16 Manual or powered?
An electric toothbrush costs between twenty and three hundred pounds. A manual one costs two. So this is the recommendation with the largest price tag attached, and it is the one where the guideline does something I have not seen it do anywhere else: it declines to choose.
The advice
Toothbrush type: use a manual or powered toothbrush
Strength of recommendation: Strong
— Delivering Better Oral Health, chapter 2, table 2a (1)
A Strong recommendation to do either of two things. It is worth pausing on how unusual that is. Strong normally means the panel is confident nearly everyone should do a particular thing. Here it means the panel is confident that it does not matter much which you pick, which is a different and less common kind of confidence.
I think it is the right call, and this chapter is mostly about why the evidence supports being that relaxed.
Following the citation
DBOH cites two reviews here (2,3), and notably not the Cochrane review of powered versus manual toothbrushing, which is the largest piece of work on the question. I will come back to that.
Wang and colleagues, 2020 pooled 21 randomized studies. According to the paper, powered brushes outperformed manual on every index:
| Outcome | SMD | 95% CI | I² |
|---|---|---|---|
| Plaque index | 0.86 | 0.58 to 1.14 | 91.5% |
| Gingival index | 0.47 | 0.12 to 0.82 | 88.7% |
| Bleeding index | 0.92 | 0.43 to 1.40 | 91.8% |
| Oscillating-rotating brushes vs manual, gingivitis | 0.07 | −0.20 to 0.33 | 57.2% |
Look at that last row, because it is the one nobody quotes. When the analysis is restricted to oscillating-rotating brushes, the design most commonly sold and most commonly recommended, the gingivitis difference is no longer statistically significant: SMD 0.07, 95% CI −0.20 to 0.33.
I wrote “the difference disappears” in an earlier draft and have changed it, because that interval does nothing of the sort. It runs from a small advantage for manual brushing to a moderate advantage for powered, and it is compatible with both. What has disappeared is the evidence of a difference, not the difference, and keeping those two apart is the whole subject of Chapter 17.
And look at the I² column. Ninety-one per cent. Heterogeneity that high means the included trials are not estimating the same thing, and the paper says as much, attributing it partly to differing assessment standards between studies. A pooled SMD of 0.86 across trials that disagree that violently is a number I would not put much weight on.
Elkerbout and colleagues, 2020 is narrower and cleaner. It examined plaque removal after a single brushing exercise and found more plaque removed by powered brushes on both of the indices it pooled: a difference of means of −0.14 (95% CI −0.19 to −0.09) on the Quigley and Hein index, from 15 comparisons, and −0.10 (−0.14 to −0.06) on the Rustogi modified Navy index, from 8.
Those are the overall figures. An earlier draft of this chapter gave two different numbers, both −0.15, which are in the paper but are the brand subanalyses: −0.15 (−0.22 to −0.08) is Procter and Gamble oscillating-rotating brushes on the first index, and −0.15 (−0.18 to −0.12) is Colgate side-to-side brushes on the second. Reporting a manufacturer’s own subgroup as the review’s headline result, in a chapter whose main complaint is industry funding, was an error worth recording rather than quietly fixing.
The authors are unusually candid about what that does and does not show:
There is a small difference in plaque removal in favour of the PTB. The single brushing design is rather direct as it does not reflect long-term use. As the risk of bias varied from “low to high” and many studies were industry-financed reporting bias cannot be ruled out.
— Elkerbout and colleagues, 2020 (3)
A single brushing exercise, with plaque allowed to accumulate for somewhere between 12 hours and 4 days beforehand, is about as far from “does this keep your gums healthy over decades” as a study can get while still being about toothbrushes. These were people brushing their own teeth, not a laboratory model, which I called it in an earlier draft; the distance from the question is about duration, not artificiality.
The outcome is plaque, not disease. Every estimate above is a plaque or gingival index. Nobody in these reviews measured tooth loss, periodontitis progression, or caries. Plaque scores are a surrogate, and a fairly distant one.
Blinding is impossible. You know which sort of toothbrush you are holding. Every trial in this literature has that problem, as in Chapter 17 and Chapter 11.
Industry funding is pervasive. Elkerbout says so directly. Toothbrush trials are largely funded by the companies that sell toothbrushes, and the empirical finding that industry-sponsored research more often favors the sponsor (4) applies with unusual force to a market this size. This does not make the results wrong. It is a reason for the certainty rating to be lower than the confidence intervals alone suggest.
On the Cochrane review. The chapter 13 row for this recommendation cites Wang 2020 and Elkerbout 2020 rather than the Cochrane review of powered versus manual brushing, and an earlier draft of this chapter said the guideline had left Cochrane out. That overstated it. Chapter 8 of the same guideline does use it: “There is moderate-certainty evidence to suggest that powered toothbrushes reduce plaque and gingivitis more than manual toothbrushing in the short and long term, and supporting their safety,” footnoted to Yaacob and colleagues, 2014. So the guideline as a whole cites the Cochrane review; the evidence table for this particular row does not. That is a smaller point than I first made, and the smaller point is the true one.
PubMed, 21 August 2026. I have full text and supplementary material for both Wang 2020 and Elkerbout 2020, supplied during writing. I did not run an independent search of the primary toothbrush literature, which is large and industry-dominated, and I would not trust a single-screener search of it.
SIGN 138 records the relevant null for caries specifically: “No evidence was identified to show that powered toothbrushes are more effective than manual toothbrushes in preventing dental caries in children” (5). Note the population in that sentence. It is children, and it is SIGN’s search rather than mine; I did not search the adult primary literature for a caries outcome, so nothing in this chapter should be read as a claim about adults.
Full record in appraisals/searches/ch16-powered.md.
Verdict
- Certainty of evidence
- Low that powered brushes reduce plaque and gingivitis more than manual ones in ordinary use. Downgraded from the reviews’ own ratings for extreme heterogeneity (I² above 88% on every pooled outcome in Wang), surrogate outcomes, impossible blinding and pervasive industry funding. Moderate that powered brushes remove more plaque in a single supervised brushing, which is a much smaller claim. No caries evidence either way. SIGN 138 identified no trials showing powered brushes better than manual for preventing caries in children, and I did not search the adult literature for that outcome.
- Directness to the advice as worded
- Adequate. The trials compare the two things the advice offers, in the right people. The weakness is outcome: plaque and gingival indices over weeks, for a choice people make for decades.
- Is the strength label defensible?
-
Yes, and it is the right shape. A Strong recommendation that either option is acceptable is well supported by evidence showing small, heterogeneous, surrogate-outcome differences. If the evidence were stronger, “use a manual or powered toothbrush” would be the wrong recommendation. Because it is weak, it is the right one.
I would go further than the guideline in one respect: it is worth saying out loud that for the oscillating-rotating subgroup, the brushes most people actually buy, the gingivitis estimate is SMD 0.07 (95% CI −0.20 to 0.33). That is compatible with a small benefit, with nothing, and with a small harm. It does not show the brushes fail; it shows this evidence cannot tell you either way.
- What would change my mind
- A trial measuring periodontitis progression or tooth loss over five years or more, independently funded. Failing that, an analysis explaining the heterogeneity: with I² above 90%, the interesting question is not the average effect but why the trials disagree so much.
What this means for you
Use whichever you will actually use, twice a day, for two minutes, with fluoride toothpaste. That last clause is doing all the work, and the guideline agrees.
If you already own an electric toothbrush, keep using it: it probably removes slightly more plaque, and nothing suggests it does harm. If you do not, the evidence does not justify the purchase on health grounds alone. The pooled differences come from trials that disagree with each other enormously, measured on plaque scores rather than disease, and mostly funded by manufacturers. For the commonest type of powered brush the gingivitis estimate is too imprecise to separate a small benefit from a small harm.
The honest summary is the guideline’s own: either is fine.