Wisdom teeth and your medications

Nearly every patient on a blood thinner reaches the same conclusion before the consultation does. Surgery means bleeding, the tablet makes bleeding worse, so the sensible move is to skip a few doses. It is a reasonable-sounding chain of thought and it is the wrong one, for a reason worth understanding rather than simply accepting.

Bleeding from a socket at the back of the jaw is a local problem with local solutions. It is visible, it is reachable, and it responds to pressure, a suture, a haemostatic material packed into the socket and, if needed, a return visit the same day. A clot forming on a coronary stent, on a mechanical heart valve, or in an artery supplying the brain is none of those things. It is not visible, not reachable, and the damage it does is frequently permanent.

Whoever prescribed the drug knows why it was prescribed: which valve, which stent, how long ago, which clot, and what happens if it recurs. None of that information lives in the mouth. The surgeon supplies the other half, stated precisely — how many teeth are coming out, how deep they sit, how much bone will be removed, and whether the work can be split across more than one visit. A prescriber asked whether it is alright to have a tooth out can only answer generally. A prescriber asked about two lower bony impactions planned together can answer specifically.

So the sequence runs: consultation first, plan agreed, then a letter or a call to the prescriber with the actual plan in it, then a date. Working in the other order, booking the surgery and sorting out the medication afterwards, is how appointments get moved on the morning.

The medications on this page change how a socket behaves, and they change how much time you have to react when something is wrong. Three findings carry a clock, and all three are worth knowing before the day rather than off a discharge sheet afterwards.

Altered sensation is the finding patients most often sit on. Local anaesthetic in the lower jaw wears off within a few hours; numbness of the lip, chin or tongue that outlasts it means the nerve has been disturbed. Most such disturbances settle, but the assessment, the record of what was found and when, and any decision about onward referral are all time-dependent. Waiting a fortnight to mention it removes options that were available in the first week.

Bleeding is the finding these drugs most obviously affect, and the pattern matters more than the volume at any single moment. Oozing that gradually settles through the evening of surgery is ordinary while an anticoagulant is running. Bleeding that had stopped and then restarts briskly on day two or three, or a clot that keeps forming and washing away, is not, and neither is a mouth filling faster than you can swallow.

Infection is the third, and swelling rather than pain is the signal to act on: a spreading infection sometimes hurts less as it moves, because the pressure that caused the pain has escaped into the surrounding tissue. Swelling confined to the gum around the socket is one thing. Swelling that has crossed into the cheek, tracked under the jaw, reached the eye or lifted the floor of the mouth has entered a fascial space, and that is a hospital problem.

For most third molar surgery in most patients, the answer that comes back from the prescriber is to continue the drug and manage the bleeding locally. That is the direction of the published dental guidance rather than a local preference. The Scottish Dental Clinical Effectiveness Programme guidance on managing dental patients taking anticoagulants and antiplatelet drugs is the document most often worked from, and American dental literature runs the same way. Both moved away from an earlier period in which routine interruption was common.

Warfarin is the one drug in the group with a number attached to it. An INR taken close to the appointment — commonly within seventy-two hours where readings are stable, and nearer the day where they move about — tells the surgeon what they are working with. Where the INR is stable and sits below the threshold used in that guidance, extraction generally proceeds with the drug continued. Above it, the case waits for the anticoagulation clinic or moves to a hospital setting. Note what that means in practice: the INR sets the timing and the setting, not the dose.

The direct oral anticoagulants — apixaban, rivaroxaban, edoxaban and dabigatran — have no equivalent routine test. What matters instead is kidney function, because that governs how long the drug stays in the system, and the timing of the appointment relative to the daily dose. Some prescribers suggest an adjustment to timing on the day. That suggestion comes from them. Reading somewhere that the morning dose can simply be skipped is not the same thing.

Antiplatelet drugs are a different mechanism and a different conversation. Aspirin on its own is rarely a reason to change the plan, and clopidogrel on its own seldom is. The situation that genuinely constrains things is dual antiplatelet therapy — aspirin combined with clopidogrel, ticagrelor or prasugrel — in the months after a coronary stent. Interrupting the second agent in that window carries a real risk of the stent clotting off, and the usual answer is not to interrupt but to defer elective surgery until the window has passed, treating only what cannot wait. Whether a wisdom tooth can wait is a clinical judgement made against symptoms, infection, and the condition of the second molar in front of it.

Continuing is a default rather than a rule, and it becomes less straightforward when several things stack up: four teeth in one sitting instead of one, deep bony impactions instead of erupted teeth, an anticoagulant alongside an antiplatelet, liver or kidney disease, an inherited bleeding disorder, chemotherapy affecting platelet counts, heavy alcohol use. The response to stacking is usually to change the operation rather than the medication — fewer teeth per visit, one side at a time, earlier in the day and earlier in the week, so that a problem on the evening of surgery is a problem during working hours.

The table below maps which conversation each group of drugs starts, and with whom. It is not a set of instructions, and no row in it authorises a change to anything you are taking.

For nearly every row, what changes is decided somewhere other than this office. That is the reason the medication list has to arrive at the consultation rather than at the surgery.

The local measures are unglamorous and they are what makes continuing the drug workable. The tooth is removed with as little trauma to the surrounding bone as its position allows, because bone that has been ground away bleeds and bone that has not does not. The socket is sutured. A haemostatic material, usually oxidised cellulose or a collagen sponge, is placed into the socket to give the clot a scaffold. Where a haematologist or prescriber advises it, a tranexamic acid mouthwash is used. Where the number of teeth and the medication together warrant it, the work is staged across more than one visit rather than taken in a single sitting.

Then the part that belongs to the patient, which is pressure. Firm, continuous biting on folded gauze for a full thirty minutes, without lifting it to check. The urge to look is strong and every look restarts the clock. A damp black tea bag substitutes for gauze if you run out. Keep the head elevated, do not rinse at all for the first twenty-four hours, and use no straws and no cigarettes.

Painkillers after surgery are chosen around the anticoagulant rather than by habit. Anti-inflammatory drugs such as ibuprofen, naproxen and aspirin affect platelets and add to a bleeding tendency. Acetaminophen, sold outside the United States as paracetamol, does not, but taken regularly over several days it can raise the INR in someone on warfarin, which is why the prescriber is told about a course of it rather than left to discover it. Ask what to use and what to avoid. Do not assemble a plan from the medicine cabinet.

If something was paused before surgery, it restarts on the prescriber's instruction and at the time they specified. Not earlier because the bleeding looks settled, and not later because it does not. If bleeding is genuinely preventing a restart on schedule, that is a phone call to make on the day rather than a decision to take alone, and this practice will make it on your behalf if you would rather it came from here.

The routine instructions matter more on these medications, not less, because the clot is slower to organise. After the first twenty-four hours, warm salt water is allowed to fall out of the mouth rather than spat, and smoking stays out for longer than the packet of gauze lasts — it is separately the strongest modifiable factor in dry socket.

These drugs work by slowing the cells that break bone down. That is the point of them, and it is also why they matter here. The jaws remodel faster than any other bone in the body, and an extraction socket is a piece of bone being asked to remodel on demand.

The American Association of Oral and Maxillofacial Surgeons defines medication-related osteonecrosis of the jaw, MRONJ, as exposed bone in the jaw, or bone that can be probed through a sinus tract in or around the mouth, persisting for more than eight weeks, in a patient who has taken an antiresorptive or antiangiogenic drug and who has not had radiotherapy to the jaws. The definition is deliberately narrow. A socket that is taking its time is not this condition.

Risk is not a single number. It is low for someone taking a weekly osteoporosis tablet and materially higher for someone receiving monthly injections at cancer dosing, and duration of exposure moves it: several years is a different proposition from several months. So do the things sitting alongside it, particularly corticosteroids, antiangiogenic cancer drugs, poorly controlled diabetes, smoking, and active infection at the site.

The conclusion patients sometimes draw, that the tooth should therefore be left alone, is usually the wrong one. An infected or unsalvageable tooth sitting in a jaw with suppressed turnover is itself one of the situations MRONJ arises from. Leaving disease in place does not avoid the problem. It exchanges a planned extraction with modified technique and close follow-up for an unplanned one under worse conditions.

What changes about the operation is real and undramatic. Infection is settled first where there is time to do it. The flap is designed and closed so the socket is covered rather than left open, and sharp bone edges are smoothed rather than left to work their way through the gum. An antimicrobial mouthwash is used, antibiotic cover is considered case by case, and follow-up runs over weeks and months rather than ending at one post-operative visit. Where cancer dosing is involved the plan is made with the oncology team. If antiresorptive therapy has not started yet, a dental assessment before the first dose is the easiest point in the sequence at which to act.

For anyone on a bone-modifying drug the follow-up runs longer by design. Report a socket that has not closed over by around four to six weeks, any rough or exposed area of bone that persists, a persistent bad taste coming from the site, or numbness of the lip that appears later rather than immediately. Those are the findings that separate a socket healing slowly from a socket that is not healing.

Drug holidays, and why denosumab is different

The idea of a drug holiday, pausing the antiresorptive around surgery, is widely discussed and poorly settled. The evidence for it is limited, and the drugs do not behave alike. Bisphosphonates persist in bone long after the last dose, so pausing them changes less than it sounds as though it should. Denosumab does not persist, and bone turnover rebounds after a missed dose in a way that has been associated with vertebral fractures. That is why denosumab is not a drug to pause casually, and why elective dental surgery is more often scheduled within the dosing cycle than around a pause. Whether to pause anything is the prescriber's decision, taken with the fracture or cancer risk in front of them rather than the socket.

Three separate questions get asked of any drug that suppresses the immune system. Does it slow healing. Does it blunt the response to infection so that trouble presents late and quietly. Does it interact with what would otherwise be prescribed after surgery. The answers differ by drug, and the third one catches people out most often.

Long-term corticosteroids thin the tissues, slow healing, and mute the inflammatory signs that would otherwise announce an infection early. On adrenal suppression the position has moved: current guidance does not routinely call for supplementary steroid cover before a straightforward extraction under local anaesthetic. What matters is the dose history and how you have responded to previous surgery, and that is settled case by case with the prescriber rather than by a rule.

Conventional disease-modifying drugs and biologics — methotrexate, TNF inhibitors, rituximab and the JAK inhibitors among them — raise infection risk to differing degrees. Rheumatology guidance often times elective surgery towards the end of a dosing interval rather than just after a dose. That timing is set by the rheumatologist against how active the disease is, and it is a scheduling question rather than a surgical one.

Transplant medication brings its own set. Tacrolimus, ciclosporin and mycophenolate suppress immunity, and ciclosporin in particular causes gum overgrowth, which around a partially erupted lower wisdom tooth makes an already uncleanable pocket worse. The interaction that matters most is with antibiotics and antifungals: macrolides and azoles raise tacrolimus and ciclosporin levels, so what is prescribed after surgery is chosen against that list, with the transplant team in the conversation.

For anyone on chemotherapy the constraint is the blood count. Neutrophils and platelets fall and recover on a cycle, and elective surgery is timed to the recovered part of it with a recent full blood count in hand. Where cancer treatment has not yet begun, a dental assessment before it starts is the same argument as with the bone drugs, and for the same reason.

Diabetes raises two separate questions that are easy to merge: healing over the following weeks, and what happens during the appointment on the day. On the first, the variable is control rather than the diagnosis. Well-controlled diabetes changes surgical planning very little. Poorly controlled diabetes slows healing, raises the chance of infection, and is associated with more trouble at the socket. HbA1c is the number the surgeon asks for, and where it is high and the tooth is causing no symptoms, spending several weeks improving control before an elective extraction is sometimes the better plan. Where there is infection or pain, surgery does not wait for a number.

On the day, the acute risk in the chair is a low blood sugar rather than a high one. An early appointment helps. Take your medication and eat as you normally would unless you have specifically been told otherwise. Bring your meter and a fast-acting carbohydrate. Mention the diabetes on arrival even though it is already on the form.

Where the comfort plan agreed at your consultation involves more than local anaesthetic, fasting instructions may apply, and fasting collides with insulin and with sulfonylureas. GLP-1 receptor agonists slow gastric emptying and can lengthen the window required. SGLT2 inhibitors combined with a prolonged fast can produce ketoacidosis at a near-normal glucose reading, which is exactly the presentation that gets missed. Each of those is a question for the prescriber, asked in advance, with the appointment time in front of them.

Afterwards, expect glucose to run higher for a few days — partly surgical stress, partly a soft diet that skews towards carbohydrate, because ice cream, smoothies and yoghurt are what the mouth will tolerate. Test more often than usual through the first week and report the readings rather than adjusting insulin yourself. Sockets in poorly controlled diabetes are reviewed with a lower threshold, and a review appointment offered early is not a sign that something has gone wrong.

Antidepressants in the SSRI and SNRI classes reduce platelet serotonin and produce a modest bleeding tendency of their own. Alone that changes nothing. Added to an anticoagulant or an antiplatelet drug it becomes part of the picture, which is why it belongs on the list even though nobody will be asked to stop it.

Supplements are left off more often than anything else, because they were bought off a shelf and do not feel like medication. Fish oil, ginkgo, garlic, ginseng, high-dose vitamin E and turmeric all affect platelet function to varying degrees. St John's wort induces liver enzymes and changes how other drugs behave, including some prescribed after surgery. None is alarming in isolation and all are worth knowing about.

Over-the-counter painkillers count too. Anyone who has been managing a sore wisdom tooth with ibuprofen for a fortnight arrives at the consultation already on a drug that affects platelets, and for anyone on methotrexate, regular use of an anti-inflammatory is a genuine interaction rather than a theoretical one. Say what you have been taking to cope, not only what is written on a prescription.

Antibiotics before surgery

This is a narrower question than most patients expect. American Heart Association guidance restricts prophylaxis to a short list: a prosthetic heart valve or prosthetic material used in valve repair, previous infective endocarditis, certain congenital heart conditions, and a transplanted heart that has developed a valve problem. For prosthetic hips and knees, dental and orthopaedic guidance no longer supports routine prophylaxis, although some orthopaedic surgeons still request it, and that is a conversation between the two clinicians rather than something for the patient to arbitrate.

Two more belong on the list. Hormonal contraception raises the background rate of dry socket, which is worth knowing when timing is flexible. Pregnancy and breastfeeding change imaging, anaesthetic choice, painkillers and antibiotics all at once, and a wisdom tooth that can wait usually does, with the second trimester generally preferred for anything that cannot.

Finally, the ones people would rather not write down. Cocaine and methamphetamine interact with the vasoconstrictor in local anaesthetic in a way that can push blood pressure and heart rate to a level that is unsafe in the chair, so recent use has to be disclosed on the day rather than agreed to in principle at a consultation. Cannabis alters how much local anaesthetic is needed. Regular heavy alcohol use affects clotting and healing both. Answered honestly, these change the plan. Left unanswered, they change the outcome.

Saying you are on a blood thinner is not a medication list. There are at least four different mechanisms hiding behind that phrase, they behave differently in a socket, and the plan diverges depending on which one it is. The same applies to a bone medication, an immunosuppressant, or an injection for arthritis.

The item most often missing is the one that does not feel like a medication because it is not taken at home: a denosumab injection every six months at the doctor's office, a zoledronic acid infusion once a year at the hospital, a biologic given at a day unit. So ask yourself directly — has anyone given me an injection or a drip for my bones, my joints, or a cancer at any point in the past few years, and when was the last one?

A medication that surfaces on the morning of surgery usually means the day is moved, because the prescriber has to be contacted and a letter written while a patient sits in the waiting room does not come back in time. Bringing the list to the consultation gives everyone the interval they need in order to use it.

The list is then checked against a fixed set of flags: anticoagulants and antiplatelet drugs, bone-modifying agents, immunosuppressants, diabetes medication, cardiac history, pregnancy, latex allergy and any previous complication under anaesthesia. Each is checked on every patient, whether or not it came up while you were talking. The case that check exists for is a bisphosphonate history nobody remembers to mention.

Every one of these plans is written against the list you brought, and any of them can be revised if the list changes. A new prescription between the consultation and the surgery date, a stent placed in the interval, a bone injection brought forward, or a course of steroids for something unrelated are all worth a call before the day rather than a mention on it.