Background:The overarching goal of the Norwegian and Swedish Health Authorities is to provide good health and care on equal terms for the entire population. The National Health authorities spend a lot of resources to improve health and social care through evidence-based recommendations in national guidelines. It is of particular interest to explore strategies to lower costs by increasing collaboration across nations in national guideline development.

Objectives: Collaboration across the National Authorities in Norway and Sweden on the development of three national guidelines was explored as a strategy to improve efficiency and lower costs in guideline development.

Methods: The National Board of Health and Welfare in Sweden and the Norwegian Directorate of Health in Norway cooperated on the development of two national guidelines. The chosen topics were on dementia, drug abuse and stroke. The collaboration was limited to sharing of specified methodological tasks such as evidence generation and evaluation. PICOs were decided and prioritised separately by each country. Thereafter, relevant PICOs for collaboration were selected and split between the two countries. Results of the collaboration projects were evaluated in terms of usefulness of the work that was generated by the other country.

Results: The Norwegian and Swedish National guidelines on drug abuse, dementia and stroke were finalised for broad national hearings in 2014, 2016 and 2017, respectively. Unexpected differences between countries in specific methodological aspects, made the generated evidence less useful for the other collaborating country when addressing the shared PICOs.

Conclusions: The total costs to develop the national guidelines on dementia, drug abuse and stroke were somewhat different, but approximately similar per recommendation in all guidelines, in both countries. More resources may be saved if both organisations used the same tools and methods and worked even closer, including more frequent physical meetings especially during the early developmental stages. This would allow for earlier recognition and targeting of unexpected methodological differences that may occur.