 (SD=7.90). This descriptive analysis is the first real-world dataset evaluating cycle characteristics of women who are TTC.
Citation: Goodale, B. M., Sakalidis, V., & Shilaih, M. Understanding cycle dynamics of women who are trying to conceive: a real-world data approach. In: ESC Abstract Book 2020: 16th Congress of the European Society of Contraception and Reproductive Health. European Society of Contraception and Reproductive Health; 2020; 42-43.
Innovative Trial Design Using Digital Approaches: An example from reproductive medicine
Primary Research Question: Can a wearable fertility tracker facilitate site-less clinical trials and increase participants' retention, compliance and satisfaction?
Key Findings: Wearable fertility trackers enable clinicians to conduct site-less clinical trials, while simultaneously ensuring higher quality data collection. Our innovative site-less trial design led to faster recruitment time and higher overall participant satisfaction (98% reported their experience as good or very good). Taking multiple physiological tests and responding to a survey daily, 92% of participants completed the minimum trial requirements while 48.5% opted to remain in the study for an additional three months. Our trial decreased data loss through real-time virtual participant monitoring, and provides a roadmap for how incorporating digital approaches can increase the quality and quantity of data contributing to reproductive health research.
Citation: Hamvas, G., Hofmann, A., Sakalidis,V., Goodale, B., M., Shilaih, M., & Leeners, B. Innovative trial design using digital approaches: an example from reproductive medicine. Poster presented at: Annual Clinical and Scientific Meeting of the American College of Obstetricians and Gynecologists; April 24-27, 2020; virtual.
Self-reported pregnancy rate among clinically subfertile women using a wearable fertility tracker
Primary Research Question: Through two studies (S1 and S2), we examined the efficacy of wearable sensor technology and artificial intelligence (AI) in helping subfertile, real-world women conceive.
Key Findings: In S1, women who had previously purchased the Ava Fertility Tracker (n=1758) completed a survey about their trying to conceive (TTC) journey. Almost 90% of women rated the Ava bracelet as easy or very easy to use. Nearly one in three respondents had a diagnosed condition affecting their reproductive health. Of the 495 users requiring fertility services, 75% had used the Ava Fertility Tracker in conjunction with their treatment. In S2, we calculated the 1-year pregnancy rate for 26,686 real-world Ava users who met the clinical definition of subfertility. Within a year, 28% of the subfertile cohort reported a pregnancy in the Ava app (n=2251), with an average time to pregnancy of 150±1.9 days. Furthermore, pregnancies among subfertile users accounted for a fifth of all pregnancies in S2.
Citation: Goodale, B.M., Rothenbühler, M., Cronin, M. Self-reported pregnancy rate among clinically subfertile women using a wearable fertility tracker. Poster presented at: Annual Clinical and Scientific Meeting of the American College of Obstetrics and Gynecologists; April 30 - May 2, 2021.
Detecting the fertile window in irregular menstrual cycles using a wearable medical device
Primary Research Question: Can the Ava bracelet identify hormonal patterns and fertile window (FW) in women with irregular cycles, including subjects with and without diagnosed polycystic ovaries syndrome (PCOS)?
Key Findings: Results included 161 cycles from women with irregular but undiagnosed/unknown PCOS (meanduration= 28.72 days [95% CI, 28.1-29.4]) and 61 cycles from women with confirmed PCOS (meanduration = 34.9 days [95% CI, 32.1-37.6]). Ava’s algorithm accurately identified fertile days in 82.1% of irregular cycles (95% CI, 80.0-83.8). Almost 30% of irregular cycles had more than one FW (mean=1.5 FW per cycle).
Citation: Rothenbühler M, Schmutz E, Hamvas G, et al. Detection of fertile window in irregular cycles using a wearable medical device. Poster presented at: International Federation of Fertility Societies 2019 World Congress; April 11-14, 2019; Shanghai, China.
Previous cycle tracking with a wearable multiparameter device reduces time to conception
Primary Research Question: Do real-world Ava users who cycle track with a wearable device prior to trying to conceive become pregnant faster than women who did not cycle track first?
Key Findings: Data from 12,540 women who purchased the Ava bracelet and reported a positive pregnancy test were included, with subjects divided between the prior cycle tracking (PCT; n=451) and no prior cycle tracking (NCT; n=12,089) groups. Time to pregnancy was significantly faster in the PCT group (mean=75 days, SD= 51 days) versus the NCT group (mean=89 days, SD=65 days; t(587)=-8.5936, p<.001).
Citation: Gibson S, Bilic A, Sakalidis V, Goodale BM, Shilaih M, Shulman L. Previous cycle tracking with a wearable multiparameter device reduces time to conception. Poster presented at: Central Association of Obstetricians and Gynecologists; Oct. 16-19, 2019; Cancun, Mexico.
Detection of the fertile window using a wearable medical device and the calendar method: A comparative study
Primary Research Question: When identifying the six-day fertile window, how does artificial intelligence-based predictions from wearable sensor technology compare to the calendar method in accuracy and precision?
Key Findings: The Ava bracelet’s accuracy in identifying the fertile days was 88.1% (SD=9.1%) compared to 76.8% (SD=5.1%) for the Standard Days method, 69.2% (SD=15.6%) for the Rhythm Method, and 67.6% (SD=16.1%) for the Alternative Rhythm Method. Furthermore, the wearable fertility tracker had the highest precision of any of the methods analyzed (70.3%, SD=21.9% v. 42.7%-47.7% for the calendar methods [SDs=7.6%-13.0%]).
Citation: Mouriki E, Bilic A, Goodale BM, et al. Detection of the fertile window using a wearable medical device and the traditional calendar method: A comparative study. Poster presented at: American Society of Reproductive Medicine Scientific Congress and Expo; Oct. 12-16, 2019; Philadelphia, PA, USA.
Capturing the physiological characteristics of early pregnancy using wrist worn wearables
Primary Research Question: Can wearable sensors on the Ava bracelet capture physiological changes associated with early pregnancy, in particular differences in heart rate variability (HRV), pulse rate, breathing rate, and wrist-skin temperature?
Key Findings: Analysis included 131 conceptive and 853 non-conceptive cycles from 330 women. In comparison to the late luteal phase of non-conceptive cycles, conceptive cycles were characterized by: an increase in pulse rate (1.43 beats/minute, p<.001), breathing rate (0.31 breaths/minute, p<.001), and wrist-skin temperature (0.05 C, p<.05). In addition, non-conceptive cycles were more likely to have lower HRV (-3.14 standard units, p<.001).
Citation: Shilaih M, Goodale BM, Falco L, Kübler F, Dammeier F, Leeners B. Capturing the physiological characteristics of early pregnancy using wrist worn wearables. Poster presented at: European Society of Human Reproduction and Embryology Annual Meeting; July 1-14, 2018; Barcelona, Spain.
Proof of concept pilot study: Digital women’s health based on wearables and big data
Primary Research Question: The goal of this pilot study was to determine if physiological data measured via a wrist worn wearable sensor could allow for an individualized, AI-driven form of natural family planning.
Key Findings: In a first step, wrist skin temperature (WST) and pulse rate readings were analyzed. Both showed significant differences between the follicular and luteal phase. The minimum average resting pulse rate occurred in the follicular phase (mean=55.5 beats/minute) and maximum resting pulse rate in the luteal phase (mean=59.3 beats/minute). WST followed the same pattern, with 34.3 compared to 34.7, respectively.
Citation: Stein P, Falco L, Kübler F, et al. Digital women’s health based on wearables and Big Data: New findings in physiological changes throughout the menstrual cycle. Poster presented at: Germany Society of Gynecology and Obstetrics (DGGG) Annual Meeting; Oct. 19-22, 2016; Stuttgart, Germany.