Bleeding that will not stop after a wisdom tooth removal

Normal bleeding after a third molar extraction looks far worse than it is, because a small volume of blood mixes with a large volume of saliva and stains all of it red. A useful measure is the gauze, not the mirror. If a fresh folded gauze pad held under firm bite pressure for thirty minutes comes out damp and pink rather than heavy and bright red, the socket is behaving normally, and the staining you see in the sink is mostly saliva.

The pattern that matters is the direction of travel. Ordinary post-operative oozing gets lighter each hour. By four to eight hours after surgery most people are down to a faint pink tinge when they spit, and by the following morning to nothing at all. Bleeding that is the same at hour six as it was at hour one, or heavier, is not following the normal curve, and the number of hours elapsed is more informative than the amount of red you can see.

A second distinguishing sign is whether the blood is flowing or seeping. Blood that wells up and fills the floor of your mouth within a minute or two of removing gauze, or that runs steadily rather than pooling slowly, is arterial or brisk venous bleeding from the socket wall and does not usually settle with home measures. That is a call, not a wait-and-see.

Bite on a fresh, folded, moistened gauze pad placed directly over the socket for a continuous thirty minutes without checking it, then replace it and repeat once more if bleeding continues. Two uninterrupted thirty-minute cycles settle the large majority of post-extraction bleeding. The two most common reasons pressure fails are that the gauze is sitting on the teeth rather than on the socket, and that it is being lifted every few minutes to look, which strips the forming clot each time.

1. Wash your hands, then fold a clean gauze square into a firm pad roughly the size of the gap, thick enough that your back teeth do not meet through it.
2. Moisten it slightly with tap water. Dry gauze bonds to the clot and pulls it out when you remove it.
3. Place it directly over the socket, not on the chewing surface of the tooth in front of it. On a lower third molar site the pad has to sit well back, which is the step most people get wrong.
4. Bite firmly and hold for a full thirty minutes by the clock. Sit upright, keep your head above your heart, and do not talk, spit, rinse or check.
5. Remove it gently at thirty minutes. If it is damp and pink, stop. If it is soaked and bright red, replace with a fresh pad and repeat once.
6. If a second full cycle has not controlled it, telephone. Do not start a third and fourth cycle for hours on end.

A moistened ordinary black tea bag substituted for gauze on the third cycle is a reasonable home measure while you are waiting to be seen or waiting for a callback. The tannic acid in ordinary black tea helps a clot organise. It is an adjunct to pressure and a bridge to being assessed, not a reason to stop counting the hours.

Vague advice to call "if you are worried" puts the whole judgment on a person who is frightened, tired and looking at a sink full of red water. The thresholds below are deliberately conservative. Meeting any single one of them is enough. You do not need to meet several, and you are not expected to work out the cause before you pick up the phone.

Notice that two of these rows do not send you to the practice at all. That is deliberate. A dental office, this one included, is not the right place for someone who is losing enough blood to feel faint or whose airway is narrowing, and calling us first in those situations costs minutes that matter.

Go to a hospital emergency department, or call 911, if you feel faint or lightheaded when you stand, if your heart is racing while you are sitting still, if you are pale and clammy, if you are vomiting blood or coffee-ground material after swallowing it, if you cannot swallow your own saliva, or if breathing feels harder than it did an hour ago. Any one of these is enough on its own and none of them is a reason to wait for a callback.

Call 911 rather than driving if you have any symptom on that list. A person who is lightheaded from blood loss should not be behind a wheel, and a person whose airway is narrowing needs to be somewhere that can secure it. If someone is with you and you are clearly deteriorating, ask them to make the call rather than debating it.

The same applies out of hours. A surgical practice that has to reach an on-call surgeon, open an office and set up a light and suction cannot compress that into the minutes an emergency department can. If you are on the edge of the thresholds above at two in the morning, the emergency department is the correct destination and you can tell the practice afterwards.

Anticoagulant and antiplatelet medicines do prolong post-extraction bleeding, typically turning a four-hour ooze into one that runs for the better part of a day and responds more slowly to pressure. That is expected, it is manageable, and it is not a reason to alter a dose. Current guidance is that stopping an anticoagulant to allow a dental extraction generally exposes a patient to more risk than the bleeding does.

What changes is the threshold at which you should make contact, not the protocol itself. If you take warfarin, apixaban, rivaroxaban, dabigatran, clopidogrel, daily aspirin or any similar agent, treat the twelve-hour mark as a six-hour mark and telephone earlier rather than later. The same shortening applies if you have haemophilia, von Willebrand disease, liver disease, chronic kidney disease, or a low platelet count, and to anyone receiving chemotherapy.

Tell the practice about all of this before surgery rather than after it. An anticipated bleeding risk changes what is prepared in the room — suturing technique, local haemostatic materials, how long you are observed before discharge — and none of those preparations can be made retrospectively at eleven at night. If a medical history has changed since your consultation, including a new prescription, say so on the day.

• Never start, stop, skip or adjust a prescribed medicine because of bleeding. Ask the prescriber, or ask the clinician assessing you to contact them.
• Bring the actual boxes or an accurate list, including doses and the time of the last dose.
• Mention supplements. Fish oil, high-dose vitamin E, ginkgo and turmeric all affect platelet function and are frequently left off a history because people do not think of them as medicines.
• If you have a haematologist, say so early. Their input often changes the plan more than anything a surgeon can do at the chair.

Late bleeding behaves differently from day-one bleeding and is worth separating out. A small pink ooze on day two or three, usually after eating something firm or after brushing near the site, is common and generally settles with a single pressure cycle. It reflects a young clot being disturbed at the surface rather than a socket that has failed to clot at all.

Frank bleeding that returns on day three to day five, particularly with increasing rather than decreasing pain, a foul taste, or a fever, is a different pattern and needs assessment. Bleeding is rarely the main event there; it is a signal accompanying something else in the socket, and the accompanying symptoms are what determine the urgency. Two of this practice's other pages cover those patterns in more detail.

One specific combination deserves naming. Bleeding from an upper third molar site accompanied by air or liquid passing between the mouth and the nose, or a nosebleed on the same side, should be reported the same day rather than watched. Do not blow your nose, and avoid anything that raises pressure in the sinus until you have been assessed.

Knowing what an assessment involves makes the decision to come in easier, so here it is plainly. The socket is anaesthetised with local anaesthetic if it is not already numb, cleared of the loose jelly-like clot that has been preventing pressure reaching the bleeding surface, and inspected under a light and suction to identify where the blood is actually coming from. Most of the time it is a specific point, not the whole socket.

From there the measures escalate only as far as they need to: firm packing with a resorbable haemostatic material placed into the socket, a suture or two across the socket to hold soft tissue against bone, and occasionally a local agent applied directly to a persistent point. The visit is usually short. Most people are surprised how quickly a socket that had been bleeding for eight hours stops once the disorganised clot is removed and pressure reaches the right place.

You will also be assessed generally, not just locally. Pulse, blood pressure, how much blood has actually been lost, and whether a medical cause is driving the bleeding rather than a local one. If the picture suggests a systemic reason, referral onward to a physician or a hospital is the correct outcome and it will be arranged rather than deferred.

Everything above describes bleeding from a third molar socket in the first days after surgery. If you are bleeding from the gums generally, from a site that was never operated on, from both jaws at once without recent surgery, or from your nose and gums together, that pattern points away from a surgical socket and toward a medical cause, and the right first contact is your physician or an emergency department rather than an oral surgery practice.

This practice treats third molars and provides anaesthesia for that surgery. It does not manage bleeding after a facial injury, after other dental surgery performed elsewhere, or bleeding related to a medical condition being managed by another team. If your extraction was carried out somewhere else, the surgeon who did it holds your records, the operative note and the knowledge of what the socket looked like, and contacting them first is usually faster.

Nothing on this page is a diagnosis, and no article can examine a socket. It describes patterns and sets thresholds so that a reader in the middle of the night has a number to act on rather than a feeling to argue with. Where a threshold is borderline, the intended reading is to be seen. Being assessed and sent home is a good outcome, not a wasted trip.