Establishing the effectiveness of public health interventions using sequential meta-analysis: Case study using stair-promotion interventions

Adrian Bauman, Karen Milton

Research output: Contribution to journalArticlepeer-review

18 Downloads (Pure)


Prevention trials often generate significant health or risk factor improvements, but these improvements are limited to the selected samples of people enrolled in these studies. To make population health gains, we need to develop methods for translating evidence from these trials into practice at the broader community and population levels. One barrier to this process is researchers themselves, who conduct replications of small-scale trials, rather than conducting intervention research at scale. Our example of this is interventions to encourage short episodes of physical activity through signs promoting stair use. We pooled the evidence from these interventions from 1980 to 2014. We carried out a meta-analysis to estimate the proportion of people that changed from the elevator to the stairs following the introduction of signage. Our innovation was to use a sequential meta-analysis method, usually described in clinical settings and trials. We used this method to estimate when there was sufficient evidence of stair sign effectiveness for public health actions to be scaled up, and we found this was around 2006. Studies since then have not contributed new evidence to the field. Methods here enabled us to see when policy makers should have implemented this intervention to the community at large, and researchers then should have focused their investigations on identifying barriers and facilitators to their implementation and assessing intervention effects at scale.
Original languageEnglish
JournalSAGE Research Methods Cases
Publication statusPublished - 2020

Cite this