Practical considerations for transitioning early childhood interventions to scale: lessons from the Saving Brains portfolio

Abstract

Small pilot studies of young children have frequently shown promise, but very few have been successfully scaled to the regional or national levels. How can we ensure that these promising approaches move from a suite of pilots to full-scale implementation that can deliver sustainable impact for hundreds of millions of children? To elucidate concrete lessons learned and suggestions on accelerating the transition to impact at scale, we reviewed the Saving Brains portfolio to better understand three points: (1) the extent to which useful signals of impact could be extracted from data at the seed phase, (2) the ways in which innovators (project leaders) were approaching human resource challenges critical for scaling, and (3) the multisector diversity of the portfolio and the way innovators entered partnerships. The findings suggest key considerations for transitioning early childhood development interventions to scale and sustainability: strong entrepreneurial leadership, rigorous measurement and active use of data in support of adaptive learning, and champions acting at subnational levels. Together, these can enable flexible, iterative learning that can make the scaling process an opportunity to increase the level of benefit each child receives from an intervention.

Keywords: early childhood development; implementation research; international development; program evaluation; scaling up social innovation.

Figure 1

Descriptive details and characteristics of Saving Brains interventions. (A) The proportion with a psychosocial component, n = 39. (B) The minimum and maximum age of eligible children, and the minimum and maximum durations of the intervention for psychosocial interventions in the portfolio, n = 34. Interventions that began during pregnancy were counted as a negative age in months. (C) The proportion of psychosocial interventions with a home visiting component, n = 34. “Home‐visiting component” means the intervention included visits for intervention programming (and not merely for evaluation). (D) The proportion of different kinds of service providers delivering the psychosocial interventions, n = 34. (E) The proportional distribution of locations where psychosocial interventions were delivered, n = 34. “Center” includes a diverse set of locations such as childcare centers, clinics, hospitals, community centers, and other gathering places. “Technology” indicates the use, for example, of text messaging or video.

Figure 2

The relationship between signal strength size scores for child development outcomes (horizontal axis) and home/parental outcomes (vertical axis) is shown. Larger bubbles mean more projects with those scores.

Figure 3

Relationships between human resources burdens in proof‐of‐concept psychosocial “seed” grantees and child outcome size (signal strength) scores, where data were available. Saving Brains projects with child development signal‐strength scores of 1 or 2 (SB Projects with Low Scores) were compared with those with scores of 3 or 4 (SB Projects with High Scores), as well as external, published studies we coded (External, Published Projects). (A) Hours of training for an individual staff member per family served are shown, n = 24. Training includes initial baseline training, refresher training, and supervisory time. There was an insufficient n for published literature (n = 3). (B) Hours spent by staff delivering interventions, n = 38. Two outliers were removed. (C) Hours spent by families receiving interventions, either in groups or individually, n = 38.