Background: Best-practice statements (BPS) may be provided in guidelines in lieu of evidence-based recommendations when there is a high level of certainty that the benefits of the recommended intervention outweigh the harms. However, BPS are not clearly defined and terminology is inconsistent, leading to both overuse (an evidence review and standard recommendation should have been developed) and underuse (performing unnecessary evidence reviews).

Objectives: To describe BPS in World Health Organization (WHO) guidelines and to propose a new definition and typology for BPS.

Methods: All guidelines approved by the WHO Guidelines Review Committee and published from 2012 to 2016 were reviewed. We included discrete statements issued by guideline expert panels that were not based on a review of evidence. We extracted the characteristics of these statements and synthesised data using descriptive statistics. An iterative, consensus-based process was used to formulate definitions and a typology for BPS based on this cohort of guidelines.

Results: Of 86 guidelines, 31 contained BPS. These statements were variably labelled and presented, and the process by which they were developed and their rationale were often poorly reported. Several discrete types of BPS emerged, including those based on: 1) human rights and ethics principles and conventions; 2) indirect evidence based on physics or other principles; 3) indirect evidence based on established clinical principles; 4) the need to collect information; and, 5) other reasons where the BPS does not reasonably require the systematic collection of evidence.

Conclusions: This work may help guideline developers consider different types of BPS and to avoid their inappropriate use. The typology needs to be tested in different guideline cohorts for reliability and utility, and as a tool in the development of future guidelines.