MD IntelligenceClinical Tools
Due Date & Pregnancy Calendar →
Pregnancy, day by day

Will My Pregnancy Continue?

The chance of a miscarriage from today until 20 weeks, and of a stillbirth after 20 weeks, shown on your own dates. You can add your age, any miscarriages just before this pregnancy, and a heartbeat seen on ultrasound. Every number starts from a published study or from US national records, and each source is named on this page. Where the tool combines sources, it says so.

Main sources: Ammon Avalos L, Galindo C, Li DK. Birth Defects Res A Clin Mol Teratol. 2012;94(6):417-23 (miscarriage by week). Magnus MC, et al. BMJ. 2019;364:l869 (age and previous miscarriages). Tong S, et al. Obstet Gynecol. 2008;111(3):710-4 (heartbeat). National Center for Health Statistics, 2024 fetal death and birth files (stillbirth). Full list of publications.
Your due date is 40 weeks 0 days. If you do not know it, use the Due Date & Pregnancy Calendar first.
Your age now, in years. Leave it empty to see the numbers for everyone.
Miscarriages in a row, just before this pregnancy
"None" also means this is your first pregnancy, or your last pregnancy ended in a birth.
Has an ultrasound between 6 and 12 weeks shown the baby's heartbeat?

Until 20 weeks: the chance of a miscarriage from each day on

Everyone (Norway registry)Higher estimate: women enrolled at a positive test (Li)You

From 20 weeks: the chance of a stillbirth from each day on

All ages, US 2024Your age groupFewer than 20 stillbirths behind this point

When to call your doctor or midwife now

Call about heavy bleeding, bleeding with cramping pain, or pain on one side of your belly. In the second half of pregnancy, call if the baby is moving less than usual, if fluid leaks, or if you have bleeding. These graphs cannot tell you what is happening in your own pregnancy today.

Read this before you use the graphs

What changes the risk, and what does not

Used in your numbers

FactorWhat the study foundSource
Your ageMiscarriage by age: under 20, 15.8%; 20 to 24, 11.3%; 25 to 29, 9.8%; 30 to 34, 10.8%; 35 to 39, 16.7%; 40 to 44, 32.2%; 45 and over, 53.6%. Norway, 421,201 pregnancies, 2009 to 2013.Magnus 2019, Table 1 (ref 4). Registry.
Miscarriages in a row just before this pregnancy11.6% with no previous pregnancy; 19.8% after 1 miscarriage (odds ratio 1.54, 95% CI 1.48 to 1.60); 27.7% after 2 (2.21, 2.03 to 2.41); 41.9% after 3 or more (3.97, 3.29 to 4.78). Odds ratios adjusted for age.Magnus 2019, Table 3 (ref 4). Registry.
Heartbeat seen on ultrasound, no bleeding or painMiscarriage before 20 weeks by the week of that ultrasound: 6 weeks, 3 of 32 (9.4%); 7 weeks, 3 of 72 (4.2%); 8 weeks, 3 of 195 (1.5%); 9 weeks, 1 of 210 (0.5%); 10 weeks, 1 of 151 (0.7%); 11 weeks, 0 of 36. Dated by crown-rump length, women without bleeding or pain. The tool combines 9 to 11 weeks (2 of 397, 0.5%) because each of those weeks has too few losses to differ. After repeated miscarriages a heartbeat reassures less: 18% loss compared with 2% in matched women (Hyer 2004, ref 34), so the tool does not use the heartbeat after 2 or more miscarriages in a row.Tong 2008, Table 2 (ref 5). Prospective cohort, Australia.
Your age, after 20 weeksStillbirth from 20 weeks on, per 1,000 still pregnant: 4.65 at 25 to 29 years up to 7.97 at 40 to 44 and 10.1 at 45 and over (all ages 5.21). US singletons, 2024.NCHS 2024 fetal death and birth files (refs 6, 7). Counted by MD Intelligence.

Also linked to a higher risk, but not added to your number

These studies report a relative effect, such as "1.23 times the odds", measured in groups different from the ones above. Stacking them onto your number would overstate it. Talk with your doctor or midwife about any that apply to you.

FactorWhat the study foundSource
Heavy bleeding (as heavy as a heavy period for at least a day)23.7% miscarried (23 of 97) compared with 11.6% with no bleeding (381 of 3,285); adjusted odds ratio 2.84 (1.82 to 4.43)Hasan 2009 (ref 8). Prospective US cohort.
Blood collection beside the pregnancy (subchorionic hematoma) on ultrasound17.6% compared with 8.9%; odds ratio 2.18 (1.29 to 3.68)Tuuli 2011 (ref 9). Meta-analysis.
SmokingRelative risk 1.23 (1.16 to 1.30)Pineles 2014 (ref 10). Meta-analysis.
AlcoholAny alcohol: odds ratio 1.19 (1.12 to 1.28)Sundermann 2019 (ref 11). Meta-analysis.
Caffeine, 350 mg a day or more350 to 699 mg: relative risk 1.40 (1.16 to 1.68); 700 mg or more: 1.72 (1.40 to 2.13); 150 to 349 mg: 1.16 (0.94 to 1.41), not clearly raisedChen 2016 (ref 12). Meta-analysis.
Obesity before pregnancyRelative risk 1.21 (1.15 to 1.27)Balsells 2016 (ref 13). Meta-analysis.
Father aged 40 or older40 to 44: 1.23 (1.06 to 1.43); 45 and over: 1.43 (1.13 to 1.81); no clear effect under 40du Fossé 2020 (ref 14). Meta-analysis.
Thyroid antibodiesOdds ratio 3.90 (2.48 to 6.12) in cohort studies. Levothyroxine does not help women whose thyroid function is normal.Thangaratinam 2011 (ref 15); Coomarasamy 2021 (ref 31).
Antiphospholipid antibodiesLinked to miscarriage. In antiphospholipid syndrome with repeated miscarriage, aspirin plus heparin raised live births compared with aspirin alone: relative risk 1.27 (1.09 to 1.49), low-certainty evidence.Coomarasamy 2021 (ref 31). Review of trials.
Septate uterusFirst-trimester loss, relative risk 2.89 (2.02 to 4.14)Chan 2011 (ref 16). Systematic review.
Previous stillbirth17.5% miscarried; odds ratio 1.30 (1.11 to 1.53)Magnus 2019, Table 2 (ref 4).
Type 1 diabetes with poor blood sugar controlOverall, women with type 1 diabetes had the same rate as women without diabetes (16.1% and 16.2%), but the risk rose with poor control early in pregnancyMills 1988 (ref 21); Rosenn 1994 (ref 33).
Pregnancy by IVFFirst-trimester loss after a pregnancy sac was seen, by age at egg retrieval: under 30, 9.8%; 30 to 34, 11.2%; 35 to 37, 14.7%; 38 to 40, 22.0%; over 40, 36.8%. A different starting point from the graph above.Hipp 2016, Table 1 (ref 17). US national IVF data.

Not shown to raise the risk

FactorWhat the study foundSource
A previous live birth12.2% compared with 11.6% in a first pregnancy, but after allowing for age the odds were slightly lower: odds ratio 0.91 (0.89 to 0.94)Magnus 2019, Table 2 (ref 4)
Spotting or light bleeding9.2% with spotting and 11.7% with light bleeding, compared with 11.6% with noneHasan 2009 (ref 8)
ExerciseOdds ratio 0.88 (0.63 to 1.21)Davenport 2019 (ref 18)
FibroidsAdjusted hazard ratio 0.83 (0.63 to 1.08)Hartmann 2017 (ref 19)
Getting pregnant within 6 months of a miscarriageLower, not higher: relative risk 0.82 (0.78 to 0.86)Kangatharan 2017 (ref 20)
Caffeine under 150 mg a day50 to 149 mg: relative risk 1.02 (0.85 to 1.24)Chen 2016 (ref 12)
Folic acid9.0% with and 9.3% without (it is still recommended to prevent birth defects)Gindler 2001 (ref 22)
Sex during pregnancyNo study of miscarriage risk was found, and no evidence of harmNo source

Nausea: in women with one or two earlier losses, those with nausea had fewer losses (hazard ratio 0.50; with vomiting 0.25) (ref 23). Many healthy pregnancies have no nausea at all, so having none is not a warning sign.

A treatment that helps some women: for women with bleeding who also had at least one miscarriage before, vaginal progesterone raised live births from 70% to 75% (relative risk 1.09); after three or more miscarriages, from 57% to 72% (ref 24). It did not help bleeding alone. After repeated miscarriages, levothyroxine may be considered when TSH is above 4.0 mIU/L (low-quality evidence), and aspirin plus heparin when antiphospholipid antibodies are found (ref 31). Ask your doctor or midwife.

After 20 weeks: what is linked to stillbirth

Your age is used in the stillbirth graph, counted from US 2024 data. The factors below come from an umbrella review of 69 systematic reviews (ref 32). They are relative effects, often unadjusted, from studies with different definitions of stillbirth, so they are not added to your number.

FactorPooled association with stillbirth (95% CI)Strength of evidence (ref 32)
A previous stillbirthOdds ratio 4.83 (3.77 to 6.18)Suggestive
Sickle cell diseaseRelative risk 3.99 (2.63 to 6.04)Highly suggestive
Diabetes before pregnancyOdds ratio 3.52 (3.19 to 3.88)Suggestive
High blood pressure before pregnancy2.58 (2.13 to 3.13)Highly convincing
A previous small-for-gestational-age babyOdds ratio 1.85 (1.42 to 2.40)Highly suggestive
BMI over 30, compared with under 25Odds ratio 1.81 (1.69 to 1.93)Highly convincing
SmokingRelative risk 1.46 (1.38 to 1.54)Highly suggestive
A first birth1.42 (1.33 to 1.51)Highly convincing
A previous cesarean deliveryOdds ratio 1.27 (1.15 to 1.40)Suggestive
Caffeine and alcoholNot consistently associated with stillbirthAs stated by the review

The same review found that no single blood test or ultrasound predicts stillbirth well. At term, in pooled cohorts of 15 million pregnancies, the risk of stillbirth in the week from 41 weeks was 1.64 times the risk in the week from 40 weeks (1.51 to 1.77), about 1 extra stillbirth for every 1,449 pregnancies that continued (ref 30).

The numbers, week by week

WeekYour datesMiscarriages per 1,000 women that week (Li)From the start of the week: Li cohortFrom the start of the week: everyone (registry)From the start of the week: you

WeekYour datesStillbirthsLive birthsFrom the start of the week, per 1,000 still pregnant

How these numbers were made

Miscarriage, 4 to 20 weeks. The shape of the curve comes from the weekly miscarriage rates in Table 2 of Ammon Avalos and colleagues (ref 1), who pooled four US cohorts with life-table methods and dated pregnancy from the last menstrual period. The tool uses the cohort of Li and colleagues (ref 3): 1,055 women in Northern California, 1996 to 1998, enrolled at their positive pregnancy test, median age 31. The review counts women "before 20 complete weeks" through "week 20", so the tool reads its week 5 as 4 weeks 0 days to 4 weeks 6 days and its week 20 as 19 weeks 0 days to 19 weeks 6 days. Combined week by week, the Li rates give a 22.1% risk from 4 weeks and 17.8% from 6 weeks. The review's authors judged this cohort the most complete for early losses because it enrolled women earliest; the three older cohorts give 10.9% to 13.5%. Each week's rate is spread evenly over its 7 days.

The level of the curve: everyone and you. The registry in Norway (ref 4) counted losses from about 6 weeks: 12.8% of all recognized pregnancies, with a figure for each age group and for 1, 2 or 3 or more miscarriages in a row. The tool sets the curve at 6 weeks 0 days to the registry figure that fits you and lets it rise before and fall after with the shape of the Li curve, on the odds scale so that it cannot pass 100%. The part before 6 weeks is therefore an extension of the registry figure, not a registry count. With only your age, it uses your age group's figure; with only your history, the history figure; with both, it combines the age figure with the age-adjusted odds ratio for your history, which is a calculation, not a published number. The Li curve stays on the graph as a higher estimate, because studies that enroll women at a positive test count early losses that a registry misses.

Heartbeat. If an ultrasound between 6 weeks 2 days and 11 weeks 6 days showed a heartbeat with no bleeding or pain, the tool compares the figure Tong and colleagues measured for that week (ref 5) with the Li cohort on the same day, since both followed women from early pregnancy: 6 weeks 9.4%, 7 weeks 4.2%, 8 weeks 1.5%, and 9 to 11 weeks combined 0.5% (2 of 397). Your own curve is then lowered by that ratio from the scan date on. A heartbeat never raises the number. This step is a calculation. After 2 or more miscarriages in a row the tool does not apply it, because women with repeated miscarriage and a heartbeat at 6 to 8 weeks still lost 18% of pregnancies, compared with 2% of matched women (ref 34; abstract figures). Tong's numbers are small: the 6-week figure rests on 3 losses in 32 women (95% CI 0 to 19.5%). Scans after 11 weeks 6 days are not used, because Tong did not study them.

Stillbirth, 20 weeks on. From the National Center for Health Statistics 2024 fetal death file (refs 6, 7) and 2024 birth file (ref 25): single babies, women living in the US, 20 to 47 completed weeks by the obstetric estimate. That gives 18,399 stillbirths and 3,511,757 live births; 122 stillbirths and 4,371 live births with no week recorded are left out, as NCHS does. For each week, the risk is the stillbirths from that week on divided by all pregnancies still going on at the start of it, live births plus stillbirths (the fetuses-at-risk method, ref 26). By US definition, stillbirths do not include induced terminations. Reporting at 20 to 23 weeks varies by state, so those weeks are the least reliable. Your age group is counted the same way from the same files. In these counts the risk for each pregnancy still continuing rises again after 39 weeks, because the weekly risk among the pregnancies that continue goes up. ACOG recommends fetal testing from 41 weeks and delivery by 42 weeks 6 days (refs 27, 28).

How these compare with other data. A US national survey of women's own reports, 1990 to 2011, found that 19.7% of pregnancies ended in a loss, counting stillbirth and ectopic pregnancy as well as miscarriage; the risk was 18.0% at age 20 to 24 and 25.1% at 30 to 44 (ref 29). It groups all women 30 to 44 together, so it cannot replace the Norwegian age groups, but it shows the same rise with age. For stillbirth at 37 to 41 weeks, the US 2024 weekly risks are lower than the pooled older cohorts in ref 30, and both rise week by week.

What this cannot tell you. These numbers do not include twins, ectopic pregnancy or molar pregnancy. They do not account for your health conditions or for anything seen on your own ultrasounds beyond a heartbeat. The miscarriage and stillbirth figures come from different times and places.

Publications used in this tool (34) →