A new systematic review and meta-analysis in Diabetes Research and Clinical Practice reports that, across 15 studies, the pooled completion estimate for the reported follow-up milestone in dental-setting referral pathways was 56.0%. In seven studies, the pooled yield of confirmed prediabetes or diabetes among people who completed diagnostic assessment was 39.5%. The authors warn that the completion estimate is highly heterogeneous and that closed-loop referral still needs comparative evaluation.
Dental visits may reach adults whose abnormal blood glucose has not been recognized. That is the premise of a review published online on October 31, 2026, by Ruan and Wang, whose first author lists affiliations in Fujian and Shanghai, China. As of November 2026, the paper addresses the question that matters to a practice considering this work: what happens after a dental team flags someone?
The authors frame it plainly. Clinical value, they write, depends on progression from risk identification to medical diagnosis and care. In our reading, identifying risk is only the first step. This report is based on the paper's published abstract and citation record; EBIKO Dental did not have access to the full text.
What did the review examine?
The reviewers searched MEDLINE/PubMed, Embase and Web of Science Core Collection from inception through October 12, 2026, looking for dental-setting studies that reported screening, referral, confirmatory assessment or linked-care outcomes. Two reviewers independently screened 751 deduplicated records. Of 71 reports sought for retrieval, 64 were retrieved and assessed for eligibility.
Forty-eight unique eligible reports informed a qualitative evidence map. Only 15 studies provided referral denominators, and those 15 supplied the main pooled estimate. The abstract does not say why the other reports did not contribute, but the pooled follow-up figure rests on those 15 studies.
What did the meta-analysis find?
Across the 15 studies with referral denominators, 1,269 participants completed the reported follow-up milestone, against study-defined denominators totalling 2,727. The random-effects pooled completion was 56.0% (95% confidence interval 45.8% to 65.7%), with an I² of 95.1% and a prediction interval of 21.6% to 85.5%.
In seven studies, 217 of 715 diagnostic completers had confirmed prediabetes or diabetes. The pooled yield was 39.5% (95% confidence interval 26.2% to 54.5%; I² 89.1%).
Dividing the raw counts gives about 46.5% for completion and about 30.3% for yield. Those two simple ratios are EBIKO Dental's arithmetic, not figures reported in the abstract. They sit below the pooled estimates, which is one more reason to treat every number here as an indication rather than a rate your practice should expect.
Why should readers be careful with the 56% figure?
The authors say so directly: the highly heterogeneous completion estimate should not be treated as a universal programme effect. They also report that endpoint definitions and diagnostic criteria differed materially between studies. "Completed follow-up" did not mean one fixed thing.
The prediction interval carries the practical message. In general, a prediction interval estimates where the result of a new, similar setting might fall, and here it runs from 21.6% to 85.5%. A program could plausibly see roughly one in five people complete follow-up, or more than four in five.
The 39.5% yield also needs a careful reading. Its denominator is people who completed diagnostic assessment, not everyone screened in a dental chair. Those who complete testing may differ from those who do not, so the figure does not tell you how many of your screened patients have dysglycaemia.
Do integrated care pathways work better?
The evidence does not yet settle that. Integrated pathways were associated with high completion in some observational reports, but the authors describe comparative evidence as sparse and confounded. Their conclusion is measured: dental screening can identify dysglycaemia, while closed-loop referral and related implementation components remain proposed strategies requiring pragmatic comparative evaluation.
The abstract does not break results down by country, so it cannot tell a Canadian reader whether any Ontario or other Canadian practices were included, or how referral into a Canadian primary care system would perform.
How does this fit Canadian diabetes screening guidance?
For Canadian readers, the backdrop is this. The Public Health Agency of Canada's 2023 snapshot reports that, between 2007 and 2019, 1.9% of adults aged 20 to 79 had undiagnosed diabetes, and 22.5% of those who met the diagnostic criteria were unaware of their condition. A further 6.3% had prediabetes.
Diabetes Canada's clinical practice guideline recommends screening for type 2 diabetes with fasting plasma glucose and/or A1C every three years in people aged 40 or older, or in people at high risk on a risk calculator. The federal CANRISK questionnaire is intended for adults aged 40 to 74, and scores 33 and over as high risk, 21 to 32 as moderate risk, and lower than 21 as low risk. It is a risk questionnaire, not a diagnosis.
What should Ontario practices take from this?
For dental professionals in Toronto, Mississauga, Markham and across the GTA who already take detailed medical histories, the paper's useful lesson is about the handoff. If your practice wants to act on risk, the referral design deserves as much attention as the screening method.
- Define the endpoint first. Decide what counts as done: a referral letter sent, a physician visit booked, or a test result returned. These are our examples; the review's abstract reports only that endpoint definitions differed materially between studies.
- Name the route. Agree how a patient reaches a family physician or nurse practitioner, and what your practice does when a patient has neither.
- Close the loop. Assign someone to check whether follow-up happened, and record the answer in the chart with the patient's consent. The review treats closed-loop referral as a proposed strategy, so measure whether it helps in your own practice.
- Confirm scope before testing. If you are considering chairside glucose or A1C testing, ask the Royal College of Dental Surgeons of Ontario (RCDSO) and your professional liability program what applies before buying anything.
Pro Tip: Before launching any screening initiative, run a 30-day chart review of the medical referrals your practice already sends. If you cannot say how many reached a physician, fix that tracking first.
Pro Tip: If you use CANRISK, keep to its intended 40-to-74 age range and tell patients plainly that a score indicates risk only. A physician or nurse practitioner can arrange testing to confirm or rule out a diagnosis.
EBIKO Dental will continue monitoring research on oral-systemic care pathways and its relevance to Canadian practices. More practice and clinical updates are available at EBIKO Dental.
Sources
- Ruan C, Wang L. From dental screening to diabetes care: effectiveness of screening, referral, and integrated care pathways for dysglycaemia in dental settings: a systematic review and meta-analysis. Diabetes Research and Clinical Practice, vol. 242 (2026); PubMed ID 42966402
- Public Health Agency of Canada: Snapshot of Diabetes in Canada, 2023
- Diabetes Canada Clinical Practice Guidelines: Screening for Diabetes in Adults
- Canadian Diabetes Risk Questionnaire (CANRISK)
Frequently Asked Questions
Q: Can a dental visit detect diabetes?
A 2026 meta-analysis concluded that dental screening can identify dysglycaemia, with cases confirmed through diagnostic assessment. In seven studies, the pooled yield of confirmed prediabetes or diabetes among people who completed diagnostic assessment was 39.5%, with wide uncertainty.
Q: How often do people referred from dental settings reach follow-up?
The review's pooled estimate was 56.0% across 15 studies, but its prediction interval ran from 21.6% to 85.5%. The authors say the highly heterogeneous estimate should not be treated as a universal programme effect. They also report that endpoint definitions and diagnostic criteria differed materially.
Q: Who should be screened for type 2 diabetes in Canada?
Diabetes Canada's guideline recommends screening every three years in people aged 40 or older, or in people at high risk on a risk calculator. The CANRISK questionnaire is intended for adults aged 40 to 74.