Air Polishing for Peri-Implantitis in Severe Periodontitis: What the 2026 BMC RCT Found

A new 2026 randomized controlled trial in BMC Oral Health found that submarginal erythritol air-polishing produced clinical outcomes broadly comparable to ultrasonic-plus-manual debridement for non-surgical peri-implantitis treatment in patients with severe periodontitis — but with a significantly larger reduction in soft-tissue bleeding and complete resolution of suppuration at 3 months. The trial also flagged a non-treatment finding that probably matters more in daily practice: keratinized mucosa width and concurrent periodontitis control were the dominant predictors of treatment success.

The study is Dai, Feng, Gao and colleagues, BMC Oral Health 26:192, published online 19 January 2026 (DOI 10.1186/s12903-026-07721-z; PMC12853578). It is the latest addition to a small but growing set of randomized data on air-polishing in peri-implantitis, and it lands while a Preventio Hub pre-registered systematic-review protocol on this exact question is in progress.

What the new 2026 trial tested

The Dai trial was conducted at the Department of Periodontology, Peking University School of Stomatology, between 2022 and 2024 and reported under CONSORT 2010 (ethics approval PKUSSIRB-202056099; Chinese Clinical Trial Registry ChiCTR2100048446). It enrolled 40 patients (38 completed) diagnosed under the 2017 World Workshop classification with stage III/IV periodontitis and peri-implantitis on at least one implant with at least one site of probing pocket depth ≥ 6 mm. Each patient contributed one implant — the deepest pocket where multiple were eligible. Smokers, patients on bisphosphonates or anticoagulants, those with diabetes or osteoporosis, and patients who had received antibiotics within 3 months were excluded.

Both arms first received a week of supramarginal cleaning to standardise baseline. Then, randomly:

• Test arm (n=19 completers): submarginal air-polishing with EMS PLUS powder — a 14 µm erythritol formulation containing 0.3% chlorhexidine — delivered with the nozzle parallel to the implant surface, 5 seconds per site at six sites per implant.
• Control arm (n=19 completers): submarginal piezoelectric ultrasonic scaling with a PEEK-coated implant tip (EMS PI instrument) plus manual debridement with LM titanium curettes (Finland) for at least 30 seconds.

Both arms received post-treatment submarginal irrigation with 0.12% chlorhexidine acetate. No systemic antibiotics were used during the study. Clinical parameters were assessed at baseline, 6 weeks and 3 months; radiographs at baseline and 3 months. The primary outcome was mean peri-implant PPD at 3 months; secondary outcomes included deepest PPD, modified Sulcus Bleeding Index (mSBI), modified Plaque Index (mPlI), suppuration, mucosal recession, and marginal bone level (MBL).

Key clinical results

Outcome	Erythritol air-polishing	Ultrasonic + manual	Between-group
Mean PPD at 3 months	4.2 ± 1.6 mm	4.7 ± 1.5 mm	Not significant
PPD reduction from baseline	−1.8 ± 1.6 mm	−1.3 ± 1.6 mm	Not significant
mSBI at 3 months	1.2 ± 1.0	1.8 ± 1.1	p < 0.05 (favours test)
Suppuration at 3 months	0 / 19 (resolved)	1 / 19	Test arm complete resolution
Strict 2023 EFP success	1 / 19 (2.6%)	0 / 19	Rare in both arms
Looser success (PPD ≤ 5 mm + no suppuration)	9 / 19 (47.4%)	9 / 19 (47.4%)	Identical
Complications	None (no emphysema, no severe pain)	None	Both safe

Full-mouth periodontal status improved in both arms — the patients were under active periodontal care concurrent with implant treatment, which the authors call out as relevant. Numerical mPlI values and the precise MBL change figures appear in the published tables; both groups reported MBL gain that exceeded the 0.1 mm threshold the authors used for clinical relevance.

What changed, and what didn't — interpretation

Three observations frame what to take from this.

First, the primary outcome was a draw. If a clinic is choosing between submarginal erythritol air-polishing and a careful piezoelectric-plus-curette debridement on the same indication, this trial does not say the air-polishing route gives meaningfully deeper pockets at 3 months. The numerical advantage in PPD change (0.5 mm) is within the noise of the standard deviations.

Second, soft-tissue inflammation responded better to air-polishing. The mSBI difference was statistically significant and the suppuration finding is clinically meaningful. For an implant where soft-tissue inflammation is the primary clinical concern — bleeding, mild suppuration, refractory mucositis — the trial supports air-polishing as a reasonable choice. This aligns with separate evidence on air-polishing producing less mechanical trauma to peri-implant soft tissue than instrumentation. The companion sodium-bicarbonate-on-implants evidence review covers the soft-tissue and titanium-surface mechanics that probably underlie the difference.

Third, neither arm hit strict treatment success often. The 2023 European Federation of Periodontology peri-implantitis success criteria — PPD ≤ 5 mm, no suppuration, no BoP at more than one site — were met by 1 in 19 in the test arm and 0 in 19 in the control arm. That is a sobering reminder of how difficult non-surgical peri-implantitis treatment remains, regardless of the decontamination tool. Looser criteria (PPD ≤ 5 mm plus no suppuration) gave a 47.4% success rate in both arms — identical between groups.

The keratinized mucosa finding clinicians shouldn't miss

The trial ran a generalized estimating equation analysis to identify factors associated with PPD at 3 months. Two findings stood out far more strongly than the treatment-arm difference:

• Keratinized mucosa width (KMW) ≥ 2 mm was associated with an odds ratio of 9.778 (p < 0.001) for reaching PPD ≤ 5 mm at 3 months. In the authors' own phrasing, implants with adequate keratinized mucosa were about 9.8 times more likely to achieve clinical success.
• Probing depth reduction at adjacent natural teeth was also a significant predictor. This supports the authors' core message: peri-implantitis and periodontitis must be treated concurrently, not sequentially, in stage III/IV periodontitis patients.

Translated into clinic decisions: the choice of decontamination tool — air-polishing versus ultrasonic — appears smaller than two things that already require attention before that choice is made. Whether the implant has at least 2 mm of keratinized mucosa around it, and whether the patient's periodontitis is under control. The treatment-arm comparison is the headline; the keratinized-tissue and periodontitis-control findings are arguably more actionable.

How this fits with the 2023–2025 evidence

The Dai 2026 RCT sits alongside a small but consistent body of work. A 2023 systematic review by Bi and colleagues (PMID 38258587) of air-polishing in peri-implant disease management concluded that air polishing produced outcomes broadly comparable to ultrasonic scaling for BoP and PPD reduction in non-surgical cases, and comparable to mechanical cleaning as a surgical adjunct — no superiority, but no inferiority either. A 2025 systematic review by Delucchi and colleagues on erythritol in implant dentistry (PMID 38825804) flagged a favourable biofilm-reduction signal for erythritol specifically. Earlier RCTs on non-surgical peri-implantitis with and without erythritol air-polishing — including the 12-month trial in PMC10125257 — found comparable outcomes between arms.

Dai 2026 is consistent with that pattern: air-polishing is a credible option in this indication, with a possible edge on soft-tissue inflammation, and not a demonstrated superior outcome on pocket depth. Our broader review of whether air polishing can be used on implants at all covers the longer evidence arc and the comparative case for glycine and erythritol on titanium surfaces. The erythritol-versus-glycine evidence article compares the two powders head-to-head on the available trials.

Limitations to keep in mind

The trial's limitations matter for how clinicians should weigh it.

• Sample size and single-centre design. 40 randomized patients, 38 analysed, at one Beijing institution. Population characteristics, operator skill, and implant systems may not generalise across European practice patterns.
• 3-month follow-up. Peri-implantitis is a chronic disease with long natural history. 3 months captures early response, not durable resolution. The 12-month Hentenaar et al. 2023 RCT (PMC10125257) on a related question is the longer-horizon reference point.
• Both arms used EMS hardware. The test arm used EMS PLUS powder and (by implication) an EMS AIRFLOW delivery system; the control arm used the EMS PI piezoelectric tip. The control was therefore not a generic ultrasonic — it was a specific implant-safe PEEK-coated tip from the same manufacturer. This isolates the air-polishing variable cleanly but constrains generalisation to clinics using different ultrasonic equipment.
• EMS holds the EU patent on erythritol air-polishing powder. Clinics in the European Union running non-EMS air polishers — including Acteon, Mectron, NSK, Woodpecker and others — cannot reproduce the test-arm intervention exactly. They run glycine or sodium bicarbonate (Woodpecker's "Super Powder" sold in Europe is a glycine formulation, not erythritol). A glycine-powder substitution is plausible on separate evidence, but it is not the same intervention this RCT tested. The erythritol powder brand comparison and how to choose prophy powder articles cover the powder-availability picture in detail.
• Strict success was rare in both arms. A 2.6% versus 0% strict-success rate is a low absolute outcome, and 3 months may not be enough time for the criteria to be met. The looser endpoint (47.4% in both) is more interpretable but also less stringent than the EFP-defined target.
• Funding and competing-interests declaration. Not captured in our reference extraction. Clinicians should consult the published paper directly for the statement, particularly given that both arms used hardware from a single manufacturer.

Practical takeaways for European clinics

For day-to-day practice, the Dai 2026 trial nudges the clinical decision in a few small ways without overturning anything.

• Submarginal air-polishing is a reasonable non-surgical peri-implantitis option, particularly when soft-tissue inflammation is the primary concern. It is not demonstrated to outperform careful piezoelectric scaling with an implant-safe tip on pocket depth, but it does not appear to underperform either.
• Keratinized mucosa first. If a peri-implantitis-affected implant has less than 2 mm of keratinized mucosa, the trial suggests the choice of decontamination tool is secondary. Surgical augmentation of keratinized tissue, where indicated, may matter more than the next-tool decision.
• Treat the periodontitis concurrently. The trial's GEE analysis identified probing depth reduction in adjacent teeth as a significant predictor of peri-implant response. Implant-only protocols in a periodontitis patient are likely to underperform.
• For non-EMS European clinics: the trial's exact intervention is not reproducible. Glycine-powder air-polishing with a submarginal nozzle is the closest substitute and is supported by separate evidence on glycine safety and efficacy around implants. Specific sub-gingival technique points are covered in our subgingival air polishing guide.
• For EMS-equipped clinics: the protocol is directly reproducible. PLUS powder (14 µm erythritol with 0.3% chlorhexidine) and a PEEK-coated implant tip are both standard configurations. The 5-seconds-per-site, 6-sites-per-implant cadence is straightforward to standardise.
• For protocol comparison. The Dai trial sits within a wider set of comparison data — see prophylaxis protocols compared for how the air-polishing-led families relate to conventional decontamination workflows.

Conclusion

The Dai 2026 RCT in BMC Oral Health adds a useful data point: submarginal erythritol air-polishing matches a well-executed piezoelectric-plus-manual debridement on pocket depth in non-surgical peri-implantitis among stage III/IV periodontitis patients, with a measurable edge on soft-tissue inflammation and suppuration resolution. With 40 patients and a 3-month horizon, it is not a definitive trial. The more durable finding is the dominance of keratinized mucosa width and periodontitis control as predictors of treatment success — a reminder that the decontamination tool is one variable inside a larger clinical system. For European clinicians outside the EMS ecosystem, glycine remains the working powder; the underlying rationale for an air-polishing-led approach travels across, even where the precise intervention does not. The pre-registered head-to-head protocol on Preventio Hub will continue to track this evidence as it accumulates.