When to save a tooth and when to replace it: a specialist's decision framework for failed root canal treatment

Last year I retreated three teeth that had been recommended for extraction. Two were upper first molars, both with separated instruments left from the original treatment. One was a lower second premolar where the original obturation had stopped 4 mm short of the apex. All three healed. All three were saved at roughly half the cost the patient had been quoted for extraction, implant placement, and crown.

This is not a heroic anecdote. It is a frequency claim. In my own practice — 23 years, around 14,000 cases — I see this pattern weekly. The decision between saving a tooth and replacing it with an implant is one of dentistry’s most common high-stakes calls. It is also, in 2026, one of the most consistently mishandled, in both domestic and international markets, for reasons I will name.

This piece is the decision framework I use when a patient walks in with a failing root canal and asks what to do. It is not individual treatment advice. It is the structure of the decision so you can evaluate the recommendation you have received.

The three options, named honestly

There are exactly three options for a tooth with a previously failed root canal:

1. Orthograde retreatment. Re-enter the canal system through the existing access, remove the previous filling material, find any missed canals, disinfect, re-obturate. Done by a specialist endodontist with magnification, the published five-year tooth-survival rates sit in the 78–86% range across systematic reviews, with retreatment success (resolution of periapical lesion) around 77% [1, 2].
2. Microsurgical retreatment (apicoectomy). Surgical access through the bone, remove the apical 3 mm of the root, place a retrograde filling. Indicated when orthograde retreatment is not feasible — calcified canals, post that cannot be removed safely, persistent infection after orthograde retreat. Success rates with modern microsurgical technique (microscope, ultrasonic preparation, MTA or bioceramic retro-fill) sit at 89–94% at one year, dropping to 74–88% at long-term follow-up [3].
3. Extraction and implant. Pull the tooth, place a titanium fixture into the bone, restore with an abutment and crown. Implant survival at five years is around 95% in selected cohorts, but survival is not success — peri-implantitis prevalence at the 9–14 year mark sits between 18% and 28% in registry data, and there is no equivalent of “retreatment” for a failed implant. You explant and graft.

The first option — orthograde retreatment by a specialist — is the under-used one. The third option is the over-prescribed one. The reason is not clinical. It is structural.

Why extraction is over-prescribed

Three reinforcing pressures push the recommendation toward extraction-and-implant when retreatment would work:

Time. A specialist endodontic retreatment takes 90–150 minutes per visit, often across two visits. A general dentist who quotes retreatment is committing more chair time at a lower margin than an extraction-implant pathway with a referred surgical placement.

Skill ceiling. Retreatment is harder than primary treatment. Locating a missed MB2 canal in an upper first molar, removing a separated file from a curved root, bypassing a ledge — these are specialist procedures requiring an operating microscope, ultrasonic instrumentation, and a body of case experience the average general dentist has not built. The honest constraint, often unstated: “I cannot do this case well, so I will recommend the option I can do.”

Margin. The full pathway of extraction → graft → implant → abutment → crown generates 3–5x the practice revenue of a retreatment + new crown. In high-volume international markets where the business model is throughput, this incentive is amplified, not muted.

I am not saying every dentist who recommends extraction is acting on these incentives. I am saying that when you receive a recommendation to extract and replace, the question to ask is: was retreatment by a specialist endodontist considered, and if so, on what specific clinical grounds was it ruled out? If the answer is vague — “the prognosis isn’t good” — that is not a clinical answer. That is a placeholder.

The actual decision criteria

A tooth is a candidate for retreatment if all four of these hold:

• Restorability. Enough sound coronal tooth structure remains, after caries removal, to support a ferrule of at least 1.5–2 mm circumferentially. Without ferrule, no crown will hold long-term, and the retreatment is futile.
• Periodontal support. Bone support to within 3 mm of the apex on at least one root, with no Grade 3 mobility, no Class III furcation involvement on multi-rooted teeth.
• Identifiable cause of failure. Missed canal, inadequate obturation length, coronal leakage from a failed restoration — these are correctable. Vertical root fracture is not. CBCT imaging identifies vertical root fracture in roughly 70% of cases pre-operatively; the remaining 30% are diagnosed at re-entry and the tooth is then extracted. This is acceptable; the alternative is extracting teeth that would have been saved.
• Patient-side factors. The patient understands the procedure takes two visits, costs roughly $2,200–3,500 AUD in Australia for the retreatment plus a new crown at $1,800–2,400 AUD, and carries a 15–20% risk of failure that would then require microsurgery or extraction.

If any of these four fails, the tooth is a candidate for extraction-and-implant. If all four hold, retreatment is the first option to consider, not the second.

The cost numbers, with currency and what’s-included

For a single upper first molar, failing previous root canal, otherwise restorable:

Numbers reflect treatment-only costs from quoted prices at four Australian specialist endodontic practices and four Vietnamese clinics holding ISO certification, Q2 2026. Excludes CBCT imaging (~$220 AUD / ~$60 USD), travel, accommodation, and the consultation visit. Implant numbers assume premium-tier fixtures (Straumann, Nobel Biocare); economy-tier fixtures reduce the Australian numbers by roughly 20% and the Vietnamese numbers by roughly 35%.

The international price differential is real, the implant pathway shows the largest absolute saving, and that is — for the patient comparing quotes on a spreadsheet — exactly the trap. The pathway with the highest savings is the pathway most likely to be over-prescribed in the first place. The cheaper route is often a pathway you should not have been on.

When the implant is genuinely the right answer

I want to be specific about when I do recommend extraction-and-implant for a failed root canal:

• Vertical root fracture confirmed. No retreatment will work. The tooth has to come out.
• Inadequate ferrule after caries removal. The crown has nothing to hold onto. Even a successful retreatment will fail mechanically within five years.
• Strategic considerations in a multi-tooth treatment plan where the tooth’s contribution to occlusal load is low and an implant fits a broader prosthodontic design.
• Patient preference after informed consent, where the patient understands the alternative, the costs, the success rates of each, and chooses the implant pathway anyway. This is a legitimate choice. It just needs to be a choice, not a default.

In around 25–30% of the failed-root-canal cases that come into my practice, extraction-and-implant is the right answer. In the other 70–75%, retreatment is the first option to consider, and in roughly half of those it is the better option full stop. That ratio — implant in three of ten, retreatment in seven of ten — is approximately the inverse of the recommendation distribution patients arrive carrying.

What would change my view

I hold this view because of (a) my own clinical caseload over 23 years and (b) the systematic reviews cited above. The evidence that would update it:

• A registry-grade five-year cohort study, N>1,000, comparing matched cases (same tooth, same baseline restorability, same periodontal support) randomised to retreatment vs extraction-and-implant, with tooth survival, peri-implantitis incidence, and patient-reported outcome measures as endpoints. No such study exists. The closest comparators are non-randomised cohort studies with significant selection bias.
• A demonstration that modern implant designs have reduced peri-implantitis prevalence below the 18–28% range at long-term follow-up. Current data do not show this.
• A demonstration that the cost differential between retreatment and implant in international markets is large enough to offset the additional risk of complication, revision, or peri-implant disease over a 15-year horizon. The cost differential is real, but it does not, in the data I have seen, offset the difference.

Until those exist, the framework above is the framework I will continue to apply.

How to evaluate the recommendation you have received

If you have been told to extract and replace, ask:

1. Was the case reviewed for orthograde retreatment by a specialist endodontist? If not, ask for that referral specifically.
2. What is the specific clinical reason retreatment was ruled out — not “the prognosis isn’t good” but a named anatomical or restorative finding (vertical root fracture on CBCT, inadequate ferrule, calcified canals, separated instrument that cannot be bypassed).
3. What CBCT imaging supports the diagnosis? If no CBCT was taken, ask why.
4. What is the cost of retreatment + new crown at this clinic, compared to the implant pathway, with currency, dates, and what’s-included specified for both?
5. Where would the retreatment be performed if you chose it — by the same dentist, or referred to a specialist endodontist?

A clinician who answers all five clearly is a clinician you can trust on this question. A clinician who answers any of them with a wave of the hand is not, on this question, a clinician whose recommendation should be taken at face value. That is true in Sydney. It is true in Da Nang. It is true in Bangkok. It is true everywhere.

For complex full-arch reconstruction or multi-implant strategic cases, this single-tooth framework does not apply, and you should be reading the all-on-X framework piece when it publishes. If your treatment plan calls for bone grafting before placement, the question of whether you actually need it is its own decision; see why most dental implants do not need bone grafting. For the cost numbers across ten countries with the trip-cost framework attached, see the implant cost comparison reference. For the patient-side decision tree on whether to travel at all, see the dental tourism trust gap long read.