Abstract
Objective:
To identify determinants of selected pregnancy testing behaviors among pregnancy planners and to elucidate the relationship between pregnancy testing and detection.
Methods:
In Pregnancy Study Online (PRESTO), a North American prepregnancy cohort study of pregnancy planners (2018–2024), participants reported day-specific information about pregnancy testing spanning four days before day of expected menstruation through four days after. We used generalized linear models to estimate the association between maternal attributes and pregnancy testing behavior, characterized as timing of first test and testing frequency. We used quantile regression to estimate timing of pregnancy detection among participants who conceived by maternal characteristics and pregnancy testing behavior. We estimated the adjusted risk of having a negative test and the probability of detecting a very early loss by timing of pregnancy test.
Results:
We analyzed data from 20,458 pregnancy tests across 6,569 unique participants. 40.7% of participants reported they engaged in “very early” testing, defined as testing more than four days before their expected period. We observed a range of pregnancy testing intensity, with some participants testing only once and others testing every day. Among participants who detected pregnancy, very early testers were over five times more likely to have a negative test before a positive test than those who waited until the day of expected period to test (adjusted risk ratio [aRR] 5.89, 95% confidence interval [CI] 4.73, 7.33). Very early testers were over three times more likely to have an initial positive test followed by a negative test, likely reflecting increased detection of very early losses (aRR 3.80, 95% CI 2.12, 6.80).
Conclusions:
Patterns in home pregnancy testing varied widely among pregnancy planners. Early initiation of pregnancy testing was associated with slightly earlier pregnancy detection, but also a marked increase in risk of negative tests and detection of very early losses.
PRÉCIS:
Early pregnancy testing was associated with slightly earlier detection of pregnancy, an increase in initial negative tests, and greater detection of very early losses.
INTRODUCTION
When to test for pregnancy is one of the first decisions an individual makes for their pregnancy health. Little is known about how those planning to conceive approach home testing. Most individuals receive little or no professional guidance about home pregnancy testing. Decisions about how to test likely reflect the coalescence of pregnancy test manufacturers’ instructions, informal advice, financial and logistical considerations, an awareness of one’s own menstrual cycle, and an understanding of how test timing relates to test sensitivity.
Pregnancy testing is naturally linked to pregnancy detection and, subsequently, timing of behavior modification1, 2 and healthcare system engagement.3, 4 Earlier pregnancy detection is associated with improved birth outcomes,5 which may be a function of who detects pregnancy early and earlier initiation of the cascade of events that follows a first positive pregnancy test. However, an emphasis of detecting pregnancy as early as possible discounts drawbacks of very early testing such as experiencing negative tests before human chorionic gonadotropin levels are sufficiently high and the risk of detecting an early loss.
Understanding different approaches to home pregnancy testing and their implications is the first step in identifying best practices. We leveraged data from a prospective cohort of pregnancy planners to characterize patterns of home pregnancy test use in terms of frequency and timing relative to the date of expected menses. We also identified how testing behaviors relate to timing of pregnancy detection, risk of having an initial negative test when pregnant, and risk of detecting early losses.
METHODS
Participants were enrolled in Pregnancy Study Online (PRESTO), an ongoing internet-based cohort of pregnancy planners in the United States and Canada.6 Individuals eligible to participate were aged 21–45, not currently using contraception, and planning a pregnancy without the use of assisted reproductive technology. Upon enrollment, participants completed a self-administered baseline questionnaire that queried basic demographics, medical history, reproductive history, lifestyle factors, and menstrual cycle characteristics.
On bimonthly follow-up questionnaires, participants reported their date of last menstrual period (LMP) and day-specific information about whether they tested for pregnancy and the result of each pregnancy test in a window centered on day of expected period. Participants were able to select the pregnancy testing timing scale of their choice: 1) relative to their day of expected menstrual period (≥4 days prior, 3 days prior, 2 days prior, 1 day prior, day of expected menstrual period, 1 day after, 2 days after, 3 days after ≥4 days after; default scale) or 2) as days from their LMP (Appendix 1, available online at http://links.lww.com/xxx). Responses were harmonized to be centered on day of expected menstrual period using participants self-reported cycle length. If a participant reported irregular cycles, their most recent self-reported cycle length was used. If both typical and most recent cycle length was missing, a cycle length of 28 days was used for harmonization. Participants who had a positive pregnancy test by the first follow-up survey also reported the date of their first positive pregnancy test. A subset of U.S. participants were offered free home pregnancy tests.7 However, participants were not given any instructions on how or when to test for pregnancy. Instead, participants were mailed a letter that included the home pregnancy tests and information about pregnancy testing via the company website: Clearblue (4/19/2017–12/14/2023) or PREGMATE (1/3/2024–present). All participants provided informed consent and the study protocol was approved by the Institutional Review Boards at Boston University Medical Campus (H-31848) and Vanderbilt University Medical Center (#241910).
We analyzed PRESTO data collected from July 2018 through December 2024, reflecting when detailed questions regarding home pregnancy testing frequency, timing, and test results were introduced to the follow-up questionnaire. We restricted analyses to participants who provided information about pregnancy testing in the first follow-up cycle after enrollment.
Given the non-normal distribution of data on frequency of pregnancy testing and timing of first pregnancy test, we describe these behaviors using median and inter-quartile range (IQR) and visually using histogram plots. We categorized timing of first pregnancy test into three groups: very early (four or more days before expected period), early (one to three days before expected period), or expected or late (day of expected period or later). We categorized frequency of pregnancy testing within the window queried (four days before expected menstrual period through four days after) into three groups: infrequent (one testing day), moderate (2–3 testing days), or frequent (four or more testing days). Cut points for frequency categories were informed by tertiles of testing frequency among participants who did not detect a pregnancy. We assessed bivariate associations between maternal characteristics and timing of first home pregnancy test and frequency of testing using generalized linear models to calculate risk ratios (RR) and 95% confidence intervals (CI).
We created nine categories based on the three levels of timing of first test and testing frequency to better understand patterns of testing. To represent different pregnancy testing behaviors visually, we depicted timing of pregnancy testing stratified by cumulative number of positive or negative tests at time of test by day of pregnancy testing
Since a random subset of participants received free pregnancy tests, we used quantile regression to test for differences in frequency or timing of first test by receipt of free test. We used generalized linear models to estimate the risk of having a positive pregnancy test followed by a negative test, by timing of first home pregnancy test, adjusting for cycle length, maternal age, history of miscarriage, and parity. Covariates included in the model were selected a priori using directed acyclic graphs. We performed a sensitivity analysis to assess if estimates changed when receipt of free home pregnancy test was included in the model.
Next, we evaluated timing of the first positive home pregnancy test among participants who detected a pregnancy to understand timing of pregnancy confirmation. We used date of first positive pregnancy test to depict the distribution of timing of pregnancy recognition in reference to date of expected period, which accounts for typical cycle length, and in reference to LMP. We present median and IQR for timing of first positive pregnancy test by maternal characteristics. We used quantile regression to calculate unadjusted median day of first positive pregnancy test from LMP and corresponding 95% CI for level of maternal characteristic and receipt of free pregnancy test.
Among participants who detected a pregnancy, we used generalized linear models to estimate the risk of having an initial negative pregnancy test according to the timing of first home pregnancy test. Effect estimates were adjusted for cycle length modeled using restricted cubic splines. We performed a sensitivity analysis to assess if estimates changed when receipt of free home pregnancy test was included in the model.
Analyses were performed using Stata (Version 18.0, StataCorp, College Station, Texas). We used two-sided tests with a significance of P<0.05. The study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.
RESULTS
This analysis includes data on 20,458 pregnancy tests reported at the first questionnaire post-enrollment, among 6,569 unique participants. 79.4% of participants reported pregnancy testing behavior in reference to day of expected menses and 20.6% of participants reported pregnancy test behavior as number of days from LMP. Median day of first home pregnancy test was three days before expected period. 40.7% of participants reported that their first pregnancy test was four or more days before day of expected period and 26.4% of participants reported their first test was on the day of their expected period or later (Figure 1A). Within the window queried, 33.5% of participants reported testing only on one day, whereas 8.6% reported testing every day (median frequency: two testing days, IQR: 1, 5; Figure 1B).
Figure 1.
Distribution of timing of first pregnancy test (A), frequency of pregnancy testing (B), and testing behavior by both timing of first test and frequency of testing (C) between four days before expected period through four days after expected period among a prospective cohort of pregnancy planners (N=6,569). In C, timing of first test was categorized as very early (four or more days before expected period), early (one to three days before expected period), or expected or late (day of expected period or later) and frequency of pregnancy testing was categorized as infrequent (one testing day), moderate (2–3 testing days), or frequent (four or more testing days).
When considering both timing of first test and frequency of testing together, the most commonly observed pattern was initiating testing very early (four or more days before expected menses) and testing very frequently (four or more days in the window; 22.3% of participants). The next most commonly observed pattern was using a single test on or after the day of expected menstruation (16.6%). The third most commonly observed behavior was testing early (one to three days before expected period) with moderate frequency (one to three testing days; 14.6% of participants; Figure 1C).
We observed a range of pregnancy testing intensity, with some participants testing only once and others testing every day, even after a positive test (Figure 2). 73.5% of participants reported taking a pregnancy test before the day of expected menses and 66.5% of participants tested for pregnancy on multiple days. Among participants who detected a pregnancy, 66.3% of participants reported taking another test after their first positive test. 2.0% of participants reported having a positive test followed by a negative test.
Figure 2.
Distribution of cumulative pregnancy testing history within window centered on date of expected menses by day of test in a cohort of pregnancy planners (N=6,569).
Timing of first pregnancy test and frequency of pregnancy testing did not vary appreciably by maternal age, education, race and ethnicity, or income in bivariate analyses. Participants who were multiparous and participants who had a history of prior loss were more likely to be very early and frequent testers (Table 1). Participants with two or more prior live births were more likely to be very early, frequent testers compared with nulliparous participants after adjusting for maternal age and miscarriage history (RR 1.44, 95% CI 1.26, 1.66). Participants with a history of miscarriage were also more likely to be very early, frequent testers compared with their counterparts after adjusting for maternal age and parity (RR 1.26, 95% CI 1.14, 1.39). Receipt of free home pregnancy test from the study was not associated with increased frequency in reported home pregnancy testing (quantile regression coefficient 0, 95% CI −0.14, 0.14) or early detection of pregnancy (quantile regression coefficient 0, 95% CI, −0.43, 0.43), but was associated with the median timing of initial pregnancy test being taken one day earlier (quantile regression coefficient −1, 95% −1.24, −0.76).
Table 1.
Maternal Characteristics and Pregnancy Testing Behavior (N=6,569)
| Timing of First Pregnancy Test |
Frequency of Pregnancy Testing |
|||||
|---|---|---|---|---|---|---|
| Characteristic | Count | Very Early Tester* |
Frequent Tester † |
|||
| N | % | Median (IQR) |
Unadjusted RR (95% CI) |
Median (IQR) |
Unadjusted RR (95% CI) |
|
| Overall | ||||||
| Maternal Age | ||||||
| <25 | 337 | 5.1 | −3 (≤−4, 0) | 1.07 (0.94–1.22) | 3 (1, 5) | 1.13 (0.97–1.32) |
| 25-29 | 2,211 | 33.7 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| 30-34 | 3,043 | 46.3 | −3 (≤−4, 0) | 0.96 (0.90–1.03) | 2 (1, 4) | 0.98 (0.91–1.07) |
| ≥35 | 978 | 14.9 | −2 (≤−4, 0) | 0.86 (0.78–0.95) | 2 (1, 4) | 0.88 (0.79–0.99) |
| Maternal Education | ||||||
| Some college or less | 1,057 | 16.1 | −3 (≤−4, 0) | Referent | 2 (1, 5) | Referent |
| College | 2,180 | 33.2 | −3 (≤−4, 0) | 0.92 (0.85–1.00) | 2 (1, 4) | 0.91 (0.83–1.02) |
| Graduate school | 3,332 | 50.7 | −3 (≤−4, 0) | 0.90 (0.83–0.97) | 2 (1, 4) | 0.88 (0.80–0.97) |
| Income (USD) | ||||||
| <$50,000 | 674 | 10.3 | −3 (≤−4, 0) | 0.97 (0.87–1.08) | 2 (1, 4) | 1.02 (0.89–1.16) |
| $50,000-$99,999 | 1,856 | 28.3 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| $100,000-$149,999 | 1,896 | 28.9 | −3 (≤−4, 0) | 1.00 (0.93–1.09) | 2 (1, 4) | 1.06 (0.96–1.16) |
| ≥$150,000 | 2,016 | 30.7 | −3 (≤−4, −1) | 0.99 (0.91–1.07) | 2 (1, 4) | 1.04 (0.94–1.14) |
| Missing | 127 | 1.9 | ||||
| Race and Ethnicity | ||||||
| Asian, Non-Hispanic | 144 | 2.2 | −2 (≤−4, 0) | 0.92 (0.74–1.14) | 2 (1, 4) | 0.91 (0.70–1.19) |
| Black, Non-Hispanic | 140 | 2.1 | −2 (≤−4, 0) | 0.97 (0.78–1.19) | 2 (1, 4) | 0.92 (0.70–1.19) |
| Hispanic | 422 | 6.4 | −3 (≤−4, −1) | 1.08 (0.97–1.22) | 3 (1, 5) | 1.09 (0.95–1.26) |
| Mixed race | 227 | 3.5 | −3 (≤−4, 0) | 0.96 (0.81–1.13) | 2 (1, 5) | 1.06 (0.88–1.30) |
| Native American | 11 | 0.2 | −1 (≤−4, 0) | 1.11 (0.58–2.14) | 1 (1, 5) | 1.17 (0.53–2.55) |
| White, Non-Hispanic | 5,581 | 85.0 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| None of the above | 44 | 0.7 | −3 (≤−4, 0) | 0.67 (0.41–1.09) | 2 (1, 4) | 0.95 (0.60–1.50) |
| BMI | ||||||
| <18.5 | 106 | 1.6 | −3 (≤−4, −1) | 1.00 (0.78–1.28) | 2 (1, 4) | 0.90 (0.65–1.23) |
| 18.5-24.9 | 3,046 | 46.4 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| 25-29.9 | 1,655 | 25.2 | −3 (≤−4, 0) | 1.09 (1.02–1.17) | 2 (1, 4) | 1.02 (0.92–1.11) |
| ≥30 | 1,756 | 26.7 | −3 (≤−4, 0) | 1.11 (1.03–1.19) | 2 (1, 4) | 1.09 (1.00–1.18) |
| Missing | 6 | 0.1 | ||||
| Gravidity | ||||||
| 0 | 3,268 | 49.8 | −2 (≤−4,0) | Referent | 2 (1,4) | Referent |
| 1 | 1,657 | 25.2 | −3 (≤−4, −1) | 1.10 (1.02, 1.18) | 2 (1,4) | 1.11 (1.02, 1.22) |
| ≥2 | 1,644 | 25.0 | −3 (≤−4, −1) | 1.17 (1.09, 1.25) | 2 (1,5) | 1.26 (1.16, 1.37) |
| Parity | ||||||
| 0 | 4,393 | 66.9 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| 1 | 1,537 | 23.4 | −3 (≤−4, −1) | 1.04 (0.97–1.12) | 2 (1, 4) | 1.13 (1.03–1.22) |
| ≥2 | 639 | 9.7 | −3 (≤−4, −1) | 1.24 (1.14–1.35) | 2 (1, 5) | 1.23 (1.10–1.38) |
| History of miscarriage | ||||||
| Yes | 1,737 | 26.4 | −3 (≤−4, −1) | 1.12 (1.05–1.19) | 3 (1, 5) | 1.21 (1.12–1.31) |
| No | 4,832 | 73.6 | −3 (≤−4, 0) | Referent | 2 (1, 4) | Referent |
| Cycle Regularity | ||||||
| Regular | 5,030 | 76.6 | −3 (≤−4, 0) | Referent | 2 (1,4) | Referent |
| Irregular | 1,291 | 19.7 | −3 (≤−4, 0) | 1.14 (1.07–1.23) | 2 (1,4) | 1.09 (0.99–1.18) |
| Missing | 248 | 3.8 | ||||
Abbreviations: CI, confidence interval; IQR, inter-quartile range; RR, risk ratio
Participants were considered very early testers if they reported a first pregnancy test four or more days before the expected period
Participants were considered frequent testers if they reported four or more testing within the window
Participants who initiated testing very early were also more than three times likely to have an initial positive test followed by a negative test, likely indicating increased detection of very early pregnancy loss (RR 3.80, 95% CI 2.12, 6.80, adjusted for cycle length, history of miscarriage, parity, and maternal age). Estimates were unchanged when receipt of free home pregnancy test was included in the model.
Among participants who detected a pregnancy, the median timing of first positive pregnancy test was one day before expected period (IQR: three days prior, two days after expected period). In reference to the LMP, the median first positive pregnancy test was at 28 days’ gestation (IQR: 25, 31; Figure 3).
Figure 3.
Distribution of timing of first positive pregnancy test among study participants who detected a pregnancy by first follow-up questionnaire in reference to date of expected period (A) or last menstrual period (LMP) (B) in a prospective cohort of pregnancy planners (N=3,055). The solid line represents median value, and the dotted line represents interquartile range.
We did not detect meaningful differences in timing of pregnancy recognition by maternal age, race and ethnicity, maternal education, income, or BMI. Individuals with irregular cycles detected pregnancy slightly later than participants with regular cycles (30 versus 28 days’ gestation). Individuals with shorter cycles reported a first positive pregnancy test earlier than those with longer cycles when timing of detection was measured in reference to LMP (Figure 4). Participants who reported very early or early testing detected pregnancy later than those who did not test until on or after their day of expected menses (median 26 and 27 days’ gestation versus 31 days’ gestation). Participants who tested frequently also detected pregnancy slightly earlier than those who reported moderate or infrequent testing (median 27 versus 29 days’ gestation).
Figure 4.
Timing of first positive pregnancy test from last menstrual period by maternal characteristic. Unadjusted marginal estimates for median and 95% CIs from quantile regression models (N=3,055). *Includes participants who reported Asian, Native American, or mixed raced. IQR, interquartile range; USD, U.S. dollars; BMI, body mass index.
Among participants who ultimately detected pregnancy, participants who initiated testing very early (four or more days prior to expected menses) detected pregnancy on average four days earlier, after adjusting for cycle length, than those who first tested for pregnancy on the day of expected menses (95% CI 3.6, 4.4). Participant who started testing very early were less likely to only have positive tests within the cycle compared to those who waited until three days prior to expected menses or later to take their first test (51.3% versus 83.1%; RR 0.62, 95% CI 0.57, 0.66, adjusted for menstrual cycle length). Those who reported very early testing were over five times more likely to have a negative test before a positive test compared with those who waited until day of expected menses to test (RR 5.89, 95% CI 4.73, 7.33, adjusted for menstrual cycle length). Estimates were unchanged when receipt of free home pregnancy test was included in the models.
DISCUSSION
In this analysis of 6,569 individuals planning a pregnancy, patterns of participant-reported home pregnancy test use widely varied. Approaches to pregnancy testing ranged from daily testing starting well before expected menses to performing a single test on day of expected menses. Despite striking differences in pregnancy testing approaches, variations in timing of pregnancy detection were modest. 40.7% of participants reported testing for pregnancy four or more days before expected period. These participants detected pregnancy slightly earlier, but were five times more likely to have a negative test before having a positive test. Early testing was also associated with an over three-times greater risk of detecting a very early loss. The high prevalence of early testing in this cohort highlights the importance in understanding the implications of very early testing.
Very early pregnancy testing was common in this cohort of pregnancy planners. While very early pregnancy testing was linked to slightly earlier pregnancy detection, it was also associated with several undesirable outcomes. The sensitivity of home pregnancy tests increases with each day after fertilization.8, 9 Thus, very early testing increases the risk of having a negative test even when a pregnancy is present. Participants who initiated testing very early were over five times more likely to have their first test be negative than those who waited until day of expected period to start testing. This experience may contribute to stress, is more costly, and requires additional testing for accurate determination of pregnancy status. Further, those who tested very early were over three times as likely to detect a very early loss (a positive test followed by a negative test) compared with those who initiated testing on day of expected menses. For some, this experience brings the grief of a pregnancy loss that would otherwise have gone undetected if home pregnancy testing had been initiated later.
Given the wide variation in pregnancy testing frequency and timing of initiation in this cohort, we sought to understand drivers of pregnancy testing behavior. Multiparous participants were more likely to test earlier and more frequently than nulliparous participants. Earlier testing may indicate earlier recognition of pregnancy symptoms in parous patients based on recalled experience. Participants with a history of miscarriage were more likely to test earlier and more frequently than their counterparts. This behavior may reflect increased vigilance or concern in the setting of a prior miscarriage. Participants with a history of irregular cycles reported initiating test earlier compared with those with regular cycles, which may signify uncertainty about timing of expected menstruation, and therefore uncertainty about when to start testing. Despite earlier testing, participants with irregular cycles detected pregnancy later than those with regular cycles, evidencing how menstrual irregularity can be a biological obstacle to prompt pregnancy awareness.10 Although use of a home pregnancy test varied strikingly between participants, the distribution of gestational age at which cohort participants detected pregnancy was relatively narrow and did not meaningfully differ by maternal attributes.
These data provide a representation of the variety in home pregnancy testing behavior among individuals planning to conceive. As participants were not provided instructions on how to test for pregnancy, this study provides insight into real-world practices. However, since participants reported behavior bimonthly, there is opportunity for recall bias. Given that only one response per day was permitted, true test frequency may be greater than measured if participants tested for pregnancy multiple times per day. The receipt of free tests by a subset of participants did not alter timing of pregnancy detection or testing frequency. However, individuals who received a free test reported initiating testing a day earlier. Additionally, day-specific data on home pregnancy testing was limited to four days before expected menstrual cycle through four days after. This window was chosen to reflect when modern pregnancy tests claim to be reliable.8, 11 However, since over forty percent of participants in this cohort reported initiating testing before this time point, more granular information about pregnancy testing practices earlier in the cycle would be valuable.
This study was conducted in high-income countries (United States and Canada) among individuals actively trying to conceive. Studies of timing of pregnancy detection in other settings show non-planners detect pregnancy later and have greater variability in timing of pregnancy recognition compared with this cohort.2, 12, 13 Studies of pregnancy detection among individuals who are seeking induced abortion care demonstrate even more extreme delays and variation in pregnancy detection, which can have meaningful implications for access to care.3, 14-16 In the setting of planned pregnancies, our findings indicate that timing of pregnancy detection is largely uniform and within several days of expected menses, even when home pregnancy testing practices varied.
In this analysis, we demonstrated the wide variety of home pregnancy testing behavior among pregnancy planners. The lack of guidance in best practices for home pregnancy testing beyond manufacturers’ instructions leaves each person making their own decisions about when and how to test. For those desiring a pregnancy, that first positive test marks the beginning of a hope realized and as home tests become increasingly more sensitive, the capability to detect pregnancy early will only increase. However, this capacity is coupled with the increased chance of detecting very early losses, which carry their own ramifications for patients and clinicians. Understanding variations in approaches to home pregnancy testing and their implications is the first step towards being able to provide evidence-based recommendations about best practices.
Supplementary Material
Acknowledgments:
The authors thank the participants of Pregnancy Study Online (PRESTO) for their commitment to the study and the PRESTO investigators and staff who devoted their time and expertise to the success of the cohort.
Footnotes
The other authors did not report any potential conflicts of interest.
Each author has confirmed compliance with the journal’s requirements for authorship.
Meeting Presentation: Portions of this analysis were presented at the Society for Reproductive Investigation Annual Meeting in Charlotte, NC, March 25-29, 2025, the Society of Pediatric and Perinatal Epidemiologic Research Annual Meeting in Boston, MA, June 9-10 2025, and the Society of Epidemiologic Research 58th Annual Meeting in Boston, MA, June 10-13 2025.
Financial Disclosure:
In the last three years, Lauren A. Wise, ScD has received in-kind donations from Swiss Precision Diagnostics (home pregnancy tests) and Kindara.com (fertility app) for primary data collection in PRESTO. She has also served as a paid consultant for AbbVie, Inc. and the Gates Foundation.
Funding Sources:
This research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health (R01HD086742), the National Institutes of Health through Building Interdisciplinary Research Careers in Women's Health (grant number K12AR084232), and the National Science Foundation (1914792).
REFERENCES
- 1.Pryor J, Patrick SW, Sundermann AC, Wu P, Hartmann KE. Pregnancy intention and maternal alcohol consumption. Obstet Gynecol 2017;129:727–33. doi: 10.1097/AOG.0000000000001933 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dott M, Rasmussen SA, Hogue CJ, Reefhuis J, National Birth Defects Prevention Study. Association between pregnancy intention and reproductive-health related behaviors before and after pregnancy recognition, national birth defects prevention study, 1997-2002. Matern Child Health J 2010;14:373–81. doi: 10.1007/s10995-009-0458-1 [DOI] [PubMed] [Google Scholar]
- 3.Finer LB, Frohwirth LF, Dauphinee LA, Singh S, Moore AM. Timing of steps and reasons for delays in obtaining abortions in the United States. Contraception 2006;74:334–44. doi: 10.1016/j.contraception.2006.04.010 [DOI] [PubMed] [Google Scholar]
- 4.Ralph LJ, Foster DG, Barar R, Rocca CH. Home pregnancy test use and timing of pregnancy confirmation among people seeking health care. Contraception 2022;107:10–6. doi: 10.1016/j.contraception.2021.10.006 [DOI] [PubMed] [Google Scholar]
- 5.Ayoola AB, Stommel M, Nettleman MD. Late recognition of pregnancy as a predictor of adverse birth outcomes. Am J Obstet Gynecol 2009;201:156.e1–6. doi: 10.1016/j.ajog.2009.05.011 [DOI] [PubMed] [Google Scholar]
- 6.Wise LA, Rothman KJ, Mikkelsen EM, Stanford JB, Wesselink AK, McKinnon C, et al. Design and conduct of an internet-based preconception cohort study in North America: Pregnancy Study Online. Paediatr Perinat Epidemiol 2015;29:360–71. doi: 10.1111/ppe.12201 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wise LA, Wang TR, Willis SK, Wesselink AK, Rothman KJ, Hatch EE. Effect of a Home pregnancy test intervention on cohort retention and pregnancy detection: A randomized trial. Am J Epidemiol 2020;189:773–8. doi: 10.1093/aje/kwaa027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gnoth C, Johnson S. Strips of Hope: Accuracy of Home Pregnancy Tests and New Developments. Geburtshilfe Frauenheilkunde 2014;74:661–9. doi: 10.1055/s-0034-1368589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nepomnaschy PA, Weinberg CR, Wilcox AJ, Baird DD. Urinary hCG patterns during the week following implantation. Hum Reprod 2008;23:271–7. doi: 10.1093/humrep/dem397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nobles J, Cannon L, Wilcox AJ. Menstrual irregularity as a biological limit to early pregnancy awareness. Proc Natl Acad Sci U S A 2022;119:e2113762118. doi: 10.1073/pnas.2113762118 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Clearblue Pregnancy Tests: 99% accurate from day of expected period. Professional Series: Pregnancy. Geneva, Switzerland: Swiss Precision Diagnostics, 2018. Accessed July 16, 2025. https://www.clearblue.com/sites/default/files/hcp_cb6_cb11_cb12.pdf. [Google Scholar]
- 12.Ayoola AB. Late recognition of unintended pregnancies. Public Health Nurs 2015;32:462–70. doi: 10.1111/phn.12182 [DOI] [PubMed] [Google Scholar]
- 13.Branum AM, Ahrens KA. Trends in timing of pregnancy awareness among us women. Matern Child Health J 2017;21:715–26. doi: 10.1007/s10995-016-2155-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Foster DG, Gould H, Biggs MA. Timing of pregnancy discovery among women seeking abortion. Contraception 2021;104:642–7. doi: 10.1016/j.contraception.2021.07.110 [DOI] [PubMed] [Google Scholar]
- 15.McCarthy M, Upadhyay U, Biggs MA, Anthony R, Holl J, Roberts SCM. Predictors of timing of pregnancy discovery. Contraception 2018;97:303–8. doi: 10.1016/j.contraception.2017.12.001 [DOI] [PubMed] [Google Scholar]
- 16.Swanson M, Karasek D, Drey E, Foster DG. Delayed pregnancy testing and second-trimester abortion: can public health interventions assist with earlier detection of unintended pregnancy? Contraception 2014;89:400–6. doi: 10.1016/j.contraception.2013.12.008 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.