|Home | About | Journals | Submit | Contact Us | Français|
Clinical practice guidelines, which are systematically developed statements aimed at helping people make clinical, policy-related and system-related decisions,1,2 frequently vary widely in quality.3,4 A strategy was needed to differentiate among guidelines and ensure that those of the highest quality are implemented.
An international team of guideline developers and researchers, known as the AGREE Collaboration (Appraisal of Guidelines, Research and Evaluation), was established to create a generic instrument to assess the process of guideline development and reporting of this process in the guideline. Based on rigorous methodologies, the result of the collaboration’s efforts was the original AGREE instrument, which is a 23-item tool comprising six quality-related domains that was released in 2003 (www.agreetrust.org).
As with any new assessment tool, ongoing development was required to improve its measurement properties, usefulness to a range of stakeholders and ease of implementation. Over the years, a number of issues were identified. For example, the original four-point response scale used to answer each item of the AGREE instrument is not in compliance with methodologic standards of health measurement design. This noncompliance threatens the performance and reliability of the instrument.5 In addition, data on the usefulness of the AGREE items has never been gathered systematically from the perspectives of different groups of users. Further, we were interested in identifying strategies to make the evaluation process more efficient, such as reducing the number of items or the number of required raters, while ensuring the instrument was reliable and valid. Therefore, an exploration of the role of shorter versions of the AGREE instrument, comprising fewer items that are tailored to the unique priorities of different stakeholders, was warranted. Finally, there was a need to establish the fundamentals of construct validity — in other words, whether the AGREE items could measure what they purport to measure, and that is variability in quality of practice guidelines.
In response to these issues, the AGREE Next Steps Consortium was established and undertook two studies.6,7 As part of the first study, the consortium introduced a new seven-point response scale and evaluated its performance and measurement properties, analyzed the usefulness of the AGREE items for decisions made by different stakeholders, and systematically elicited stakeholders’ recommendations for changes to the AGREE items and domains.6 In the second study, the consortium evaluated the construct validity of the tool and designed and evaluated new supporting documentation aimed at facilitating efficient and accurate use of the tool.7
The following key findings emerged from the two studies:
Based on these results and three rounds of interpretation and consensus by the consortium, several refinements were made to the items and supporting documents, culminating in the release of AGREE II, which consists of 23 items, two overall assessment items and a user’s manual (see Appendix 1, available at www.cmaj.ca/cgi/content/full/cmaj.090449/DC1).
The 23 items in AGREE II are grouped into the same six domains as in the original AGREE instrument. These domains are scope and purpose, stakeholder involvement, rigour of development, clarity of presentation, applicability, and editorial independence. The key changes from the original document involved refinements to the purpose, response scale and items of the instrument.
The purpose of the AGREE II is more explicitly stated. The new version of the instrument is designed to assess the quality of practice guidelines across the spectrum of health, provide direction on guideline development, and guide what specific information ought to be reported in guidelines. The four-point response scale was replaced by a seven-point response scale, in compliance with key methodologic principles of test construction.5 A score of 1 indicates an absence of information or that the concept is very poorly reported. A score of 7 indicates that the quality of reporting is exceptional and all of the criteria and considerations articulated in the user’s manual were met. A score between 2 and 6 indicates that the reporting of the AGREE II item does not fully meet criteria or considerations. As more criteria are met and more considerations addressed, item scores increase (see user’s manual below). Finally, modifications, deletions and additions were made to approximately half of the original 23 items (Table 1).
The user’s manual (Appendix 1) was rewritten and extended with the following information linked to each item:
The consortium recommends that the AGREE II replace the original AGREE instrument8 as the preferred instrument for guideline development, reporting and evaluation. We used high-quality methods to direct the improvements made, with strong empirical evidence supporting the changes.6,7
As with the first version of the AGREE, the items and domains in AGREE II focus on methodologic issues relevant to guideline development and reporting. However, they do not evaluate the clinical appropriateness or validity of the recommendations themselves. While rigorous development and explicit reporting are necessary, they do not guarantee optimal and acceptable recommendations or better health outcomes for patients and populations.9,10 The new item assessing the description of strengths and limitations of the body of evidence (i.e., item 9) can be considered as a precursor for clinical validity or appropriateness of the recommendations. The consortium is targeting this area as its next priority for further study in the AGREE A3 initiative. This research initiative, funded by the Canadian Institutes of Health Research, is focused on the application, appropriateness and implementability of recommendations in clinical practice guidelines.
Similarly, some of the concepts in AGREE II could be improved. For example, the consortium considerably debated the representation of patient–public engagement in guideline development, as well as the items related to applicability and implementability in the instrument. These areas are also being targeted for future research.
Depending on the structure and length of the guideline document, quality-related assessment of a guideline using AGREE II will take 1.5 hours, on average, per appraiser. Although basic knowledge of the principles of evidence-based decision-making and health care methodology can facilitate its use, the new user’s manual should allow novices to use the instrument with confidence. Furthermore, although content-specific expertise on the topic of a guideline is not necessary, it may improve the ease of interpretation of the findings. At this time, we recommend that at least two appraisers, and preferably four, rate each guideline to ensure sufficient reliability as the consortium continues its formal reliability testing.
The AGREE II has been used to evaluate several hundred guidelines related to the control of cancer (www.cancerview.ca; select “Services” in the menu bar and click on the “SAGE” link). It will be available on the AGREE Research Trust website (www.agreetrust.org).
AGREE II has myriad uses. Guideline developers can incorporate the concepts of the AGREE II framework into their development protocols, procedural documents and reporting templates. The instrument can also be used to evaluate the quality of guidelines that are candidates for use in clinical practice, for formulating policy-related decisions or for adaptation of recommendations from one context to another. Journal editors and reviewers may use AGREE II as a framework to help define reporting requirements for guidelines submitted for publication, as has been done with the CONSORT (Consolidated Standards of Reporting Trials)11 and STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statements.12 Finally, given the increasing number of guidelines developed worldwide, AGREE II provides a framework for reaching consensus on methodologic principles and reporting requirements for transnational cooperation.
Other tools to support the application of AGREE II are being developed, including a translation into French, an online version and an interactive online AGREE II training tool. The AGREE Research Trust, an independent body established in 2004, manages the interests of the AGREE project, supports an agenda of research regarding its development and formally endorses AGREE II.
The AGREE II, along with support tools and information about ongoing research-based initiatives associated with the instrument, is available at www.agreetrust.org.
The AGREE Next Steps Consortium thanks the US National Guidelines Clearinghouse for helping to facilitate the identification of eligible practice guidelines for the research program. The consortium also thanks Ms. Ellen Rawski for her support on the project as research assistant from September 2007 to May 2008.
This article has been peer reviewed.
Competing interests: Melissa Brouwers, Francoise Cluzeau and Jako Burgers are trustees of the AGREE Research Trust. No competing interests declared by the other authors.
Contributors: Melissa Brouwers conceived and designed the study, led the collection, analysis and interpretation of the data, and drafted the manuscript. All of the authors made substantial contributions to the study concept and the interpretation of the data, critically revised the article for important intellectual content and approved the final version of the manuscript to be published.
Funding: This work was supported by the Canadian Institutes of Health Research (CIHR). Michelle Kho is supported by a CIHR Fellowship Award (Clinical Research Initiative).
Members of the AGREE Next Steps Consortium: Dr. Melissa C. Brouwers, McMaster University and Cancer Care Ontario, Hamilton, Ont.; Dr. George P. Browman, British Columbia Cancer Agency, Vancouver Island, BC; Dr. Jako S. Burgers, Dutch Institute for Healthcare Improvement CBO, and Radboud University Nijmegen Medical Centre, IQ Healthcare, Netherlands; Dr. Francoise Cluzeau, Chair of AGREE Research Trust, St. George’s University of London, London, UK; Dr. Dave Davis, Association of American Medical Colleges, Washington, USA; Prof. Gene Feder, University of Bristol, Bristol, UK; Dr. Béatrice Fervers, Unité Cancer et Environement, Université de Lyon – Centre Léon Bérard, Université Lyon 1, EA 4129, Lyon, France; Dr. Ian D. Graham, Canadian Institutes of Health Research, Ottawa, Ont.; Dr. Jeremy Grimshaw, Ottawa Hospital Research Institute, Ottawa, Ont.; Dr. Steven E. Hanna, McMaster University, Hamilton, Ont.; Ms. Michelle E. Kho, McMaster University, Hamilton, Ont.; Prof. Peter Littlejohns, National Institute for Health and Clinical Excellence, London, UK; Ms. Julie Makarski, McMaster University, Hamilton, Ont.; Dr. Louise Zitzelsberger, Canadian Partnership Against Cancer, Ottawa, Ont.