Health Research Methods (HRM), Department of Clinical Epidemiology and Biostatistics, McMaster University, 1280 Main Street West, Hamilton, Ontario L8N 3Z5, Canada.
Mayo Graduate School, Mayo Clinic, Plummer 3-35, 200 First Street SW, Rochester, MN 55905, USA; Knowledge and Evaluation Research Unit, Mayo Clinic, Plummer 3-35, 200 First Street SW, Rochester, MN 55905, USA.
J Clin Epidemiol. 2016 Feb;70:111-22. doi: 10.1016/j.jclinepi.2015.09.006. Epub 2015 Sep 21.
Many strong recommendations issued by the World Health Organization (WHO) are based on low- or very low-quality (low certainty) evidence (discordant recommendations). Many such discordant recommendations are inconsistent with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) guidance. We sought to understand why WHO makes discordant recommendations inconsistent with GRADE guidance.
We interviewed panel members involved in guidelines approved by WHO (2007-2012) that included discordant recommendations. Interviews, recorded and transcribed, focused on use of GRADE including the reasoning underlying, and factors contributing to, discordant recommendations.
Four themes emerged: strengths of GRADE, challenges and barriers to GRADE, strategies to improve GRADE application, and explanations for discordant recommendations. Reasons for discordant recommendations included skepticism about the value of making conditional recommendations; political considerations; high certainty in benefits (sometimes warranted, sometimes not) despite assessing evidence as low certainty; and concerns that conditional recommendations will be ignored.
WHO panelists make discordant recommendations inconsistent with GRADE guidance for reasons that include limitations in their understanding of GRADE. Ensuring optimal application of GRADE at WHO and elsewhere likely requires selecting panelists who have a commitment to GRADE principles, additional training of panelists, and formal processes to maximize adherence to GRADE principles.
世界卫生组织（WHO）发布的许多强烈建议都是基于低质量或极低质量（低确定性）的证据（不一致的建议）。许多这样的不一致建议与 Grading of Recommendations Assessment, Development and Evaluation（GRADE）指南不一致。我们试图了解为什么世卫组织会提出与 GRADE 指南不一致的不一致建议。
我们采访了参与世卫组织批准的指南（2007-2012 年）的小组成员，这些指南包括不一致的建议。访谈记录并转录，重点关注 GRADE 的使用，包括其背后的推理以及导致不一致建议的因素。
出现了四个主题：GRADE 的优势、GRADE 的挑战和障碍、改进 GRADE 应用的策略以及对不一致建议的解释。不一致建议的原因包括对提出有条件建议的价值持怀疑态度；政治考虑；尽管评估证据为低确定性，但对收益的确定性很高（有时有道理，有时没有）；以及担心有条件的建议将被忽视。
世卫组织小组成员提出与 GRADE 指南不一致的建议，原因包括他们对 GRADE 的理解有限。要确保在世卫组织和其他地方最佳应用 GRADE，可能需要选择承诺遵循 GRADE 原则的小组成员，对小组成员进行额外培训，并制定正式流程，以最大程度地遵守 GRADE 原则。
