Bleeding After a Positive Test: When to Go In, What a Scan Shows

If the bleeding is light and there is no pain, the right step is to arrange to be seen rather than to rush anywhere - and not to treat it as the answer about how the pregnancy will end, because in the largest cohorts it is not. If there is pain, faintness, shoulder-tip pain, or bleeding heavy enough to soak pads, that is an emergency and it needs looking at now, not in the morning.

The rest of this page is what is actually known, which is narrower than most pages imply and more useful than a colour chart.

Any of these alongside bleeding means going in, or ringing your clinic or a hospital straight away:

• Pain that is severe, or worse than any period pain you have had
• Pain that is on one side, low in the abdomen or pelvis
• Feeling faint or dizzy, or fainting
• Pain at the tip of the shoulder
• Bleeding that soaks through pads, or large clots
• Pain or pressure on opening the bowels
• A racing pulse, or feeling cold and clammy

These are taken from NICE NG126 1.4.3 and 1.4.4, which set out the symptoms and signs of ectopic pregnancy. NG126 1.4.1 refers women directly to emergency care if they are haemodynamically unstable or there is significant concern about the degree of pain or bleeding. NG126 1.4.7 asks for immediate referral to an early pregnancy assessment service where a woman has a positive test plus pain and abdominal, pelvic or cervical motion tenderness.

At or after six weeks, the threshold is lower again: NG126 1.4.9 advises referral for assessment for women with bleeding who have pain, or a pregnancy of six weeks or more, or a pregnancy of uncertain dates.

Common enough that it is not, by itself, a signal. Hasan 2010 followed 4,539 women through the first trimester in a community-based study. Bleeding was reported by 1,207 of them - roughly a quarter of the 4,539 - across 1,656 separate episodes. Only 8 in 100 of the women who bled described any episode as heavy. Most episodes lasted under three days, and most fell between gestational weeks 5 and 8.

A tighter, earlier study points the same way with a smaller number. Harville 2003 followed 221 women with daily urine samples; of the 151 who became clinically pregnant, 14 recorded any bleeding in the first eight weeks, and 12 of those 14 pregnancies ended in a live birth.

The two figures differ because they count different windows and ask in different ways, which is worth knowing before trusting any single percentage on this subject. If the bleeding happened before you had a positive test, the question of whether it was a period or early pregnancy is a different one, and the answer there is that a test settles it and the appearance of the blood does not.

For light bleeding, the honest answer from the best data is no. In the same cohort of 4,539, miscarriage occurred in 12% of the women who bled and 13% of the women who did not - a difference in the wrong direction to be a warning sign.

Hasan 2009 tested it directly on 4,510 of those women, with bleeding in the four days before a loss excluded so that the bleeding of a miscarriage already happening was not counted as a prediction of one. Any bleeding gave an odds ratio for miscarriage of 1.1, with a confidence interval of 0.9 to 1.3 that includes no effect. Heavy bleeding was a different finding: among the 97 women who reported it, the odds ratio was 3.0, from 1.9 to 4.6, and most of that sat in the women whose heavy bleeding came with pain.

There is a real tension in the literature here and it should not be smoothed over. The PRISM trial's own background states that bleeding in early pregnancy is strongly associated with pregnancy loss. PRISM recruited women who had presented at hospital with bleeding; Hasan followed a community cohort in which 92% of those who bled never had a heavy episode. Both can be true of their own populations. What follows for a reader is that heaviness and pain are the features that carry information, and that the strength of the association depends on which of those two groups the bleeding belongs to.

NICE NG126 1.5.1 sets the first step: a transvaginal ultrasound to find where the pregnancy is and whether there is a fetal pole and a heartbeat. A scan through the abdomen is offered instead if a transvaginal scan is not acceptable to you, with its limits explained, or if there is an enlarged uterus, a fibroid or an ovarian cyst.

The part nobody prepares you for is that one scan often cannot decide. NG126 1.6.1 says in terms that a diagnosis of miscarriage made from a single scan cannot be relied on as fully accurate, and that there is a small chance it is wrong, particularly at very early gestations. Where the crown-rump length measures under 7.0 mm with no visible heartbeat, 1.6.3 requires a second scan at least seven days later before a diagnosis is made. Where no fetal pole is visible and the mean gestational sac diameter is under 25.0 mm, 1.6.6 requires the same seven-day wait.

So being sent home to return in a week is not evasion or bad news withheld. It is the guideline, and it exists because scans done too early have been wrong in both directions. Separately, for bleeding before six weeks without pain and with no risk factors, NG126 1.4.10 uses expectant management: repeat a urine pregnancy test after 7 to 10 days, return if it is positive, and a negative means the pregnancy has ended.

This matters in India because progesterone is prescribed readily for early pregnancy bleeding, and the largest trial ever run on exactly that question did not find it worked.

PRISM randomised 4,153 women at 48 UK hospitals who presented with bleeding in early pregnancy to vaginal progesterone 400 mg twice daily or matching placebo, from presentation through 16 weeks. Live birth after at least 34 weeks occurred in 75% of the progesterone group (1,513 of 2,025) and 72% of the placebo group (1,459 of 2,013) - a relative rate of 1.03, with a confidence interval from 1.00 to 1.07 and a P value of 0.08. Adverse events did not differ significantly between the two groups.

That is not a finding of harm, and it is not a reason to stop medication somebody has prescribed you. It is a reason to ask what it is expected to do. Progesterone in fertility treatment is a separate question with a separate evidence base, because luteal support after an embryo transfer is not the same intervention as progesterone for threatened miscarriage.

On this subject the gaps are the most useful thing on the page, because they are where false reassurance and false alarm both come from.

• That the colour of the blood means anything. No study has shown that pink, brown or bright red predicts how an early pregnancy ends. Hasan 2009 examined colour alongside duration, pain and timing, and what the risk tracked with was heaviness, and heaviness with pain.
• A single reliable figure for how common early bleeding is. The published range runs from 14 of 151 pregnancies in Harville's prospective cohort to 1,207 of 4,539 women in Hasan's, because they count different windows and ask differently.
• That light bleeding is harmless in every group. Hasan's confidence interval of 0.9 to 1.3 is consistent with no effect and with a small one. It rules out a large effect, not every effect.
• That progesterone does nothing for anyone with early bleeding. PRISM's overall result was not statistically significant at P=0.08 with an interval touching 1.00, which is a null primary result rather than a demonstration of no benefit anywhere.
• That an early scan can settle the outcome. NICE 1.6.1 states the opposite, and 1.6.3 and 1.6.6 build a mandatory waiting period into the diagnosis for exactly that reason.
• How any of this differs for Indian women. Hasan's cohort was recruited in the United States, PRISM's in the United Kingdom, and Harville's in the United States. No ICMR or Indian registry data on early pregnancy bleeding patterns was located for this page, so these figures are reported as what they are rather than as national rates.

Bleeding when you badly want a pregnancy to continue is frightening in a way that a confidence interval does not touch, and there is no version of this page that makes the waiting shorter. What can be said is that most early bleeding in these cohorts did not end the pregnancy, that heaviness and pain are the things worth acting on, and that being asked to come back in a week is the protocol rather than a verdict. If a loss has happened before, that changes what gets investigated and is worth raising at the appointment.