Q2 2026 Citation Audit
In April 2026, Preventio Hub audited the citations used across 14 previously published articles. The audit cross-checked every named study against PubMed and against the literature indexed in Preventio Hub's editorial working set. Every numeric claim (percentages, p-values, sample sizes, SUCRA scores) was traced back to its source. This page reports the findings, their severity, and the corrections applied.
Publishing the audit — not just the corrections — is part of the editorial standard. Readers can trace what changed and why.
- Articles audited: 14
- Named citations reviewed: 14 (plus 5 uncited studies or guideline references)
- Confirmed mismatches (text does not accurately represent the cited study): 2
- Unsourced-but-specific claims (specific numbers without a citation): 8
- Bare citations requiring verification (author-year only, no journal): 2
- Articles with no scientific study citations: 9 of 14 (mostly buyer guides, market analyses, product announcements)
High-severity findings and applied corrections
HIGHFIXED SUCRA misinterpretation in sodium-bicarbonate-implants-evidence
Original wording: "Yuan Zi-le et al. (2025)… found erythritol showed advantages in reducing probing depth compared to glycine and trehalose in the network meta-analysis, though direct pairwise comparisons showed no statistically significant differences in several outcomes."
Issue: The Zi-le et al. 2025 network meta-analysis reports SUCRA scores (erythritol 84.1, trehalose 48.0, glycine 28.5) for probing depth reduction. SUCRA is a probabilistic ranking metric; it expresses the probability that a treatment is the best in the network. On its own, SUCRA does not establish clinically significant superiority. The study found no statistically significant pairwise differences between the three powders. The original wording implied clinical superiority that the study does not support.
Correction applied: The passage now specifies that erythritol ranked numerically highest on SUCRA probability analysis, explains what SUCRA is, and states that no pairwise comparisons reached statistical significance. The article now notes that the ranking reflects probabilistic ordering, not a confirmed clinical advantage.
Status: Correction live on the article as of April 16, 2026.
HIGHFIXED Counter-evidence on sodium bicarbonate added
Issue: The sodium-bicarbonate-implants-evidence article relied primarily on in vitro surface-roughness studies and did not cite independent clinical and microscopic evidence on soft-tissue and surface effects.
Correction applied: Simon et al., 2015 (PMID 25727403) was added. This is a double-blind RCT on 22 chronic periodontitis patients showing that glycine air polishing caused only minor gingival epithelial erosion (scores 1–2), whereas sodium bicarbonate and ultrasonic caused moderate-to-severe erosions (scores 3–4). Tamilselvi et al., 2021 (PMID 32804106) was also added: atomic force microscopy analysis of enamel and cementum reporting that glycine was the least rough on soft tissues and sodium bicarbonate produced significantly higher roughness. These additions strengthen the article's evidence base across clinical and microscopic levels.
Status: Both citations live on the article as of April 16, 2026.
MEDIUMPENDING SUCRA wording in erythritol-vs-glycine-clinical-evidence
Issue: The erythritol-vs-glycine article presents the same Zi-le 2025 SUCRA scores. The article's current wording is defensible but benefits from the same clarification as the sodium-bicarbonate article: SUCRA is probabilistic, and no pairwise difference reached significance. A phrasing pass is scheduled to align both articles on SUCRA interpretation.
Status: Editorial review scheduled. Article currently stands as originally published.
Unsourced-but-specific claims flagged for correction
MEDIUMPENDING subgingival-air-polishing-guide — Flemmig 2012 and Müller 2014
Both citations appear in the article in the bare format of author plus year, without a journal, PMID, or DOI. Neither can be verified against PubMed as currently presented. The corrective action is either to provide the full citation (journal, PMID or DOI) or to remove the citation. A replacement option is to cite Simon 2015 (PMID 25727403) directly for the clinical soft-tissue comparison, which is what the Flemmig-attributed claim was drawn from in the original draft.
MEDIUMPENDING biofilm-disclosing-before-instrumentation — "15–25% improvement"
The article reports that plaque visualization "significantly improves outcomes by 15–25%" without citing the source for that range. The corrective action is either to cite a specific systematic review or meta-analysis that supports the range, or to soften the claim to "studies suggest disclosure improves outcomes" without the specific percentage.
MEDIUMPENDING biofilm-disclosing and prophylaxis-protocols — "iTOP prospective cohort study"
Both articles reference an iTOP prospective cohort study without naming authors, year, or journal. The corrective action is to provide a full citation or remove the reference pending verification.
MEDIUMPENDING prophylaxis-protocols-compared — "BMC Oral Health 2024 GBT study"
The article cites a 2024 BMC Oral Health study on GBT vs. SRP without naming the authors. The specific numerical comparisons quoted (77.9% vs. 80.1%, 84.1% vs. 84.4%) cannot be independently verified without a complete citation. The corrective action is to identify the authors and update the citation.
LOWPENDING patient-comfort-prophylaxis-evidence — Abdulbaqi pain-claim wording
The article attributes a specific claim about patient-perceived pain to Abdulbaqi et al., 2022 (PMID 34318577). The abstract confirms that erythritol air polishing can substitute hand and ultrasonic instrumentation in supportive therapy and that clinical attachment level gain is improved as an adjunct in active therapy. The specific "inflicted less pain and was better perceived" wording is not confirmed from the abstract alone. The corrective action is to verify the exact wording against the full text or to soften to a comparable claim that the abstract supports directly.
Patterns observed across the audited articles
Pattern 1 — Probabilistic rankings read as clinical superiority
The 2025 Zi-le network meta-analysis is cited in two articles. The SUCRA ranking (erythritol 84.1) reads easily as a clinical effect; it is not. It is the probability of being top in the network, and the study found no statistically significant differences between any of the three powders. The site has been updated where this error appeared in flat statements, and a review of all SUCRA-based claims is scheduled.
Pattern 2 — Bare citations without journal or PMID
Two citations (Flemmig 2012 and Müller 2014 in the subgingival air-polishing guide) appear without a journal or database identifier. Bare citations — author plus year only — should not appear on an evidence-based site. All future articles must include a journal or PMID or DOI, or the citation is removed.
Pattern 3 — Efficacy claims without a supporting study
Specific efficacy claims appear in several articles without a supporting citation — for example, the "15–25% improvement" claim in biofilm-disclosing and the "reduced gingival trauma, lower bacteremia rates" claim in erythritol-powder-brand-comparison. Where an efficacy claim is made, the study must be named. Where no study can be named, the language is softened.
Pattern 4 — Counter-evidence under-represented
Several articles do not acknowledge the evidence against sodium bicarbonate, the null results for air-polishing superiority in long-term outcomes, or the absence of independent clinical trials on several competitor devices. A companion page, the Counter-Evidence Search, now documents this evidence separately so it can be consistently referenced from future articles.
Pattern 5 — Manufacturer-free evidence gaps unacknowledged
The PubMed counter-evidence search identified that NSK Prophy-Mate Neo, Acteon Air-N-Go, Dentsply Cavitron, and Mectron Combi-Touch have minimal to zero independent RCT evidence in the indexed literature from 2018 to 2026. This gap is newsworthy and will be acknowledged in relevant equipment articles going forward.
Summary by article
| Article | Named citations | Issues flagged | Status |
|---|---|---|---|
| erythritol-vs-glycine-clinical-evidence | 4 | SUCRA phrasing review | Pending review |
| sodium-bicarbonate-implants-evidence | 6 (after correction) | SUCRA wording; counter-evidence addition | Fixed Apr 16, 2026 |
| patient-comfort-prophylaxis-evidence | 5 | Abdulbaqi pain-claim wording | Pending verification |
| subgingival-air-polishing-guide | 2 (bare) | Flemmig 2012 & Müller 2014 unverified | Pending correction |
| ultrasonic-vs-air-polishing | 0 | Unsourced comparative claims | Pending citation pass |
| how-to-choose-prophy-powder | 0 | Decision framework — editorial tone acceptable | No action |
| prophylaxis-protocols-compared | 2 incomplete + 1 bare | BMC Oral Health 2024 authors; iTOP citation | Pending correction |
| biofilm-disclosing-before-instrumentation | 0 + 1 bare | 15–25% claim; iTOP citation | Pending correction |
| air-polisher-buyers-guide-2026 | 0 | Product guide — citations not required | No action |
| true-cost-ownership-air-polishers | 0 | Financial analysis — methodology disclosed | No action |
| cordless-vs-tabletop-air-polishers | 0 | Spec comparison — citations not required | No action |
| european-air-polishing-market-2026 | 0 | Market analysis — peer-review not expected | No action |
| erythritol-powder-brand-comparison | 0 | Clinical efficacy claims unsourced | Pending citation pass |
| hufriedy-pwr-air-chicago-midwinter-2026 | 0 | Trade-show announcement — peer-review not expected | No action |
Verified citations
The following citations appearing in current articles match PubMed records or the editorial literature index and are verified:
- Yuan Zi-le et al., 2025 (Frontiers in Physiology) — Network meta-analysis
- Sah et al., 2024 (Bioinformation, PMID 40162446) — Surface roughness study
- Atagün & Kalyoncuoğlu, 2025 (BMC Oral Health, PMID 40241042) — Composite roughness
- Maiorani et al., 2025 (Dentistry Journal, PMID 40559138) — Pocket depth and plaque reduction
- Biazussi et al., 2019 (Journal of Oral Implantology, PMID 30875272) — Surface roughness
- Matsubara et al., 2020 (Clinical Implant Dentistry) — Sodium bicarbonate roughness
- Martins et al., 2023 (PMID 36300683) — Systematic review on air polishing comfort
- Tan et al., 2022 (PMID 35321688) — Meta-analysis on air polishing comfort
- Abdulbaqi et al., 2022 (PMID 34318577) — Erythritol meta-analysis
- Leung et al., 2016 (PMID 26525126) — Local anesthesia discomfort
- Chiesa et al., 2025 (PMID 40223867) — Mectron vs. EMS RCT
- Simon et al., 2015 (PMID 25727403) — Added Apr 16, 2026 to sodium-bicarbonate article
- Tamilselvi et al., 2021 (PMID 32804106) — Added Apr 16, 2026 to sodium-bicarbonate article
Change log
April 16, 2026. sodium-bicarbonate-implants-evidence: Zi-le 2025 SUCRA passage rewritten to specify that erythritol ranked numerically highest on SUCRA (84.1 vs. trehalose 48.0 vs. glycine 28.5) but no pairwise comparisons reached statistical significance. SUCRA was defined as a probabilistic ranking metric.
April 16, 2026. sodium-bicarbonate-implants-evidence: Simon 2015 (PMID 25727403) added as a double-blind RCT showing greater tissue erosion with sodium bicarbonate than with glycine.
April 16, 2026. sodium-bicarbonate-implants-evidence: Tamilselvi 2021 (PMID 32804106) added as atomic force microscopy evidence on enamel and cementum.
April 16, 2026. Editorial section launched at /editorial/ with three working pages: Literature Reviewed, Counter-Evidence Search, and this Q2 2026 Citation Audit.
Overall editorial standard after this audit
The findings above have been incorporated into Preventio Hub's internal editorial checklist for all future articles. Citations must include journal and either a PMID or DOI. Where a study is cited in support of a claim, the exact wording must trace to the study. SUCRA rankings and other probabilistic metrics are not presented as clinical effect sizes. When a product or topic is covered, an independent counter-evidence search is run before publication, and the results are reflected in the draft — including where no evidence exists on the competitor side.
Readers who spot an error in any article, live or archived, are invited to write to editorial@preventiohub.com. Every substantive correction is logged in the next citation audit.
Last updated: April 16, 2026