Systemic Disease and Dentistry, a Complete Guide: a domain map from bidirectional relationships and medication risk to disclosure of medical history

Systemic Disease and Dentistry, a Complete Guide: a domain map from bidirectional relationships and medication risk to disclosure of medical history

TL;DR

Diabetes and periodontitis are risk factors for each other [Fn7]; antiresorptive drugs are associated with osteonecrosis of the jaw [Fn68]; before dental treatment, disclose your medication history in full [Fn3]. Whether an antithrombotic drug is adjusted is in every case decided by a doctor.

(50 characters in the Chinese original, not counting the [Fn] markers)

Introduction

This article is general oral-health education based on international literature. It does not address any country's insurance or regulations; consult local rules for care pathways and costs.

This article deliberately does not answer the question "I am on this drug, can I have a tooth taken out?" That is a question only the physician and the dentist, holding your complete medical record, medication list and laboratory values, can answer together. What this article deals with is the gap at the domain level: why systemic disease and dentistry pull on one another, which axes the professional community uses to describe this border, what the frequently quoted percentages are each actually measuring, and where the evidence is in fact still thin.

Three things need saying first, because they govern how every later passage should be read:

1. This article contains no instruction to stop, continue, or dose any medication. All the conclusions quoted here of the "continue" or "no need to interrupt" kind are conclusions at the level of study populations in the literature, not instructions for action addressed to the reader; any adjustment of medication can be decided only by the physician who wrote that prescription.
2. The strength of the evidence in this domain is extremely uneven. Within the same article, some passages rest on a moderate-certainty meta-analysis of 30 trials and 2,443 people [Fn16], while others rest on nothing but an honest blank: the evidence is insufficient to either claim or refute [Fn79]. This article marks that out passage by passage.
3. Population figures are not individual probabilities. Every rate quoted here is a value observed in a particular study population under particular conditions; change the population or change the criteria and the number changes.

1. Why the dental operatory is a setting for systemic medicine

1-1 Dental patients walk in already carrying systemic disease

An observational, retrospective, cross-sectional study of 862 patients attending one dental clinic, based on medical-history questionnaires over a period of 6 months [Fn115], reports that allergies, cardiovascular disease and metabolic disorders were the conditions most commonly reported in that population [Fn1]. Its conclusion is that a thorough medical history should be obtained for all dental patients to ensure safe dental care [Fn2], and it stresses that identifying and taking account of chronic systemic diseases at the treatment-planning stage is essential to minimise the risk of intraoperative and postoperative complications, particularly in oral surgery [Fn105].

The significance of this passage lies not in those three categories themselves (a single-clinic sample cannot represent the composition of patients worldwide), but in what it says in reverse: the dental medical-history questionnaire is not an administrative formality, it is part of the safety process.

1-2 Oral procedures have systemic reach

A systematic review and meta-analysis including 64 non-randomised controlled trials and 25 randomised controlled trials [Fn116] reports that dental extractions showed the highest incidence of bacteraemia among dental procedures (62%–66%), followed by scaling and root planing (44%–36%) and by oral health procedures without root planing (27%–28%) [Fn4]. The same analysis also records a fact that is easily overlooked: everyday activities cause bacteraemia too — flossing and chewing at 16%, toothbrushing at 8%–26% [Fn5].

This thread matters because it is the physiological basis for the whole conceptual turn in section 6 (antibiotic prophylaxis for endocarditis): if everyday toothbrushing also causes bacteraemia, then the logic of placing the entire preventive weight on a single dose of antibiotic on the day of the appointment is itself open to question.

1-3 Systemic disease also shows itself in the mouth

The MedlinePlus patient-education page on diabetes complications, from the US National Library of Medicine (NLM/NIH), lists gum disease and other dental problems among the complications of diabetes, and explains the mechanism as high levels of glucose in saliva helping harmful bacteria in the mouth to grow [Fn6].

This is the most basic bidirectional intuition in the domain, and it has been examined in the literature fairly thoroughly — see the next section.

2. Diabetes × periodontal disease: how to read the bidirectional relationship

This is the passage where the evidence in this domain is thickest, and also the one most easily oversimplified.

2-1 What "bidirectional" precisely means in the literature

The consensus report of the joint workshop of the International Diabetes Federation (IDF) and the European Federation of Periodontology (EFP) states at the outset that diabetes and periodontitis are both chronic non-communicable diseases, each independently associated with mortality, and that the two have a bidirectional relationship [Fn7]. The same consensus states the two directions separately:

• Periodontitis → metabolism: there is strong evidence that people with periodontitis have an elevated risk of dysglycaemia and insulin resistance [Fn8]; periodontitis is also associated with an increased risk of incident type 2 diabetes [Fn9].
• Metabolism → periodontium: the consensus notes that cohort studies of people with diabetes show significantly higher glycated haemoglobin in those with periodontitis than in those who are periodontally healthy [Fn8], while flagging honestly at the same time that there are insufficient data among people with type 1 diabetes [Fn10].

2-2 The magnitude of the two directions: two numbers, two directions

A systematic review and meta-analysis that included 15 cohort studies [Fn117] and evaluated risk of bias with the QUIPS tool [Fn118] calculated a summary relative risk for each of the two directions:

How to read this table: the effect magnitudes in the two directions are close, and that is exactly what the word "bidirectional" means concretely in epidemiology. But the authors of that study put the main limitation in their own conclusion — there is high, unexplained heterogeneity between the studies, whose sources include differences in the methods used to diagnose the diseases [Fn13]. These two numbers are therefore suited to understanding direction and magnitude, and are not suited to being treated as a risk prediction for any individual.

2-3 Mechanism: inflammatory mediators are the shared language

The IDF/EFP consensus report summarises the mechanistic links between the two diseases as elevations in interleukin (IL)-1-β, tumour necrosis factor-α and IL-6 [Fn14], and in the same passage includes the receptor activator of nuclear factor-kappa B ligand/osteoprotegerin ratio, oxidative stress and Toll-like receptor 2/4 expression [Fn14].

In plain terms: periodontitis is not a local problem confined to the mouth; it shares the same set of mediators with systemic inflammation — and that is the premise on which the later question, whether treating periodontitis feeds back onto blood glucose, can be posed at all.

2-4 The effect of treating periodontitis on blood glucose: two independent sources, pointing the same way

• The consensus-report version: the wording used in that consensus report is that periodontal therapy is safe and effective in people with diabetes, and that it is associated with reductions in glycated haemoglobin of 0.27%–0.48% after 3 months, although studies involving longer-term follow-up are inconclusive [Fn15]. ⚠ Limits on how to read this: the evidence base for that conclusion is mainly the population with type 2 diabetes together with periodontitis, and data on type 1 diabetes are limited; "safe" is that consensus's own word, and the abstract cited here does not report item by item how safety endpoints were assessed, so it should not be read as an endorsement of safety covering every population with diabetes and every periodontal procedure.
• The Cochrane 2022 version: 35 randomised controlled trials, 3,249 participants randomised [Fn119]. Moderate-certainty evidence shows an absolute reduction in glycated haemoglobin of 0.43% at 3 to 4 months after treatment of periodontitis (30 studies, 2,443 analysed participants) [Fn16]; at 12 months the absolute reduction is 0.50%, but that figure comes from only 1 study with 264 participants [Fn17]. The review authors' conclusion is that there is now moderate-certainty evidence that periodontal treatment using subgingival instrumentation, compared with no treatment or usual care, improves glycaemic control in people with both periodontitis and diabetes by a clinically significant amount [Fn18].

Two limits that must always be stated together:

1. Population limit: the studies in the Cochrane review focused on people with type 2 diabetes, other than one study that included participants with type 1 or type 2 diabetes [Fn19]. The effect size above therefore cannot be extrapolated directly to the type 1 diabetes population — this and the consensus report's statement that data on type 1 are insufficient [Fn10] are two formulations of the same gap.
2. Insufficient safety data: the review notes that adverse effects of periodontal treatments were not evaluated in most studies [Fn20]. Adverse effects not having been reported is not the same as adverse effects having been shown not to occur.

2-5 Where it sits in the diagnostic framework: diabetes is a modifying factor, not another disease

The consensus report of workgroup 3 of the 2017 World Workshop (classification of periodontal and peri-implant diseases) sets this out explicitly: diabetes-associated periodontitis should not be regarded as a distinct diagnosis, while diabetes should be recognised as an important modifying factor and included in a clinical diagnosis of periodontitis as a descriptor [Fn21]. The same consensus handles smoking by the same logic — likewise an important modifier to be included in the clinical diagnosis as a descriptor [Fn22].

This point is pivotal in health education: it shows that the professional side does not treat "diabetic periodontitis" as another disease to be treated, but adds, on top of the same periodontal diagnosis, a modifying condition that changes treatment expectations and follow-up frequency.

2-6 What this section implies in practice: comanagement, not each side treating separately

The IDF/EFP consensus report positions its own output as consensus guidelines for physicians, oral healthcare professionals and patients, to improve early diagnosis, prevention and comanagement of diabetes and periodontitis [Fn23].

Adjacent concrete questions in this domain (each has its own canonical card; not expanded here)

- What periodontal treatment involves, how long the course runs and what its costs are made of — see canonical card KM-DENTAL-48 (in production).

- How to read swollen, painful gums in the moment and when to seek care — see canonical card KM-DENTAL-05 (in production).

- For the full domain framework of periodontal treatment, see domain article P05 (Periodontics and gums).

3. Diabetes × dental implants: survival and health are two different outcomes

This is the passage in this domain that is most easily misread. The error almost always occurs at the same point: treating "is the implant still there" and "are the peri-implant tissues healthy" as one and the same thing.

3-1 Survival: most meta-analyses show no significant difference

• A systematic review and meta-analysis including 9 clinical studies [Fn120], published in the official journal of the American Dental Association, shows no significant difference in implant failure rates between patients with type 2 diabetes and non-diabetic patients [Fn24]; the authors' conclusion is written as: patients with type 2 diabetes mellitus seem to be able to achieve a rate of implant survival similar to that of healthy patients [Fn26].
• A systematic review and meta-analysis of immediately loaded implants (with 5 studies in the quantitative synthesis [Fn121]) shows no significant difference in the survival rate of dental implants between people with type 2 diabetes and people without (RR = 1.00, 95% CI 0.96–1.04) [Fn33]; its conclusion attaches three conditions: type 2 diabetes does not seem to be a risk factor for immediately loaded implants if the glycaemic level is controlled, oral hygiene is satisfactory and the technical steps are strictly followed [Fn34].
• A systematic review and meta-analysis pooling 60 studies and assessing the effect of advanced age and systemic conditions likewise records that high implant survival rates were reported for patients with type 2 diabetes mellitus or Parkinson's disease [Fn37].

3-2 Health: the other half of the conclusions within the same studies

Take those same studies and change the outcome measure, and the direction changes:

• Bleeding on probing and peri-implant bone loss were significantly worse in the diabetic group [Fn25], on which basis the authors state that hyperglycaemia is an important risk factor for peri-implant inflammation [Fn27].
• A systematic review and meta-analysis including 10 studies and 634 participants [Fn122] reports that implant survival was not compromised in the study population with glycated haemoglobin below 10% [Fn28]; but the dose-response meta-analysis in the same paper shows that, as glycated haemoglobin rises, advanced glycation end-product accumulation, probing depth and marginal bone loss worsen in a dose-response dependent manner [Fn29].
• A systematic review and meta-analysis including 22 studies [Fn123] reports that, for the study population with glycated haemoglobin below 8%, survival over the first 3 years fell across the included studies within a range at or above 92.6% [Fn30]; whereas in those with glycated haemoglobin above 8%, osseointegration progressed at a slower rate and inflammatory cytokines and bone biomarkers were adversely affected [Fn31].

⚠ How to read those two rows of figures: they are values observed in particular study populations under short- to medium-term follow-up, not the expected result for any individual, and not a claim of efficacy. Once the follow-up lengthens or the population changes, the numbers differ — which is exactly why the same body of studies points in a different direction on the health outcome (see the next subsection).

3-3 Putting the two halves together: this is the complete conclusion

Survival and health are two outcomes, and the evidence gives a different answer on each. The short- to medium-term data on whether the implant is still in place are relatively optimistic [Fn24][Fn33], while inflammation and bone loss in the peri-implant tissues are clearly related to glycaemic status [Fn25][Fn29][Fn31].

This is also why the clinical-implication sentence of that dose-response review reads: glycated haemoglobin values must be considered for risk assessment before placement and throughout the lifespan of the implant placed in a patient with diabetes [Fn32]. Note that the subject of that sentence is risk assessment on the clinical side, not a threshold instruction issued to patients — this article gives no individual advice about how low glycated haemoglobin must be before an implant can be placed.

3-4 Another supporting line: where diabetes sits on the list of risk factors for peri-implantitis

A systematic review including 57 studies [Fn124] lists diabetes mellitus among the risk factors for peri-implantitis (effect summary OR 2.5, 95% CI 1.4–4.5), alongside smoking, lack of prophylaxis, and a history or presence of periodontitis [Fn35]. The same paper also marks out why figures in this domain are hard to compare with one another: the reported prevalence of peri-implantitis at the implant level ranges from 1.1% to 85.0% [Fn36].

The same review also records honestly one side of the evidence gap: currently there is no convincing or low evidence available that identifies osteoporosis, absence of keratinised mucosa, implant surface characteristics or edentulism as risk factors for peri-implantitis [Fn110].

A portable way of reading this: whenever you see any percentage for peri-implantitis or for implant survival, first ask what is being measured — implants or patients, over how long a follow-up, under which criteria. 1.1% and 85.0% coexist inside the same review [Fn36]; that is not a contradiction, it is a difference of definition and of population.

Adjacent concrete questions in this domain (each has its own canonical card; not expanded here)

- The full course of an implant and what each stage involves — see canonical card KM-DENTAL-25 (in production).

- Situations of regret after an implant, and the key points of assessment beforehand — see canonical card KM-DENTAL-07 (in production).

- For the full domain framework of implants, see domain article P01 (Dental implants).

- Costs and the boundaries of insurance coverage belong to local systems: for local systems and costs see the corresponding canonical cards (TW) and domain article P12 (Complete guide to costs and insurance systems).

4. Anticoagulant and antiplatelet drugs × tooth extraction: general principles (this article gives no instruction to stop any drug)

⚠ This section sets its boundary first: everything below is a conclusion at the level of study populations in the literature, not an instruction for action addressed to anyone. Whether medication is adjusted can be assessed only jointly by the prescribing physician and the dentist; this article provides, and implies, no recommendation to stop or to continue any drug.

4-1 What the risk is at each end of this decision

A Cochrane systematic review sets out the structure of this decision very clearly in its background section: individuals on continuous treatment with vitamin K antagonists (VKAs) or direct oral anticoagulants (DOACs) are at increased risk of bleeding complications during and after oral or dental procedures [Fn42]; yet anticoagulant treatment is preferably continued at the same dose, since dose reduction or discontinuation of treatment is associated with an increased risk of thromboembolism [Fn43]. The same passage also notes that the use of haemostatic measures during or after the procedure (or both) could enable continuation of the oral anticoagulant treatment [Fn44].

Those three sentences together are the skeleton of this section: the risk does not sit at one end, it sits between two ends; and the role of haemostatic measures is to make it possible not to have to choose between two harms.

⚠ This passage describes the structure of the decision as found in the literature; it is not a medication instruction for you. The statement above about continuation at the same dose is a general principle at the level of study populations, and says nothing about whether any individual should adjust their medication; any anticoagulation-related adjustment is judged jointly by the physician who prescribed the drug and the dentist. This article does not provide, and must not be cited as, advice on medication, on stopping, on continuing, or on dosing.

4-2 Vitamin K antagonists (such as warfarin): relatively mature evidence

• The conclusion of a systematic review including 6 randomised controlled trials [Fn125] is written as: patients with an international normalised ratio (INR) within the therapeutic range can safely continue taking the regular dose of warfarin before dental extractions [Fn45]. Two of those studies compared continuation with temporary discontinuation and found that continuation did not increase the risk of bleeding in patients who had an INR within the therapeutic range [Fn46].
• The Cochrane review's conclusion on locally applied tranexamic acid (TXA) is that there seems to be a beneficial effect of locally applied TXA in preventing oral bleeding in people on continuous treatment with VKAs undergoing minor oral surgery or dental extractions [Fn47]; but the same review immediately adds its limitations: the small number of included trials, the relatively small number of participants, and the differences in standard care and procedural protocol between trials do not allow a definite efficacy of antifibrinolytic therapy in this population to be concluded [Fn48].

4-3 DOACs: a clearly different level of evidence, not to be conflated with the passage above

The same Cochrane review states explicitly that it was unable to identify any eligible trials in people on continuous treatment with DOACs undergoing oral or dental procedures [Fn49].

Practical evidence on DOACs therefore comes mainly from observational data. A real-world audit conducted in accordance with the guidance of the Scottish Dental Clinical Effectiveness Programme (SDCEP) (98 patients, 119 dentoalveolar procedures [Fn126]) reports that persistent bleeding followed 17 (14.3%) procedures, of which 11 (9.2%) required specific intervention [Fn50]; its conclusion is that the guidance is safe to follow, with no patients experiencing major haemorrhage [Fn51]. The same audit also notes that a diagnosis of heart failure and advanced age were identified as contributory factors to post-operative bleeding [Fn52].

This is half of the most honest answer this article can give to the question of whether bleeding becomes unstoppable when a tooth is taken out under anticoagulation: what the existing literature supports is a group-level picture in which most episodes can be managed and a few need a return visit for intervention [Fn50][Fn51], rather than any individual probability. This article gives no figure of the "X% of people bleed uncontrollably" kind, because the existing sources provide no such single authoritative value.

4-4 Antiplatelet drugs: the dividing line between monotherapy and dual therapy

A systematic review and meta-analysis including 35 papers, of which 23 provided the requisite information for meta-analysis [Fn127], reports:

• Compared with healthy controls, no immediate or delayed bleeding risk was found in patients treated with aspirin [Fn58].
• Patients receiving dual antiplatelet therapy (DAPT) had a high risk of both immediate (RR = 10.3) and delayed (RR = 7.72) bleeding compared with healthy controls [Fn59].
• The authors' conclusion: dental extraction can be performed safely in patients on aspirin monotherapy; in contrast, patients receiving DAPT should be considered at risk for immediate and continued bleeding [Fn60].

An umbrella review pooling 4 systematic reviews gives the other face of the same thing: DAPT does increase the risk of bleeding-related complications after dental extractions, but the differences may not be clinically significant, as local haemostatic measures were adequate in controlling bleeding [Fn61]; its conclusion is written as: despite the increased risk of bleeding after dental extractions in patients on DAPT, it may not be necessary to interrupt the antiplatelet therapy [Fn62], and it cautions that the complications of discontinuing DAPT may be more severe and fatal [Fn63].

The two papers are not in contradiction: one measures the rate at which bleeding events occur, the other addresses whether those events can be handled by local measures, and what the cost of interrupting the medication would be. Only when both are read together is the decision picture complete.

4-5 The ceiling of the evidence on local haemostatic measures

• A systematic review and network meta-analysis including 8 randomised controlled trials [Fn128] reports that no thromboembolic complications were reported in any of the studies [Fn53]; that the several bleeding episodes associated with anticoagulant drugs were mild and generally happened on the first day after the treatment [Fn54]; and concludes that the use of local haemostatic measures is generally effective for bleeding control, with no further pharmacological drug management or suspension [Fn55].
• But haemostatic adjunct materials are not a universal answer. A PROSPERO-registered systematic review and meta-analysis [Fn135] shows that, in patients on oral anticoagulant therapy, the use of platelet-rich fibrin (PRF) in extraction wounds did not reduce the risk of bleeding after extraction [Fn56], and that the certainty of the evidence in that analysis ranged from moderate to low [Fn57].

The reason this negative result is written in is that it demonstrates the honesty standard of this domain: measures that work have to be stated, and so do measures that have not been shown to work.

Adjacent concrete questions in this domain (each has its own canonical card; not expanded here)

- The healing course of an extraction wound and post-operative care — see canonical card KM-DENTAL-01 (in production).

- For the full domain framework of extraction and oral surgery, see domain article P06 (Extraction and oral surgery).

5. Antiresorptive drugs and osteonecrosis of the jaw (MRONJ): the concept, risk stratification, and the duty to disclose

5-1 The name itself is a piece of history

The 2022 update of the position paper of the American Association of Oral and Maxillofacial Surgeons (AAOMS) records that this condition is now called medication-related osteonecrosis of the jaws (MRONJ), formerly referred to as bisphosphonate-related osteonecrosis of the jaws (BRONJ) [Fn64]. The change of name reflects an expansion in the range of causative drugs, not merely a preference in wording.

The same position paper explains its reason for updating: the knowledge base and experience in addressing MRONJ continues to evolve and expand, necessitating modifications and refinements to the previous position papers [Fn65]; this update contains revisions to diagnosis and management strategies and highlights the current research status [Fn66].

Why this is placed first: this is a domain in which definitions and management strategies are still moving, and any statement made in a settled, final tone should be treated with suspicion.

5-2 What kind of risk it is: rare, but not mild

The 2022 Cochrane review gives, in its background section, two descriptions that must be set side by side:

• Severity: MRONJ is associated with significant morbidity, adversely affects quality of life, and is challenging to treat [Fn67].
• Frequency depends on the drug, the dose and the duration of exposure: depending on the drug, its dosage and the duration of exposure, this adverse drug reaction may occur rarely (for example following the oral administration of bisphosphonate or denosumab treatments for osteoporosis, or antiangiogenic agent-targeted cancer treatment), or commonly (for example following intravenous bisphosphonate for cancer treatment) [Fn68].

A comprehensive literature review states the same direction: the incidence and prevalence of MRONJ are relatively low, although they are clearly higher in cancer patients receiving high-dose antiresorptive agents or angiogenesis inhibitors than in osteoporosis patients receiving oral bisphosphonates or denosumab [Fn69].

5-3 Why tooth extraction is singled out in this topic

The Cochrane review notes that dentoalveolar surgery is considered a common predisposing event for developing MRONJ [Fn70].

A systematic review in cancer patients (7 studies included, all of them observational studies [Fn130]; 550 patients treated with bisphosphonates and denosumab, of whom 271 had received tooth extractions after medication onset [Fn131]) reports that the incidence of MRONJ after tooth extractions varied between 11% and 50% at the patient level [Fn71]; MRONJ was observed in that study to occur up to 3 years after the tooth extraction [Fn72] (this is the study's observation window, not a ceiling on risk); and teeth affected by inflammation before the extraction, together with additional osteotomy during the surgical procedure, were identified as risk factors [Fn73]. The same review also states its methodological limitation: due to significant heterogeneity in the collected data, only a qualitative analysis was performed [Fn74]; and it says plainly that reliable methods of diagnosing MRONJ and adequate follow-up periods are important factors in obtaining the actual incidence [Fn108].

There is one misreading here that has to be blocked: the range of 11%–50% is a value observed in cancer patients on high-dose antiresorptive treatment, after tooth extraction [Fn71]; it cannot be read as saying that somewhere between one in ten and one in two people taking a drug for osteoporosis will develop osteonecrosis of the jaw after an extraction. Change the indication and the dose, and the magnitude changes [Fn68][Fn69].

5-4 A different set of figures for the osteoporosis indication (and it cannot be set directly against the previous set)

A systematic review and meta-analysis carried out for the 2024 update of the International Task Force on Osteonecrosis of the Jaw (in populations with osteoporosis or osteopenia) reports that the pooled rate of MRONJ following implantation in those exposed to antiresorptive therapy was 0.5%, pooled from 21 cohorts [Fn75]; and a single risk-adjusted report found that bisphosphonates increased MRONJ by 3 cases per 1,000 patients (adjusted hazard ratio 4.09, 95% CI 2.75–6.09, moderate certainty) [Fn76].

⚠ These two sets of figures cannot be compared directly: section 5-3 measures the outcome after tooth extraction in a high-dose cancer population [Fn71], while this section measures the outcome after implant placement in an osteoporosis population [Fn75] — different populations, different procedures, different follow-up designs. They are placed in adjacent sections in order to show that indication and dose determine the magnitude [Fn68][Fn69], not so that readers can subtract one from the other.

A third supporting line in the same direction comes from a systematic review of implant survival: patients with bone metastases receiving high-dose antiresorptive therapy carry a high risk of complications after implant surgery [Fn39]; whereas implant survival was reported to be high in patients receiving low-dose antiresorptive therapy for the treatment of osteoporosis [Fn40].

5-5 The evidence on preventive measures: there is a direction, but the certainty is very low

The 2022 Cochrane review included 13 randomised controlled trials with 1,668 participants [Fn129], and recorded one key limitation: the studies were clinically diverse and examined very different interventions, so meta-analyses could not be performed [Fn112]; the authors rated their overall certainty about the available evidence on this topic as low or very low [Fn113]. Within it:

• One open-label randomised controlled trial compared usual care with regular dental examinations at three-month intervals and preventive treatments [Fn77]; that intervention seemed to lower the risk of MRONJ (RR 0.10, 95% CI 0.02 to 0.39, 253 participants) [Fn111]; but two limits have to be stated at the same time: the population of that trial was men with metastatic prostate cancer treated with zoledronic acid [Fn114], and the review authors assessed the certainty of this evidence as very low [Fn78].
• For the various preventive interventions in people on antiresorptive therapy undergoing dentoalveolar surgery, the review's conclusion is written as: there is insufficient evidence to either claim or refute a benefit [Fn79].
• For the treatment of MRONJ once it has occurred, the conclusion is likewise that the available evidence is insufficient to either claim or refute a benefit, in addition to standard care, of any of the interventions studied [Fn80].

This section yields a single portable conclusion: the direction the literature currently supports is that oral assessment and preventive treatment beforehand are worth doing [Fn77], but the certainty of the evidence supporting it was rated by the reviewers themselves as very low [Fn78] — so it is a reasonable clinical direction, not an established conclusion.

5-6 The drug holiday: the evidence does not stand on either side

This is the part of the section that needs the most care, because it is the part most often misread as a recommendation for action.

• A systematic review (including 1 systematic review, 1 randomised controlled trial, 5 observational studies and 3 case reports [Fn132]) records that no studies showed a benefit — that is, a reduction in the incidence of osteonecrosis of the jaw — of short-term withdrawal of antiresorptive agents for tooth extraction [Fn81]; and that, equally, no studies examined the harm of withdrawal, that is, an increase in femoral and vertebral fractures and in skeletal-related events during bone metastasis [Fn82]. Its conclusion is that the authors were unable to determine whether withdrawal before and after tooth extraction is necessary with a high certainty of evidence [Fn83].
• Another systematic review, on drug holidays from high-dose antiresorptive agents, likewise writes that there is no evidence for using a drug holiday [Fn84]; and that most of the studies recommended assessing each patient separately [Fn85].

How this passage should be read: insufficient evidence is neither the same as it should be stopped, nor the same as it should not be stopped. The first of those reviews states its own position clearly: what it provides is evidence-based information for multidisciplinary collaborations [Fn86] — which is to say that the arena for this decision lies between physicians and dentists, not in a health-education article, and still less within the scope of a patient's own judgement.

Adjacent concrete questions in this domain (each has its own canonical card; not expanded here)

- The specific timeline of extraction-wound healing and post-operative care — see canonical card KM-DENTAL-01 (in production).

- For the full domain framework of extraction and oral surgery, see domain article P06; for the implant side, see P01.

6. Antibiotic prophylaxis for endocarditis: a conceptual history

This section covers only the conceptual history and the state of the evidence. This article lists no antibiotic name, no dose, no timing and no list of who it applies to — those belong to prescribing decisions, not to health education.

6-1 The starting point: the narrowing of 2007

The 2021 scientific statement of the American Heart Association (AHA) — whose writing group was formed jointly by experts from the AHA, the American Dental Association, the Infectious Diseases Society of America and the American Academy of Pediatrics — records in its background that in 2007 the AHA published updated evidence-based guidelines on the recommended use of antibiotic prophylaxis to prevent viridans group streptococcal infective endocarditis in cardiac patients undergoing invasive procedures [Fn87]; and that those 2007 guidelines significantly scaled back the underlying conditions for which antibiotic prophylaxis was recommended, leaving only 4 categories thought to confer the highest risk of adverse outcome [Fn88].

6-2 A review more than a decade later: no reason found to reverse it

The methods and results of the same 2021 statement note that awareness of the 2007 guidelines was generally good but that adherence was variable, and that there was no convincing evidence that the frequency, morbidity or mortality of viridans group streptococcal infective endocarditis had increased since 2007 [Fn89]. On that basis its conclusion is that there are no recommended changes to the 2007 prevention guidelines [Fn90].

6-3 The shift in conceptual weight: from a single dose of antibiotic to everyday oral health

The concluding sentence of that statement also contains the most important sentence in this section: prophylaxis continues to be recommended only for categories of patients at highest risk for adverse outcome, while emphasising the critical role of good oral health and regular access to dental care for all [Fn91].

Set that sentence beside the bacteraemia data of section 1 and the logic completes itself: since toothbrushing, flossing and chewing themselves cause bacteraemia [Fn5], the bacteraemia exposure produced by everyday activities and preventive medication before a single procedure are addressing different levels of the problem — the former points to maintaining oral health over the long term, while the latter is an existing guideline measure for particular high-risk groups before particular procedures. ⚠ The data on everyday bacteraemia cannot be used to compare with, or to replace, preventive measures where the current guidelines apply; this article does not rank the two in importance, and does not advise anyone to change an established course of management on that basis.

6-4 The state of the evidence: an honest blank

This is the part of the section that most needs to be presented exactly as it stands, because the three sources point to conclusions of different strength:

The Cochrane review also records a fact of divergence at the level of health systems: guidance from the National Institute for Health and Care Excellence (NICE) in England and Wales states that antibiotic prophylaxis against infective endocarditis is not recommended routinely for people undergoing dental procedures [Fn95]. In its conclusion, the authors place the responsibility back into the clinical conversation: ethically, practitioners should discuss the potential benefits and harms of antibiotic prophylaxis with their patients before a decision is made about administration [Fn96].

The same review also explains in its background why this topic has long resisted a settled answer: the incidence of bacterial endocarditis is low, but it has a high mortality rate [Fn109] — a rare and serious event is precisely the kind that is hardest to test in a randomised controlled trial. That is also why the AHA scientific statement writes, at the very end of its own conclusion, that randomised controlled studies to determine whether antibiotic prophylaxis is effective against viridans group streptococcal infective endocarditis are needed to further refine recommendations [Fn92] — even the side issuing the recommendation writes the evidence gap next to the recommendation, and that is the sentence in this section that most deserves to be relayed in full.

The portable conclusion of this section: this is not a topic on which the question of whether it works has been settled; it is a topic on which the question of in whom, and decided by whom, keeps being redrawn [Fn90][Fn96][Fn99]. Whether an individual patient falls into a category for which it is recommended can be determined only by physicians and dentists on the basis of that person's cardiac history, and this article provides no criteria for making that determination.

7. Medications and medical history to disclose before a dental visit: a framework (not a checklist of medical orders)

7-1 Why "thorough" matters more than "the main points"

• A paper reviewing the systemic factors that affect the prognosis of implants states plainly that a thorough medical history, including prescription and over-the-counter medications, is vital [Fn3], because there may be numerous factors that could directly or indirectly influence the prognosis of dental implants [Fn106]. The same paper stresses the importance of delineating, and taking into consideration, these systemic factors from the planning phase to the restorative phase [Fn107].
• The conclusion of the cross-sectional study says the same thing from the safety side: a thorough medical history should be obtained for all dental patients to ensure safe dental care [Fn2].

Note that the wording in both sentences is "thorough" rather than "relevant" — because what counts as relevant is itself a professional judgement, and not a piece of screening the patient should carry out on the clinician's behalf.

7-2 A gap measured in practice: what patients say and what the record says are often different

A pilot study of 100 adults attending a university dental clinic [Fn134] compared the dental medication record with the medications prescribed by physicians in the electronic medical record, and found:

• About 80% of the patients consumed at least one drug (94.2% of those over 65 years), and 19% took more than five drugs (26.4% of those over 65 years) [Fn101].
• In total, 54% of the patients had some discrepancy between the medications recorded in the dental medication record and those in the medical medication record [Fn102].
• The drug classes most often omitted from the dental medication record were analgesics/opioids, antihypertensives and anxiolytics/hypnotics/sedatives [Fn103].
• The authors note that these discrepancies may be particularly common and relevant in elderly patients, in whom multimorbidity and polypharmacy are more frequent [Fn104].

This is a pilot study, at a single clinic, with 100 people [Fn134] — it cannot be treated as a general proportion. But it is enough to support one conservative behavioural suggestion: taking the actual medicine bag or medication list to a dental appointment is more reliable than answering from memory.

7-3 The framework this article can give (each row corresponds to evidence already cited above)

⚠ This table is a framework for the scope of disclosure, not a risk-assessment table, and not a list of conditions that rule treatment out. The review cited above also records honestly that no evidence was found on implant survival in patients with dementia, respiratory diseases, liver cirrhosis or osteoarthritis [Fn41] — no evidence found is not the same as no risk; it means only that this cell is blank.

8. What costs are made of, and what makes them vary (no monetary amount)

This article provides no price, fee or coverage information. This section explains only how systemic disease changes, structurally, the items of which costs are composed.

1. The cost of comanagement: the IDF/EFP consensus positions itself as comanagement between physicians, oral healthcare professionals and patients [Fn23]. Cross-specialty communication and consultation are themselves additional acts of care.
2. Assessment is not an add-on item: a thorough medical history, including prescription and over-the-counter drugs, is listed as a key prerequisite [Fn3], and in the implant setting risk assessment is required to run from before treatment through the entire lifespan of the implant [Fn32].
3. Follow-up frequency is pushed up: the evidence on the preventive direction for MRONJ comes from designs such as regular dental examinations at three-month intervals with preventive treatments [Fn77]; and periodontal and peri-implant inflammatory measures are worse when glycaemic control is poor [Fn25][Fn31] — both point to a denser rhythm of return visits.
4. Procedures may be split into more visits: dentoalveolar surgery is a common predisposing event for MRONJ [Fn70], while teeth already inflamed before extraction and additional osteotomy are listed as risk factors [Fn73], and that affects the choice of technique and the staging of visits.
5. Handling complications is a separate line: in the DOAC audit, 9.2% of procedures required specific intervention [Fn50], and return visits of that kind fall outside the original course of treatment.

Local fee systems, insurance and the boundaries of coverage lie outside the scope of this article: for local systems and costs see the corresponding canonical cards (TW) and domain article P12 (Complete guide to costs and insurance systems).

9. Overview of signals to seek care (this article compiles no symptom checklist)

This article marks its scope honestly: symptom grading and red-flag criteria form an independent domain, carried by P13 (Symptom grading and guide to seeking care); this article compiles no symptom table of its own.

### ⚠ The minimum safety net first (it does not replace the full grading in P13, but this article does not leave it out)

Most readers of this article are taking antithrombotic or antiresorptive drugs, and belong to a population at higher risk of bleeding and of poor healing. If any of the following occurs, do not wait for the next appointment; seek medical help immediately:

- Heavy bleeding, or bleeding that cannot be stopped, especially together with dizziness, weakness, pallor or fainting

- Breathing or swallowing is affected, or rapid swelling of the floor of the mouth or of the neck

- An extraction site or an intraoral wound that fails to heal for a long time, exposed bone, persistent swelling and pain, or discharge (users of antiresorptive drugs should be especially alert — the literature cited in this article observed, among its included studies, onset as late as 3 years after extraction [Fn72] — that is the observation window of that study; it does not mean that 3 years is a ceiling on risk)

>

These three lines are an editorial reservation setting a floor of safety, not grading criteria taken from any single publication; the full symptom-grading axis and the complete red-flag table are in P13. This article provides no method of self-management.

The literature in this domain supports four conceptual signals and no more:

1. Persistent bleeding after extraction is a situation that has been quantified and that has a management pathway — in the DOAC audit, 14.3% of procedures showed persistent bleeding, of which 9.2% required specific intervention [Fn50], and no patient in that population experienced major haemorrhage [Fn51]. In other words, being seen again is part of the expected process, not a loss of control.
2. Oral symptoms in users of antiresorptive drugs have a longer time window — MRONJ was observed in that study to occur up to 3 years after tooth extraction [Fn72] (this is the study's observation window, not a ceiling on risk), and it is associated with significant morbidity, an adverse effect on quality of life, and difficulty of treatment [Fn67]. This is the reason for reporting a medication history even long after the event.
3. Self-judgement is structurally unworkable in this domain — for interventions treating MRONJ, the available evidence is insufficient to either claim or refute a benefit [Fn80]; and on drug holidays, research is missing on the benefit side and on the harm side alike [Fn81][Fn82]. If even the professional side is judging case by case within an evidential blank [Fn85], it is all the less appropriate for an individual to judge alone.
4. Determining cardiac history belongs to the physician — antibiotic prophylaxis is recommended only for particular highest-risk categories [Fn88][Fn91], and whether to administer it should be decided after the practitioner has discussed the benefits and harms with the patient [Fn96].

The advice the domain level can give therefore comes down to a single sentence: anyone on any long-term medication, or with a history in any of the categories above, should hand the complete medication and medical history to the dentist before treatment [Fn2][Fn3], rather than deciding for themselves which items probably do not need to be mentioned.

10. Risk factors (indications / side effects / contraindications and limits)

Indications (in this section, situations the literature supports including in an assessment, not treatment recommendations)

• Periodontal therapy is described as safe and effective in people with diabetes [Fn15], and moderate-certainty evidence supports an improvement in glycaemic control by a clinically significant amount [Fn18].
• Implant survival in patients with type 2 diabetes has been reported as similar to that of healthy patients [Fn26][Fn37], although the positive conclusion for immediate loading carries three conditions: glycaemic level controlled, oral hygiene satisfactory, technical steps strictly followed [Fn34].
• Where anticoagulant treatment is continued, local haemostatic measures are generally effective for bleeding control [Fn55]; and no increase in immediate or delayed bleeding risk was found with aspirin monotherapy [Fn58].
• The direction of oral prevention in populations on antiresorptive therapy comes from a trial design of regular dental examinations at three-month intervals with preventive treatments [Fn77].

Possible side effects and adverse outcomes

• Peri-implant inflammation and bone loss: bleeding on probing and peri-implant bone loss were significantly worse in the diabetic group [Fn25], and hyperglycaemia was identified as an important risk factor for peri-implant inflammation [Fn27]; as glycated haemoglobin rises, advanced glycation end-product accumulation, probing depth and marginal bone loss worsen in a dose-response manner [Fn29], and osseointegration also progresses more slowly [Fn31].
• Bleeding: people on continuous VKA or DOAC treatment are at increased risk of bleeding during and after oral procedures [Fn42]; in the DOAC audit, 14.3% had persistent bleeding and 9.2% required intervention [Fn50]; and both immediate and delayed bleeding risk were higher in DAPT patients than in healthy controls [Fn59].
• Thromboembolism: dose reduction or discontinuation of anticoagulant treatment is associated with an increased risk of thromboembolism [Fn43]; and the complications of discontinuing DAPT may be more severe and fatal [Fn63].
• Osteonecrosis of the jaw (MRONJ): associated with significant morbidity and an adverse effect on quality of life, and challenging to treat [Fn67]; dentoalveolar surgery is a common predisposing event [Fn70]; in cancer patients on high-dose antiresorptive treatment the incidence after extraction ranged from 11% to 50% at the patient level [Fn71], with onset observed in the studies as late as 3 years after extraction [Fn72]; and the pooled rate after implant placement in an osteoporosis population was 0.5% [Fn75].
• Infective endocarditis: its incidence is low but its mortality rate is high [Fn109]; bacteraemia after invasive dental procedures is highest for extractions (62%–66%) [Fn4].

Contraindications and limits of application (the evidence ceiling of this article)

• Evidence gap, anticoagulation side: no eligible trials were identified in people on continuous DOAC treatment [Fn49]; a definite efficacy of antifibrinolytic therapy cannot be concluded [Fn48]; PRF did not reduce the risk of bleeding after extraction in anticoagulated patients [Fn56], with certainty of evidence ranging from moderate to low [Fn57].
• Evidence gap, MRONJ side: the evidence on preventive interventions is insufficient to either claim or refute a benefit [Fn79]; the evidence on treatment interventions is likewise insufficient [Fn80]; no study has shown a benefit of a drug holiday [Fn81], no study has examined its harms [Fn82], it cannot be determined with high certainty [Fn83], and there is no evidence supporting the use of a drug holiday [Fn84].
• Evidence gap, endocarditis side: whether antibiotic prophylaxis is effective remains without clear evidence [Fn93], at very low certainty [Fn94]; data on benefit in the moderate-risk population are insufficient [Fn99]; and the results of time-trend studies are inconsistent [Fn100].
• Population limits: the evidence that periodontal treatment improves glycaemic control is mainly in type 2 diabetes [Fn19]; data on type 1 diabetes are insufficient [Fn10].
• Safety not fully assessed: most periodontal treatment trials did not evaluate adverse effects [Fn20].
• Figures cannot be carried across populations: MRONJ at 11%–50% (high-dose cancer population, after extraction) [Fn71] and at 0.5% (osteoporosis population, after implant placement) [Fn75] measure different populations and different procedures; the prevalence of peri-implantitis spans 1.1%–85.0% within a single review [Fn36].
• Heterogeneity limits: the bidirectional meta-analysis carries high unexplained heterogeneity [Fn13]; the review of MRONJ incidence after extraction could perform only a qualitative analysis because of heterogeneity [Fn74].
• Unresearched blanks: no evidence was found on implant survival in patients with dementia, respiratory diseases, liver cirrhosis or osteoarthritis [Fn41]; osteoporosis and others have not been identified as risk factors for peri-implantitis [Fn110].
• This article contains no medication instruction: every citation involving a drug is a presentation of a study conclusion or of the state of the evidence, and constitutes no recommendation to stop, to continue, to dose or to prescribe.

⚠ This section is a disclosure of medical risk and does not constitute an individual treatment recommendation. Actual treatment and its results vary from person to person and must be assessed by a dentist.

Every clinical statement in this article is mapped line by line to its cited source (see the sources and evidence chain below). It has not been clinically reviewed by a licensed practitioner. This is health information, not individual advice; assessment by a clinician is required.