Few absolutes, many modifiers
Risk stratification is the discipline of deciding whether and how to place an implant before any drill touches bone. A clean medical history is rarely a yes-or-no question. The mistake of the inexperienced clinician is to read systemic disease as a list of prohibitions; the reality is that the great majority of medical conditions raise risk rather than forbid treatment. The clinician's task is therefore not to memorize a roster of contraindications but to sort each finding into the bucket that determines its disposition — and to recognize that most findings are modifiers the patient and clinician can work to neutralize.2
This chapter organizes that judgment around three buckets and three verdicts. Findings divide into absolute / defer risks — the rare true contraindications that stop the pathway; modifiable / optimize risks — the common conditions that elevate failure rate but respond to optimization; and local / site risks — anatomical or infective limitations addressed surgically. Each finding maps to one of three actions: proceed, optimize (then proceed), or decline / defer. The logic is sequential and conservative: a single absolute finding halts the pathway regardless of how favorable everything else looks, while modifiable risks route first to an optimization phase and only then to surgery.1
Absolute / defer findings (IV oncologic antiresorptives, recent head-and-neck radiotherapy to the site, unstable cardiac disease, active malignancy under treatment, skeletal immaturity) stop the pathway. Modifiable / optimize findings (smoking, diabetes, oral bisphosphonates, untreated periodontitis, poor hygiene) elevate risk but are correctable before surgery. Local / site findings (insufficient bone, nerve or sinus proximity, restricted space, active local infection) are managed surgically. The pathway runs the patient against the buckets in order of severity: the highest-tier finding present governs the verdict.
Sorting risk into three buckets
The first move at screening is classification, not decision. Before weighing how aggressively to optimize a smoker or how to graft a deficient ridge, the clinician must establish which category each finding belongs to, because category dictates the route. An absolute finding renders the modifiable and local analysis moot; a purely local finding does not invoke medical deferral. The table below collects the canonical findings of each bucket — the working vocabulary of candidacy.
| Bucket | Representative findings | Default route |
|---|---|---|
| Absolute / defer | High-dose IV / oncologic antiresorptives; recent head-and-neck radiotherapy to the site; recent MI or stroke / unstable cardiac disease; active malignancy under treatment; skeletal immaturity (growth incomplete) | Decline / defer |
| Modifiable / optimize | Smoking (RR ≈ 1.9 for failure); diabetes (target HbA1c ≤ 7–8%); oral bisphosphonates (MRONJ ≈ 0.5%); untreated periodontitis; poor oral hygiene / compliance | Optimize, then proceed |
| Local / site | Insufficient bone volume or quality; proximity to inferior alveolar nerve or maxillary sinus; limited inter-arch or mesiodistal space; active infection at site | Manage surgically |
Why the buckets are ordered, not parallel
The buckets are not weighed against one another like items on a balance; they are consulted in descending order of gravity. A patient on high-dose IV antiresorptives for metastatic disease is not made a candidate by a pristine ridge and impeccable hygiene. Conversely, a healthy non-smoker with a deficient ridge is not medically deferred — the constraint is local and surgical. This ordering is what lets a busy screening visit reach a defensible verdict quickly: identify the highest-tier finding present, and let it govern.2
From dominant risk to verdict
The interactive selector below operationalizes the ordered logic. Run the patient against the three buckets in sequence and select the highest-tier finding present. A single absolute finding stops the pathway; a modifiable risk routes to an optimization phase before any commitment to surgery; absence of either yields a low-risk, standard-protocol verdict. The tool returns the corresponding pathway — proceed, optimize-then-proceed, or defer / decline — with the concrete next steps for each.
Interactive candidacy selector
Select the dominant risk finding to reveal its pathway, the rationale, and the ordered next steps.
The middle pathway is the one that distinguishes thoughtful candidacy from reflexive refusal. A modifiable finding does not mean "no" — it means "not yet, and here is the work." Cessation counseling for the smoker, glycemic control for the diabetic, periodontal stabilization for the perio patient: each is a defined optimization step that, once met, converts the patient into a candidate. Document the elevated risk in consent, intensify maintenance, and proceed once the target is reached.1
- Treating oral bisphosphonates as if they were IV oncologic antiresorptives — declining a patient whose MRONJ risk is roughly 0.5% rather than consenting and proceeding atraumatically.3
- Reading "diabetes" as a single category. Well-controlled diabetes (HbA1c ≤ 7–8%) proceeds on routine protocol; poorly controlled diabetes is optimized first. The meta-analytic signal for diabetes as an independent failure risk is, in fact, weak when control is adequate.1
- Placing into untreated active periodontitis, importing the patient's dysbiotic biofilm and inflammatory burden onto a fresh implant surface.
- Quoting population risk figures as individual predictions during consent. The relative risk of smoking (≈ 1.9) is a population estimate, not a guarantee for the patient in the chair.
Common conditions at a glance
Most conditions modify rather than prohibit. The table below collapses the screening logic into a chairside reference: each condition resolves to a status and a concrete action, with the strength of the underlying evidence noted. Verify each finding against current medical guidance and coordinate with the patient's physician where indicated — risk figures are population estimates, not individual predictions.
| Condition | Status | Action | Evidence |
|---|---|---|---|
| Smoking | Modifiable | Cessation counseling; document elevated failure risk (RR ≈ 1.9) | Meta-analysis |
| Diabetes (well-controlled) | Proceed | Confirm HbA1c ≤ 7–8%; routine protocol | Syst. review |
| Diabetes (poorly controlled) | Optimize first | Defer until glycemic control improved | Consensus |
| Oral bisphosphonates | Caution | Consent for MRONJ (~0.5%); atraumatic technique | Position paper |
| IV / oncologic antiresorptives | Avoid | Generally contraindicated; coordinate with oncology | Position paper |
| Head & neck radiotherapy | High risk | Site/dose-dependent; specialist referral, weigh ORN risk | Meta-analysis |
| Treated periodontitis | Optimize first | Stabilize before placement; strict maintenance | Syst. review |
Of the population-level risk factors quantified in the pooled literature, smoking (RR ≈ 1.9) and radiotherapy to the site (RR ≈ 2.3) carry the largest, most reproducible effects on implant failure; well-controlled diabetes, by contrast, shows no clear independent increase in pooled estimates. This is why cessation counseling and a careful radiotherapy history earn disproportionate attention at screening — they are the modifiable levers with the most evidence behind them.1
Key terms
- Absolute contraindication
- A finding that prohibits routine implant therapy outright (e.g., high-dose IV oncologic antiresorptive therapy); stops the candidacy pathway.
- Relative (modifiable) risk factor
- A condition that elevates the probability of failure but can be reduced through optimization before surgery (e.g., smoking, hyperglycemia).
- Local / site risk
- An anatomical, volumetric, or infective limitation at the proposed site, addressed surgically rather than by medical deferral.
- MRONJ
- Medication-related osteonecrosis of the jaw — exposed or probeable necrotic bone in a patient on antiresorptive or antiangiogenic medication, persisting > 8 weeks, without a history of head-and-neck radiation.
- ORN (osteoradionecrosis)
- Non-healing irradiated bone exposed for > 3 months without tumor recurrence; the principal risk driving deferral after head-and-neck radiotherapy.
- HbA1c
- Glycated hemoglobin; an index of average glycemia over ~3 months, used to gauge whether a diabetic patient is sufficiently controlled (target ≤ 7–8%) to proceed.
- Relative risk (RR)
- The ratio of failure probability in exposed versus unexposed groups; a population estimate, not an individual prediction.
- Optimization phase
- The defined interval in which a modifiable risk is treated to target before implant surgery proceeds.
Board & fellowship preparation
- Give me an example where a perfect ridge does not make the patient a candidate.
- Why are the buckets ordered rather than scored against each other?
- What is MRONJ and how is it defined?
- Does a drug holiday change your decision for the oral patient?
- What does the meta-analytic evidence actually show for diabetes versus smoking?
- How does poor glycemic control impair osseointegration mechanistically?
- What dose threshold raises your concern, and why?
- How does ORN differ from MRONJ in pathophysiology?
- What would make you escalate the smoker from optimize to defer?
- How do you phrase the RR figure honestly during consent?
References
- Chen H, Liu N, Xu X, Qu X, Lu E. Smoking, radiotherapy, diabetes and osteoporosis as risk factors for dental implant failure: a meta-analysis. PLoS One. 2013;8(8):e71955. doi:10.1371/journal.pone.0071955. PMID: 23940794
- Diz P, Scully C, Sanz M. Dental implants in the medically compromised patient. J Dent. 2013;41(3):195–206. doi:10.1016/j.jdent.2012.12.008. PMID: 23313715
- Ruggiero SL, Dodson TB, Aghaloo T, Carlson ER, Ward BB, Kademani D. American Association of Oral and Maxillofacial Surgeons' position paper on medication-related osteonecrosis of the jaws — 2022 update. J Oral Maxillofac Surg. 2022;80(5):920–943. doi:10.1016/j.joms.2022.02.008. PMID: 35300956
Evidence grades: Systematic review / meta-analysis Consensus / position paper Preclinical. Risk figures are population estimates, not individual predictions.