# The Research Says: Should We Use Posts or Not?

> A post doesn't strengthen a root-canal-treated tooth. Naumann's systematic review and five others show remaining walls and ferrule are what decide survival.

- Author: Ryan Smith, DDS
- Published: 2026-07-16
- Category: The Research Says
- Canonical: https://ryansmithdds.com/posts-after-root-canal-research

---

> **The verdict**
>
> A post does not strengthen a root-canal-treated tooth. Its only job is retaining a core when there isn't enough coronal dentin left to do that on its own. Across six independent systematic reviews, the studies that isolate post-versus-no-post keep landing in the same place: remaining structure and ferrule predict survival, and adding a post to a tooth that already has two sound walls and a ferrule buys you nothing measurable.
>
> Evidence strength: **Moderate**: six systematic reviews, no dedicated large RCT.

I finish the canal, irrigate, dry the chamber, and look at what's left of the tooth. Two intact walls, decent ferrule, a normal-sized access. And somewhere in the back of my head a voice (training from fifteen years ago) says *put a post in it, just to be safe.*

That voice is wrong. Some version of this argument comes up at almost every study club: half the room places a post in nearly everything endodontically treated, out of habit or liability anxiety; the other half knows that's outdated but can't always say exactly where the line sits. Ask ten dentists when a post is necessary and you get ten answers, most some flavor of "it depends": true, but not useful.

Here's the point a lot of us were never taught cleanly: a post does not reinforce a root. It doesn't strengthen dentin or stiffen the tooth, and an over-prepared post space can remove structure you needed. A post has exactly one job: retaining a core when the remaining coronal structure can't retain one on its own. What determines long-term survival is how much sound structure is left, and whether you can achieve a ferrule. So the real question was never "should I place a post?" It's "does this tooth have enough structure to retain a core without one?" Here's what the literature says.

## The question

The comparison this article answers: in adults with an endodontically treated permanent tooth needing a full coronal restoration, does placing an intraradicular post before core buildup, versus a bonded core alone, improve restoration or tooth survival, and does that answer change with remaining coronal structure (intact axial walls, a circumferential ferrule)? Given how small these trials run, a meaningful difference is a gap several independent reviews find in the same direction even without significance in any one study: exactly the pattern for the post's own effect, and the opposite pattern for remaining structure.

## The mechanism

A ferrule is a band of sound tooth structure, roughly 2 mm tall, that a crown margin fully encircles below the core. When present, the crown grips the tooth like a metal band around a barrel: forces distribute into that band of dentin instead of concentrating at the core-tooth interface, which is what resists flexing and cracking under function. A post sits below all of that. It doesn't touch the crown margin, doesn't resist lateral flexure, and can't substitute for a ferrule it's structurally incapable of creating. What it can do is give a core material something to grip inside the root when there isn't enough coronal dentin to bond onto otherwise. That's the whole mechanism: reinforcement was always the wrong mental model.

## What the evidence says

### Naumann et al., 2018: systematic review of post-vs-no-post trials

**What they did:** Searched every prospective human study, through mid-2017, directly comparing teeth restored with a post to teeth restored without one over a minimum 5-year observation. Seven RCTs and one prospective trial met that bar.

**What they found:** Seven of the eight included studies showed no positive effect of the post on survival [1]. Two of three studies testing ferrule separately supported the ferrule-effect concept. In the authors' words: "ferrule effect and maintaining cavity walls are the predominant factors," and "most studies do not confirm a positive effect of post placement" [1].

**Why it matters:** The closest thing this literature has to a direct answer: built on studies that isolated the post as the variable, with ≥5-year data where early success curves have had time to diverge.

### AlSaleh et al., 2021: systematic review and meta-analysis of remaining axial walls

**What they did:** Five studies (two RCTs, three cohorts) of posterior root-filled teeth restored with a single crown and bonded fiber post, stratified by remaining axial walls.

**What they found:** Pooled failure was 23% (95% CI, 10%–36%) for zero remaining walls and 15% (95% CI, 3%–26%) for one remaining wall: failure rose as walls were lost [2], and fiber post debonding tracked the same trend.

**Why it matters:** The one review that puts a number on "how much structure is enough." Even one extra wall of dentin measurably drops failure: a structural effect, since every tooth here already had a post.

### Al-Dabbagh et al., 2024: systematic review and meta-analysis of circumferential ferrule with fiber posts

**What they did:** Eight studies (407 teeth) on ferrule presence in fiber-post-restored teeth; two of the eight (123 teeth) allowed a direct head-to-head comparison.

**What they found:** In the direct comparison, a uniform circumferential ferrule produced a significant increase in success (risk ratio 1.28, 95% CI 1.06–1.54) [3]. Across all eight studies indirectly, there was no significant difference.

**Why it matters:** The ferrule effect is real when isolated cleanly: 123 teeth is a reminder of how thin the direct evidence base still is.

### Batista et al., 2020: systematic review and meta-analysis of ferrule and fiber posts

**What they did:** Four prospective or randomized studies (297 teeth: 157 with ferrule, 140 without), minimum 6-month follow-up, comparing failure of fiber post-and-core restorations with and without ferrule.

**What they found:** Survival was 88.35% with ferrule versus 78.05% without: the expected direction, but not statistically significant (risk ratio 0.71, 95% CI 0.47–1.06, P=.09) [4].

**Why it matters:** Another vote in the same direction, again underpowered. The pattern across this literature: independent groups keep finding the same effect, with no single trial big enough to prove it alone.

### Tsintsadze et al., 2024: meta-analysis of glass-fiber vs metal posts

**What they did:** RCTs with at least 2 years of follow-up comparing survival of teeth restored with glass-fiber versus metal posts; seven studies were selected from 184 identified.

**What they found:** Pooled survival was 92.8% for glass-fiber posts versus 78.1% for metal posts: a large numeric gap, not statistically significant. The authors concluded both "are reliable materials when a significant amount of coronal tooth structure is missing and treatment with a post is indicated" [5].

**Why it matters:** For teeth that genuinely need a post, fiber has the numeric edge and, as the next study shows, fails more safely.

### Figueiredo et al., 2015: meta-analysis of metal vs fiber posts and root fracture

**What they did:** Seven RCTs and seven cohort studies, over 5 years of follow-up, on root fracture (the catastrophic, usually unrestorable failure) comparing metal and fiber post-retained restorations.

**What they found:** Pooled survival was 90% for metal posts versus 83.9% for fiber posts overall, not significantly different. But prefabricated metal posts and carbon fiber posts each had roughly double the root-fracture incidence of cast metal and glass fiber posts, respectively [6].

**Why it matters:** Top-line survival hides the failure mode a patient feels: a debonded fiber post is a redo; a root fracture can mean an extraction.

## Evidence at a glance

| Study | Design | Sample | Follow-up | Finding | Weight |
|---|---|---|---|---|---|
| Naumann et al., 2018 [1] | Systematic review of RCTs | 8 studies | ≥5 years | 7 of 8 studies show no post-survival benefit | High |
| AlSaleh et al., 2021 [2] | Systematic review + meta-analysis | 5 studies | Variable, ≥1 year | Failure 23% (0 walls) vs 15% (1 wall) | Moderate |
| Al-Dabbagh et al., 2024 [3] | Systematic review + meta-analysis | 8 studies (407 teeth) | ≥6 months | Ferrule: RR 1.28 success (direct analysis only) | Moderate |
| Batista et al., 2020 [4] | Systematic review + meta-analysis | 4 studies (297 teeth) | ≥6 months | 88.4% vs 78.1% survival, not significant | Moderate |
| Tsintsadze et al., 2024 [5] | Meta-analysis of RCTs | 7 studies | ≥2 years | 92.8% (fiber) vs 78.1% (metal) survival | High |
| Figueiredo et al., 2015 [6] | Meta-analysis of RCTs + cohorts | 14 studies | ≥5 years | Metal/carbon posts ~2× root fracture rate | Moderate |

*Pooled survival of endodontically treated teeth restored with a post, by post material: a large numeric gap that did not reach statistical significance. Data from [5].*

*Pooled failure rate of posterior root-filled teeth restored with a fiber post and single crown, by number of remaining axial walls. Data from [2].*

## Where the evidence is thin

None of this is built on the trial you'd actually want: a large, prospective RCT randomizing post versus no post while stratifying cleanly by wall count and ferrule height, over a decade. What exists is six overlapping systematic reviews on a shared, small pool of primary studies: the largest head-to-head comparison is 407 teeth, and the base for isolating ferrule cleanly is smaller still, at 123. Cochrane's own review here (Bolla et al.) was withdrawn from the library; while active it pooled only two trials and 317 participants and said it "could not specify which type of post and core system should be used when two or three dentine walls remain." Most trials come from academic referral centers, and "success" isn't defined identically across studies. Anterior teeth are underrepresented, and nobody has published cost-effectiveness data on a fiber post kit against the downstream cost of an occasional redo.

## What this means Monday morning

1. **Assess remaining structure first.** After canal prep, check how many sound walls remain and whether you can achieve roughly 2 mm of circumferential ferrule.
2. **Two or more walls, achievable ferrule → skip the post.** Build the core directly. There's nothing to gain from a post here, and one less material cost and debonding point.
3. **One wall, or ferrule compromised → place a fiber post.** This is where AlSaleh's data shows failure climbing.
4. **Zero walls, large access, or high esthetic demand → fiber post, default material.** Fiber wins numerically and fails more safely.
5. **Reserve metal or cast posts for exceptions:** flared canals with no ferrule, unusual anatomy, or a crown too short for a fiber post's length.
6. **Update the script.** "This tooth has enough structure left that a post won't add anything" is more honest than "we'll add a post for extra strength", and it's what the evidence supports.

## The bottom line

A post retains a core; it does not reinforce a tooth. Six independent systematic reviews keep landing on the same structural factors (remaining walls and ferrule) as what predicts survival, even though no single trial is large enough to prove it beyond doubt. Treat the post as a retention decision made after honestly assessing what's left of the tooth, not a reflex for every endodontically treated case.

## References

1. Naumann M, Schmitter M, Frankenberger R, Krastl G. "Ferrule Comes First. Post Is Second!" Fake News and Alternative Facts? A Systematic Review. *Journal of Endodontics*. 2018;44(2):212-219. [doi:10.1016/j.joen.2017.09.020](https://doi.org/10.1016/j.joen.2017.09.020) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/29229457/)
2. AlSaleh E, Dutta A, Dummer PMH, Farnell DJJ, Vianna ME. Influence of remaining axial walls on root filled teeth restored with a single crown and adhesively bonded fibre post: A systematic review and meta-analysis. *Journal of Dentistry*. 2021;114:103813. [doi:10.1016/j.jdent.2021.103813](https://doi.org/10.1016/j.jdent.2021.103813) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/34530058/)
3. Al-Dabbagh RA, Sindi MA, Sanari MA, Manna AI, Al-Dabbagh MA. Effect of a circumferential ferrule on the survival and success of endodontically treated teeth restored with fiber posts: A systematic review and meta-analysis. *Journal of Prosthetic Dentistry*. 2024;132(6):1251-1259. [doi:10.1016/j.prosdent.2023.12.002](https://doi.org/10.1016/j.prosdent.2023.12.002) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/38216376/)
4. Batista VES, Bitencourt SB, Bastos NA, Pellizzer EP, Goiato MC, Dos Santos DM. Influence of the ferrule effect on the failure of fiber-reinforced composite post-and-core restorations: A systematic review and meta-analysis. *Journal of Prosthetic Dentistry*. 2020;123(2):239-245. [doi:10.1016/j.prosdent.2019.01.004](https://doi.org/10.1016/j.prosdent.2019.01.004) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/31227236/)
5. Tsintsadze N, Margvelashvili-Malament M, Natto ZS, Ferrari M. Comparing survival rates of endodontically treated teeth restored either with glass-fiber-reinforced or metal posts: A systematic review and meta-analyses. *Journal of Prosthetic Dentistry*. 2024;131(4):567-578. [doi:10.1016/j.prosdent.2022.01.003](https://doi.org/10.1016/j.prosdent.2022.01.003) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/35430048/)
6. Figueiredo FE, Martins-Filho PR, Faria-E-Silva AL. Do metal post-retained restorations result in more root fractures than fiber post-retained restorations? A systematic review and meta-analysis. *Journal of Endodontics*. 2015;41(3):309-316. [doi:10.1016/j.joen.2014.10.006](https://doi.org/10.1016/j.joen.2014.10.006) · [PubMed](https://pubmed.ncbi.nlm.nih.gov/25459568/)

*This article summarises published research for clinicians. It is not a substitute for your own reading of the studies or your clinical judgment.*

If you're weighing this alongside the crown itself, [when a tooth actually needs a crown instead of a filling](/crown-vs-filling-clinical-guidelines) covers the same case, and [how long crowns really last](/the-research-says-how-long-do-crowns-really-last) puts the numbers above in context against a crown's own failure curve. And if fiber post kits and remakes are padding your supply line, run the numbers in the [overhead diagnostic](/tools/overhead-diagnostic).