| Randomized village clusters (7 intervention arms and 7 control arms) | Substantial listenership (as reported) | Substantial improvements in behavior seen in control areas |
| Areas believed to be isolated from other mass media messaging (outside the “electric grid”) | Improvement in some care seeking behaviors, but mostly no difference compared with control areas | Large differences between intervention and control areas at baseline: intervention areas poorer, more likely Muslim, and farther from health facilities |
| Mass media only (radio) | Some dose effect seen based on "spot" messages but not dramas | Seasonal difference between timing of baseline and midline surveys |
| Addressed multiple behaviors | Increase in service site utilization corroborates the reported increases in care seeking | Contamination of exposure to radio messages in one control area |
| Prioritized most important behaviors based on lives saved modeling | | Various other health promoting activities occurred in both intervention and control areas |
| Extensive formative research | | No qualitative data provided as yet to give further insights |
| Short spots and interactive dramas (story-based) | | Mass media only |
| Aired multiple times per day (spots) or per week (dramas) | | Many behaviors addressed; certain topics received more emphasis |
| Potential to reach others besides primary caregivers | | No supply-side change limits service-related behavior |
| Local, rapidly developed, innovative content with quality control | | More effect may take more time |
| Partnerships with local stations resulting in cost-efficiencies | | |
| No efforts to increase or improve service delivery | | |