Design: phase 3, multisite, randomised, blind, placebo-controlled trial, 2 applications (baseline and 6 months), intention-to-treat
Population: 830 generally healthy US children aged 12-71 months with severe early childhood caries and active cavitated lesions; mean dmft 11.36
Primary outcome: proportion of arrested lesions at 6 months after one application
Effect: 54.0% vs 22.5% at 6 months, difference 31.5% (99.9% CI 21.5-41.6); 50.2% vs 17.4% at 8 months; no difference in pain

A phase 3, multisite, blind, placebo-controlled trial enrolled 830 children aged 12 to 71 months with severe early childhood caries and active cavitated dentin lesions, recruited from medical and dental clinics and preschool programmes. Mean decayed, missing and filled teeth at entry was 11.36. Children received 38% silver diamine fluoride or placebo at baseline and again at six months.

In intention-to-treat analysis, lesion arrest was 57.5% at three months, 54.0% at six and 50.2% at eight with silver diamine fluoride, against 18.8%, 22.5% and 17.4% with placebo. The differences, reported with 99.9% confidence intervals, were 38.7% (28.7-48.6), 31.5% (21.5-41.6) and 32.8% (22.3-43.2). Pain did not differ. Adverse events were similar between arms, and 7.5% discontinued for adverse events in each.

The reason this matters to a paediatrician rather than a dentist is access. Severe early childhood caries in a two-year-old conventionally means restorative treatment under general anaesthesia — a waiting list, a theatre slot, an anaesthetic risk, and a cost most families in Indian practice cannot meet. A liquid painted on the tooth in a clinic room, arresting half the lesions, changes who can be treated at all. It stains the arrested lesion black permanently, and parents must be told that before it is applied, not after.

Tell the parent the treated lesion turns black permanently — this is the commonest reason for later complaint
Arrest is not repair: the cavity remains and the tooth still needs dental follow-up
Thirty per cent of this trial was lost to follow-up, which weakens the eight-month estimates most
Consider it where general anaesthesia is unavailable, unaffordable or unsafe — that is the population it serves
It does not remove the need to address the cause: bottle feeding at night, sugary drinks and fluoride exposure

Why it matters

It offers a treatment for severe early childhood caries that does not require a theatre, an anaesthetist or a waiting list.

The statistics, in plain English

The trial reports 99.9% rather than 95% confidence intervals, a conservative choice that makes the estimates harder rather than easier to call significant — and the lower bound at six months is still 21.5 percentage points. The 30% loss to follow-up is the real limitation: if children who dropped out did worse, intention-to-treat analysis carries that forward only partly.