Consideration of person-important outcomes by guideline recommendations for the use of tests and testing strategies: protocol for a meta-epidemiological study

Ideally, tests and testing strategies are only recommended for use if they improve person-important outcomes (PIOs). However, direct evidence on PIOs is rare, and evidence on test accuracy is extensively used for guideline development. When guideline developers fail to consider PIOs, they might recommend to use tests at risk of excessive testing; thus, we will conduct this study to examine the extent to which guideline recommendations about the use of tests and testing strategies consider PIOs, and to examine factors related to such considerations. We will conduct a systematic survey of clinical practice guidelines that provide recommendations for the use of specific tests or testing strategies, for the purposes of diagnosis or screening, published in English, and in recent 1 year. A team of ten reviewers will work in pairs, independently determine the eligibility of the documents and conduct the data extraction. We will randomly sample 100 guidelines (and up to three recommendations per guideline) for data extraction and analysis. We will classify the consideration of PIOs using a prespecified framework with living feedback-and-update during data extraction to ensure proper classification of all real-world cases. We will provide descriptive analyses of the characteristics of the sampled guidelines and recommendations, their consideration of PIOs, and the underlying evidence. We will examine the association between prespecified recommendation-level characteristics and the consideration of person-important benefits among recommendations in favor of the use of tests. This study provides the first empirical evidence of the consideration of PIOs in the current development of guideline recommendations for the use of specific tests and testing strategies, which is important for evidence-based practice and reduction of medical overuse. Findings of this study will be disseminated through peer-reviewed publications and academic conferences. Systematic review registration: https://doi.org/10.17605/OSF.IO/4RQ3F .