When to Refer to an Endodontist

Endodontic referral decisions get made under time pressure, chairside, with a patient who wants an answer now. The hard part is rarely whether the tooth needs root canal treatment. It is whether this particular tooth, in this particular patient, on this particular day, is a case you should start.

The specialty exists for the cases that are harder than average. Endodontics is defined around the biology, pathology, diagnosis and treatment of the dental pulp and the periradicular tissues⁸, and the American Association of Endodontists frames the referral question as one of difficulty rather than one of competence — a case carries features that raise the risk of an unfavourable outcome, and those features can be listed in advance rather than discovered mid-appointment¹.

Difficulty, not competence

The AAE publishes an Endodontic Case Difficulty Assessment Form, available through its case assessment tools, which exists to make this judgement explicit instead of intuitive. It asks the clinician to weigh a case against a defined set of factors and arrive at a level of difficulty, and the AAE states that it may be used to assist with referral decision making and record keeping¹.

Rather than reproduce the form — it is short, free, periodically revised, and best read in its current version at source — it is worth knowing the kinds of thing it weighs. The factors fall into three families¹:

• Patient considerations — medical history, the ability to tolerate and cooperate with a long appointment, anaesthesia history, and how far the mouth actually opens
• Diagnostic and treatment considerations — how settled the diagnosis is, the position of the tooth in the arch, the state of the crown, canal and root morphology, radiographic findings, resorption, and whether the tooth has been treated before
• Additional considerations — trauma history and the complicating circumstances that sit outside the tooth itself

The point of a tool like this is not the score at the bottom. It is that difficulty gets assessed before access rather than during it, which is why the AAE files case assessment under treatment planning². A case reassessed at the moment the pulpal floor turns out to be a sclerosed maze is a case that has already spent its easy options.

Findings that move a case across the line

Some findings raise difficulty enough that most general practices treat them as a referral trigger on sight. None of them are absolute — a practice with an operating microscope and a steady endodontic caseload draws the line somewhere different from one that treats a handful of molars a year — but these are the recurring ones.

• Calcified or sclerosed canals, where the chamber and the canal orifices cannot be located readily
• Curved, dilacerated or S-shaped roots, and roots with an abrupt apical curvature
• Additional or aberrant canal systems — the second mesiobuccal canal is the standing example, but it is not the only one
• An open apex or incomplete root development, where there is no apical stop to work against
• Internal or external resorption, including cervical resorptive defects
• A previously accessed tooth with a missed canal, or a canal obstructed by a ledge, a post or a separated instrument

That last group deserves emphasis, because it is where a referring practice most often revisits its own judgement. A tooth still symptomatic after treatment that looked complete on the radiograph is a diagnostic problem before it is a technical one, and answering it usually means imaging you may not have. The AAE describes cone beam computed tomography as giving three-dimensional views of the teeth and surrounding structures that enhance the endodontist’s ability to diagnose, evaluate and treat certain cases⁴. If the answer turns on a scan you cannot take, that is itself a reason to refer.

When the diagnosis is the thing you cannot get

Persistent pain with no identifiable cause is a referral indication in its own right, and it is badly under-used as one. A patient who has had two teeth treated in a quadrant and still hurts is not a patient who needs a third tooth treated. Diagnosis of pulpal and periradicular disease is a defined part of what the specialty covers⁸, so "I cannot localise this" is a legitimate, well-formed reason for referral — arguably a better one than "lower molar, difficult anatomy".

Suspected cracks belong in the same paragraph. They are difficult to see, difficult to stage, and described inconsistently from one practice to the next; the AAE maintains a Glossary of Endodontic Terms partly because shared vocabulary is not a given⁶, and its clinical documents include guidance on the methodology of cracked tooth studies³. If you suspect a crack but cannot say where it runs or how far it goes, write exactly that on the referral. Stated uncertainty is more useful to the endodontist than a confident guess.

Retreatment, surgery, or extraction

The hardest referral is the one where the alternative to endodontic treatment is removing the tooth. Three routes are usually on the table for a tooth that has failed previous treatment, and they are not interchangeable.

1. 1Non-surgical retreatment

   The existing root filling is removed and the canal system re-treated through the crown. It is squarely specialist territory once posts, ledges or separated instruments are involved.

2. 2Surgical retreatment

   Access to the root end through bone, rather than through the crown. The AAE’s clinical documents include position statements covering surgical and non-surgical endodontic management³.

3. 3Extraction and implant

   A different specialty, a different timeline, and a different set of consequences for the adjacent teeth and the surrounding bone. The AAE maintains a position statement on implants among those same clinical documents³, because the comparison gets made often and sometimes casually.

Who makes that call matters at least as much as which way it goes. A tooth judged hopeless without an endodontic opinion has been judged by whoever happened to see it first. The workable division is that the referring dentist frames the restorative question — what is this tooth being kept for, is it restorable, does the periodontal support hold — and the endodontist answers the endodontic one, and the patient hears the trade-off in plain terms before anything is booked. Where the case is genuinely borderline, that is a consultation, and the referral should say so: consultation-only versus consultation-and-treatment is the most under-specified field on a referral form, as we set out in what a complete dental referral contains.

Patient factors that decide the case on their own

The tooth can be straightforward and the case still not be. The AAE’s assessment framing puts patient factors in the same weighing as anatomy¹, which is the right instinct: an accessible canal in a mouth that barely opens is not accessible.

• Limited opening, from trismus, temporomandibular disorder or prior surgery — it constrains instrument angle before it constrains anything else
• Medical complexity that affects appointment length, positioning, or how the visit has to be managed
• Anxiety severe enough that the appointment itself is the obstacle, including patients for whom sedation is part of any realistic plan
• A gag reflex that prevents radiographs or rubber dam placement
• Inability to tolerate a long appointment, which turns a single-visit case into a sequencing problem

These are also the factors most likely to be knowable before the patient is in the chair, from the medical history and from the last two visits. They are the cheapest referral triggers to act on early and the most expensive to discover at the point of access.

Trauma, where the clock decides

Dental trauma is the category least like the others, because the variable that dominates is elapsed time rather than case difficulty. The AAE publishes recommended guidelines for the treatment of traumatic dental injuries alongside supporting resources, and states plainly that the type and severity of the injury determine the treatment required⁵. Avulsion and luxation injuries sit at the sharp end of that: the sequencing is time-dependent in a way ordinary endodontic planning is not.

Two consequences for referral workflow follow. First, an avulsed or luxated permanent tooth is a phone call, not a referral form — the receiving practice needs to know the patient is coming and when the injury happened, and neither of those facts survives a queue. Second, the pathway has to be arranged before you need it: know which endodontic practices within reach accept emergencies, and who answers after hours. Working that out during the call is how the hour gets spent.

What an endodontic referral should carry

The general field list applies — identity, contact details, insurance, and a clinical reason stated as a finding rather than as a procedure request. Endodontics adds several of its own, and they are the ones whose absence causes retakes and callbacks.

• Pre-operative periapical radiographs, dated — and the CBCT volume itself if one was taken, not merely a report of it
• Pulp testing results, thermal and electric, with percussion and palpation findings
• Symptom history: duration, what provokes it, whether it wakes the patient, whether it has changed
• Any prior endodontic treatment on that tooth, with dates, and what was attempted
• Your restorative plan, and whether you consider the tooth restorable — the endodontist is being asked to save a tooth for a purpose, and the purpose changes the advice
• Whether you want consultation, treatment, or both, and who is placing the definitive restoration

Sending images matters more here than almost anywhere else. The ADA’s guidance on dental radiographs is built around individualised professional judgement rather than fixed intervals, and around imaging only where there is a clear clinical benefit⁹. A referral that arrives without images generally means the specialist retakes them — a second exposure the first set would have made unnecessary, and a baseline lost, since yours was taken before anything changed.

The coronal seal usually remains the referring practice’s responsibility, so the report you want back should state when it is safe to restore. That return leg is its own distinct failure point, and we treat it separately in closing the referral loop.

What this is, and what it is not

This is guidance about workflow. It describes how the referral decision tends to get made, what the profession’s own case-assessment tools weigh, and what information makes a referral usable on arrival. It prescribes no treatment, and it is not a substitute for the current text of the AAE’s clinical resources⁷, which are revised over time and should be read at source.

Case selection and every treatment decision that follows belong to the treating clinician, who is the only person holding the radiographs, the medical history and the patient. The ADA makes the same point about imaging: the dentist who knows the patient’s history is the one positioned to make the judgement, and published recommendations are not standards of care⁹. A referral threshold is a personal number, set by training, equipment, caseload and honest self-assessment, and it moves across a career. The most an article can usefully do is make sure the factors are on the table before the access cavity is.