Thirty-one RCTs (n=4,807) were included in the review.
One study reported a sample size calculation and method of allocation generation. Allocation concealment was adequately described in 4 studies, not mentioned in 3 studies, and indicated in the remaining studies. Blinding of the outcome assessment was adequately described in 4 studies, not mentioned in 2 studies, and indicated in the remaining studies. Blinding of personnel was adequately described in 8 studies and indicated in the remainder.
Studies lasting 3 months (43 comparisons in 27 studies).
Active agents significantly reduced calculus at 3 months compared with the control (ES -0.6, 95% confidence interval, CI: -0.7, -0.4). Statistically significant heterogeneity was found (p<0.0001 ), but this was no longer significant when the studies were grouped by the active agent (pyrophosphates 5.0%, pyrophosphates 3.3% plus copolymer 1.0%, and zinc citrate 0.5% and 2%). Significant reductions in calculus were seen for most agents except polyphosphates (1 trial), pyrophosphates 1.3% (4 trials) and zinc citrate 0.75% (2 trials). The largest reduction in calculus was obtained through the use of pyrophosphates 1.3% with copolymer 1.5% (5 trials) with an ES of -1.1 (95% CI: -1.7, -0.6).
Studies lasting 6 months (20 comparisons in 14 studies). Active agents significantly reduced calculus at 6 months compared with the control (ES -1.1, 95% CI: -1.5, -0.8). Statistically significant heterogeneity was found (p<0.0001). Results presented according to type of agent all showed statistically significant effects in favour of the active agents.
Studies lasting 12 months (3 studies).
Pyrophosphates 3.3% were associated with significantly reduced calculus compared with the control (ES -11.7, 95% CI: -13.8, -9.7), based on one study. There was no significant difference between pyrophosphates 1% and the control (2 studies).
The funnel plot of ESs for all studies of 3 months' duration was asymmetrical, suggesting the presence of publication bias; this was confirmed by statistically significant Begg and Majumder, and Egger test results.
Twenty studies reported no adverse effects.