Deciding what type of evidence and outcomes to include in guidelines: article 5 in Integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report.

INTRODUCTION

Professional societies, like many other organizations around the world, have recognized the need to use more rigorous processes to ensure that health care recommendations are informed by the best available research evidence. This is the fifth of a series of 14 articles that were prepared by an international panel to advise guideline developers in respiratory and other diseases on approaches for guideline development. This article focuses on what type of evidence and outcomes to include in guidelines.

METHODS

In this review we addressed the following topics and questions. (1) What methods should be used to select important outcomes? (2) What types of outcomes should be considered? (3) What sources of evidence should be considered? (4) How should the importance of outcomes be ranked? (5) How to deal with surrogate outcomes. (6) What issues related to outcomes should be considered in the evidence review? (7) What quality of evidence should be used? (8) How to interpret the effect on outcomes. (9) How to incorporate outcomes related to harm. We based our responses on a PubMed literature review, prior reviews, relevant methodological research, and workshop discussions.

RESULTS AND DISCUSSION

Guideline panels should use transparent and systematic methods to select both the evidence and important outcomes, with input from groups that represent a wide range of expertise and constituencies. Outcomes should address both benefits and downsides, with consideration of the definitions, severity, and time course of the outcomes. Guideline panels should use a transparent approach to rank outcome importance recognizing that stakeholder and patient values and preferences may vary. Intermediate and surrogate outcomes are frequently reported, but their correlation with patient important outcomes may be low. A guideline panel should determine a priori the magnitude of effect judged clinically significant, factors that may influence outcome reporting, and whether different ways of measuring the outcomes permit the outcomes to be combined. Comprehensive identification of the evidence includes the use of multiple data sources. While randomized controlled trials (RCTs) provide the highest quality evidence, reviewers of evidence also need to consider nonrandomized studies such as case series, registries, and case-control studies if randomized trials are not available. This is particularly true for harms. The outcomes reported from RCTs may not always directly apply to clinical practice settings (i.e., they may not be generalizable).