Article Text

Download PDFPDF
PP16 Can expert opinion rapidly provide useful quantitative data on policy effectiveness and inequalities? Pilot study
  1. DOS Gillespie1,
  2. K Allen2,
  3. M Guzman-Castillo1,
  4. P Bandosz1,
  5. P Moreira1,
  6. R McGill1,
  7. F Lloyd-Williams1,
  8. H Bromley1,
  9. P Diggle2,
  10. M O’Flaherty1,
  11. S Capewell1
  1. 1Department of Public Health and Policy, University of Liverpool, Liverpool, UK
  2. 2Lancaster Medical School, Lancaster University, Lancaster, UK


Background Public health policy development has many uncertainties, including a lack of data on the differential socio-economic impacts of interventions. Successfully capturing quantitative prior estimates of policy outcomes from topic experts could be valuable. We therefore aimed to develop a rapid methodology to elicit experts’ quantitative assessments of the effectiveness and equity of future policies, piloted initially on salt-reduction policies in England.

Methods We identified three possible future salt policy options: voluntary product reformulation, social marketing and nutrition labelling. We developed a conceptual framework that distinguished structural and agentic policies; this quantitatively described each policy by its efficacy, coverage, ultimate population effect and potential inequalities. We developed an elicitation process to obtain expert views on potential policy changes up to 2020. A draft questionnaire was piloted iteratively to refine content and language. The resulting seven item questionnaire elicited the estimated best, minimum and maximum effect sizes for each policy parameter, plus socio-economic differentials (across quintiles of the Index of Multiple Deprivation). We approached twenty scientists and policy experts in the non-governmental and academic sectors. After obtaining formal consent, each expert was given a brief introduction to the project context and invited to complete the questionnaire. We quantified uncertainty from 10,000 probabilistic samples of the experts’ estimates.

Results Fourteen of the twenty experts approached agreed to participate; eleven provided complete responses. The experts estimated that further voluntary reformulation could result in approximately 39% of products being reformulated by 2020 (95% CI: 13–76%). In these products, the experts expected a 22% (9–45%) salt-content reduction. Less progress was expected in the most deprived fifth, who the experts considered might experience approximately 82% (18–150%) of the reduction in the most affluent fifth. We asked the experts to assume a baseline of 10% of individuals currently exposed to social marketing and 10% exposed to nutrition labelling. They estimated that this exposure would change to approximately 24% (8–55%) with developments in social marketing and approximately 26% (9–50%) with developments in nutrition labelling. Experts considered that the most deprived fifth might experience an increase in exposure about 45% (15–89%) of that in the least deprived fifth for social marketing, and 49% (8–110%) for nutrition labelling.

Conclusion We describe the development of a user-friendly, transparent procedure to rapidly elicit experts’ quantitative estimates on population effects and inequalities. Such data could potentially inform debate or policy formation, populate models or guide further empirical studies.

  • expert elicitation
  • salt
  • public health

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.