An Exploratory Descriptive Study Evaluating Divergence in the Judgment of Clinical and Public Health Recommendations for the Management of the Novel Coronavirus Disease (COVID-19)

Abstract

Background: The emergence of the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) pandemic has unravelled a global demand for rapid and reliable guidance at the clinical, systems, and policy levels. Therefore, to equip decision-makers with the appropriate knowledge and tools, organizations have published evidence-informed guidance for its prevention and treatment. Objectives: Various organizations may produce comparable but diverging recommendations for the same intervention or health scope. Diverging recommendations are those that contain varying judgements in their strength, direction, or subgroup consideration associated with the intervention. Nonetheless, the extent of divergence between COVID-19 recommendations remains unknown. Consequently, the primary objectives of this study are to 1) describe the frequency and types of divergence between COVID-19 recommendations for the same intervention and 2) investigate differences in the guideline development process for a selected sample of diverging recommendations. Methods: We screened guidelines for divergence using the digital COVID-19 Recommendations Catalogue (covid19.recmap.org). Diverging recommendations for the same intervention were grouped into clusters, and differences in their formal judgment of strength and direction were appraised. Additionally, we compared any differences between PICO criteria for comparable recommendations addressing the same health scope. Descriptive statistics were performed to assess the frequency and types of divergence. Finally, we applied deductive content analysis to evaluate differences in the methods for a sample of 12 recommendation clusters. Results: Two-hundred twenty-three diverging recommendations resulted in the categorization of 66 clusters. Twenty-nine clusters contained clinical also stated as therapeutic recommendations, and 37 clusters contained public health recommendations. Each cluster had a range of 2-8 individual recommendations in divergence with at least one recommendation for the same intervention. Clinical recommendations were more likely to diverge in formal judgment than public health recommendations (P < 0.001). We identified differences in the date of publication, the interpretation of evidence, and in the judgments of the Evidence-to-Decision framework between comparable recommendations. Consequently, results from our study may have important implications for comparing duplicate recommendations and for making clinical practice decisions. Conclusions: From our study, we have identified diverging recommendations for a range of COVID-19 related interventions. These recommendations may have important implications for clinical practice and public health decisions.