19 Scale, polish, and being told what to do
Every chapter so far has asked whether a recommendation is better supported than it looks. This one asks the reverse question. What happens when the evidence is excellent, unambiguous, and says the thing does not work?
The answer turns out to be interesting, because Delivering Better Oral Health gets this right, and a great many dental practices do not.
The advice, and what is missing from it
Here is the thing to notice first. Search the guideline for a recommendation to give patients a routine scale and polish, and you will not find one.
What table 2a contains is this:
Correct factors that impede effective plaque control including supra and subgingival calculus, open margins and restoration overhangs and contours, which prevent effective plaque removal
Strength of recommendation: Good practice
— Delivering Better Oral Health, chapter 2, table 2a (1)
That is a recommendation to remove calculus where it is impeding plaque control, labeled honestly as clinical consensus rather than research. It is not a recommendation that healthy adults should have their teeth scaled every six months.
The absence is the finding. If you have ever been booked in for a six-monthly “scale and polish” as a matter of course, you have received something the national guideline does not recommend, and the reason it does not recommend it is that the question has been asked properly and answered.
What the evidence says
The Cochrane review of routine scale and polish included two trials in 1,711 adults, both conducted in UK general dental practice, in adults without severe periodontitis who attended regularly (2). One followed participants for 24 months, the other for 36.
According to PubMed, comparing planned six- or twelve-monthly scale and polish against no scheduled treatment:
| Outcome | Finding | Certainty |
|---|---|---|
| Gingivitis | little or no difference; SMD −0.01 (95% CI −0.13 to 0.11) for six-monthly treatment, −0.04 (−0.16 to 0.08) for twelve-monthly | high |
| Probing depths | little or no difference | high |
| Oral health-related quality of life | little or no difference | high |
| Plaque | little or no difference | low |
| Calculus | small reduction; SMD −0.32 (95% CI −0.44 to −0.20) six-monthly, −0.19 (−0.31 to −0.07) twelve-monthly | high |
| Self-reported oral cleanliness | higher with treatment | low |
Those “six-monthly” and “twelve-monthly” labels are how often the treatment was given, not when the outcome was measured. Everything above was assessed at 24 to 36 months. I had that wrong in an earlier draft and it made the follow-up sound a great deal shorter than it was.
And comparing six-monthly against twelve-monthly: little or no difference in gingivitis (SMD 0.03, 95% CI −0.09 to 0.15, I²=0%), probing depths or quality of life, all at high certainty, with a small extra calculus reduction from the more frequent option.
Read the certainty column, because it is doing something unusual for this book. High certainty, from randomized trials, in the right population, in the right setting, over two to three years. These are not weak studies that failed to find an effect. This is a well-designed body of evidence that looked hard and found little or nothing.
One qualification on the outcomes, since this book is severe about surrogates elsewhere. Gingivitis indices and probing depths are measured by a clinician, not felt by a patient, and they are intermediate outcomes for the thing that actually matters. What the trials did not measure is attachment loss, tooth loss, or progression to periodontitis. Oral health-related quality of life, which they did measure, is the one genuinely patient-reported outcome in the set, and it also showed little or no difference.
The one thing routine scaling reliably does is remove calculus. Which is to say: it achieves its immediate technical objective, and that does not translate into less gum inflammation, shallower pockets, or a patient who feels better. The review’s own phrase for the calculus finding is that “the clinical importance of these small reductions is unclear.”
Chapter 4 warns against treating absence of evidence as evidence of absence. This chapter is the case where the distinction runs the other way, and it is worth being precise.
For gingivitis, the pooled estimate is SMD −0.01 with a confidence interval from −0.13 to 0.11. That interval is narrow and centered on zero. It excludes effects of half a standard deviation or more, which is the threshold the review prespecified as an appreciable difference in the absence of an agreed minimally important difference for these indices. So the precise claim is not “it excludes any effect a patient would notice,” which I wrote before and cannot support since no patient-noticeable threshold has been established for a gingival index. It is “we measured it precisely and it is very close to nothing, on the review’s own yardstick.” GRADE’s high-certainty rating reflects exactly that: randomized evidence, consistent across the two trials, precise enough to rule out an appreciable benefit.
Compare with Chapter 6, where no study of the question exists at all. Both chapters can be summarized as “no demonstrated benefit,” and they mean completely different things. One is an empty literature. This one is a full answer.
What the review could not assess is also worth recording: neither trial measured adverse effects, changes in attachment level, tooth loss, or halitosis. Scaling is invasive and has been associated with tooth sensitivity and damage to tooth surfaces. So the harm side of the ledger is not empty because it is zero; it is empty because nobody filled it in.
The other half of table 2a
Alongside the plaque-control row sits a set of recommendations about telling patients what to do, and these deserve their own look because they are graded Conditional:
Use behaviour change methods with oral hygiene instruction
Advise best methods of plaque removal to prevent gingivitis and achieve lowest risk of periodontitis and tooth loss
Strength of recommendation for both: Conditional
— Delivering Better Oral Health, chapter 2, table 2a (1)
These look like one thing and are two, resting on different evidence at different certainties. Chapter 13 keeps them apart even though the summary table does not:
| Recommendation | Evidence, in chapter 13’s words | Source |
|---|---|---|
| Advise best methods of plaque removal | “very low certainty evidence for one-to-one oral hygiene advice reducing gingivitis and plaque” | Soldani and colleagues (3) |
| Use behavior change methods | “low certainty evidence that goal setting, self-monitoring and planning improve oral hygiene-related behaviour” | Newton and Asimakopoulou |
An earlier draft of this chapter attributed both rows to the Cochrane review of one-to-one advice. Only the first is Soldani. The second rests on separate behavior-change work that I have not appraised, and I am not going to pronounce on a label whose evidence I have not read.
On the first row, the Conditional label is a fair reflection of very low certainty evidence. This is the guideline doing precisely what Chapter 2 asks: where the evidence is uncertain and reasonable people might weigh it differently, it says Conditional rather than Strong.
I note this without a complaint attached, which is the point of including it. Note also the outcome in the second row: “improve oral hygiene-related behaviour.” That is a behavioral outcome, not a disease outcome, and it is honestly labeled as one.
PubMed, 21 August 2026. The Cochrane review’s own search ran to 10 January 2018 across the Cochrane Oral Health Trials Register, CENTRAL, MEDLINE Ovid and Embase Ovid, with two trial registries for ongoing studies. I checked for a .pub6 update and found none; CD004625.pub5 remains current.
I have the full text of this review, with its characteristics of included studies and risk-of-bias tables, supplied during writing. The statements above about follow-up at 24 to 36 months, the review’s 0.5 SD threshold for an appreciable difference, the eligibility criteria of BPE 0 to 3, and the minimization on gingival bleeding are from the review body and its study characteristics, not from the abstract. An earlier draft carried a blanket “abstract only” disclosure inherited from another chapter, which was not true here.
The age gap matters here. Only two trials have ever addressed this question, both in the UK, and the evidence has not moved since 2018. High certainty from two trials is high certainty about two trials.
Full record in appraisals/searches/ch19-scale-and-polish.md.
Verdict
- Certainty of evidence
- High that routine scale and polish makes little or no difference to gingivitis, probing depths or quality of life in adults without severe periodontitis who attend regularly. High that it produces a small reduction in calculus of unclear clinical importance. Absent on adverse effects, tooth loss and attachment level, which no included trial measured.
- Directness to the advice as worded
- Excellent, and the advice is correspondingly narrow. DBOH does not recommend routine scaling, and the evidence is why. The recommendation it does make, to correct calculus that is impeding plaque control, is a different and narrower claim which these trials do not address, and which the guideline correctly labels Good practice.
- Is the strength label defensible?
-
Yes, and this is the guideline at its best. Faced with high-certainty evidence of no benefit, it declines to recommend the intervention. Faced with a narrower clinical judgment it cannot support with research, it says Good practice. Faced with uncertain evidence on oral hygiene instruction, it says Conditional. Three different situations, three different labels, each correct.
The gap in this chapter is not in the guideline. It is between the guideline and routine practice.
- What would change my mind
- A trial in adults with periodontitis, or in irregular attenders, either of whom might benefit where the studied population did not; the review’s population is explicitly healthy regular attenders. A trial measuring attachment level or tooth loss over longer than three years. Or evidence on the harm side, which is currently unmeasured rather than absent.
What this means for you
If you are a healthy adult who sees a dentist regularly and does not have severe gum disease, the best available evidence says a routine six-monthly scale and polish will not measurably improve your gums, your pocket depths, or how your mouth feels. It will remove calculus, and your teeth will feel cleaner afterwards, which the trials did measure and which is a real thing to value if you value it.
What you should not believe is that it is reliably reducing gum inflammation or pocket depths. Two randomized trials in exactly your situation looked for that over two to three years and found little or nothing, at high certainty. They did not measure whether it prevents teeth being lost decades later, and nobody has.
This is different advice from most of the chapters here, and worth being clear about. Elsewhere I say: keep doing it, the evidence is just weaker than the label suggests. Here I say: this is a treatment you may reasonably decline, and the national guideline does not ask you to have it.
One boundary, and it is narrower than I first drew it. An earlier draft said that none of this applied if your gums bleed. That was wrong: the larger trial enrolled adults with BPE scores 0 to 3, two thirds of whom had visible gingival inflammation, and it explicitly balanced the arms on whether or not gums bled on probing. Bleeding gums were not an exclusion; they were most of the sample.
What was excluded is severe periodontal disease, BPE 4. So if you have been diagnosed with periodontitis, or you do not attend regularly, this evidence is not about you, and Chapter 4 explains why that matters. If your gums simply bleed when you brush, it is about you, and that is the more useful and more uncomfortable finding.