Design: Three-arm effectiveness-implementation randomised trial
Population: 630 Latino caregiver-infant dyads (infants aged 6 to 12 months)
Primary outcome: Child language and social-emotional development, and parenting behaviours
Effect: Texts improved language (effect size 0.39) at 18 months; added referral improved social-emotional development

Reading promotion in primary care, such as Reach Out and Read, supports early development, but how to amplify it is less clear. This three-arm randomised trial enrolled 630 Latino caregiver-infant dyads, most with limited formal education and limited English, and tested adding text-message reminders, with or without referral to a community-resource navigator.

Both enhancements helped, in different domains. Adding texts improved child language scores at 18 months and reading behaviours at 9 months versus literacy promotion alone. Adding texts plus community referral improved social-emotional development at 9 and 18 months and verbal responsiveness at 9 months. Effect sizes were in the small-to-moderate range and survived adjustment for multiple comparisons.

The practical point is that low-cost, scalable additions, a text reminder and a link to community resources, measurably strengthen a familiar clinic intervention, and are worth building into primary-care reading promotion for families who face the steepest barriers.

Three-arm randomised trial of 630 Latino caregiver-infant dyads in primary care.
Text reminders improved child language at 18 months and reading behaviours at 9 months.
Texts plus community referral improved social-emotional development and verbal responsiveness.
Effect sizes were small-to-moderate and adjusted for multiple comparisons.
Both additions are low-cost and scalable within existing reading-promotion programmes.

Why it matters

It shows a familiar reading-promotion programme can be made measurably more effective with cheap, scalable additions.

The statistics, in plain English

Effect sizes around 0.3 to 0.4 are small-to-moderate differences in development scores, meaningful at a population level even if modest for one child. Adjusting for multiple comparisons makes the significant findings more credible than raw p-values would.