Celiac Disease

CD: Methods (2021)

Celiac Disease Systematic Review and Guideline Methods
(Full methods in PDF)

According to the Institute of Medicine (National Academy of Sciences), “Clinical practice guidelines are statements that include recommendations intended to optimize patient care that are informed by a systematic review of evidence and an assessment of the benefits and harms of alternative care options”1. Clinical practice guidelines are developed by subject matter experts based on multiple factors like evidence, patient values as well as other crucial factors (such as financial cost, feasibility of implementation, or stakeholder buy-in). This section outlines and details the process and methods used in the development of the 2021 Evidence-based Nutrition Practice Guideline for registered dietitians working with individuals with celiac disease (CD).  The methodology for this project was developed using the process of the Academy of Nutrition and Dietetics2,3, in accordance with the Standards for Developing Clinical Practice Guidelines from the National Academy of Science using grading and guideline development tools from GRADE (Grading of Recommendations Assessment, Development and Evaluation) group.4,5 

Overview of the Guideline Development Process

Guideline development is a detailed and comprehensive process. The steps followed to develop the guideline are below (some steps were completed concurrently):

  1. Conduct an evidence-based scoping review to determine the availability of literature;
  2. Recruit an expert workgroup panel (content experts and patient advocates) that works with the evidence review team;
  3. Orient the work group the 5-step systematic review process of the Academy of Nutrition and Dietetics’ Evidence Analysis Center;
  4. Develop research questions and a priori eligibility criteria for the systematic review;
  5. Design a search plan and register it on the PROSPERO database;
  6. A medical librarian conducts a search of multiple databases;
  7. Screen abstracts and full-text articles based on a priori eligibility criteria;
  8. Extract data and critically assess the quality of included studies (risk of bias of studies);
  9. Synthesize evidence narratively (evidence summary and conclusion statements) and in table format (Study characteristics and findings table). Grade the quality of evidence for each outcome and provide GRADE tables;
  10. When evidence is available, expert workgroup panel members complete GRADE’s evidence-to-decision (EtD) framework to determine the best recommendations based on evidence, clinical expertise and patient values.
  11. When no evidence is available from the systematic review, expert workgroup panel members used these same principles (supporting evidence outside of the systematic review, clinical expertise and patient values) to develop consensus recommendations; 
  12. Recommendations are rated according to Academy principles and voted on and approved by workgroup members;
  13. For nutrition topics outside of the scope of the guideline, the expert workgroup panel members identified external evidence-based practice guidelines, and these were assessed for quality and individual recommendations voted on by workgroup members. 
  14. The evidence-based practice guideline was reviewed externally by 10 individuals with content expertise using the AGREE II tool;
  15. Respond to reviewer comments and update publication.

Expert Workgroup Panel Selection Process

To assure appropriate expertise and limit bias, the Work Group Selection sub-committee of the Academy’s Council on Research followed a transparent process of selecting an expert workgroup panel of subject matter experts. An open recruitment message with a link to an online application was circulated via stakeholders for experts in the topic of celiac disease.  Applications were reviewed by this sub-committee and six RDNs with extensive experience in nutrition care and /or research into individuals with celiac disease. Additionally, one patient advocate was recruited from Consumers United for Evidence-based Healthcare (CUE). This organization provides expertise in recruiting and training patient advocates. The members of the expert panel participated in all steps of the systematic review process and guideline development process described below. Academy staff and contractors supporting the expert workgroup panel included:  systematic review and guideline methodologists, a medical librarian, project manager, lead analysts, and trained evidence analysts. The expert workgroup panel and evidence review team would meet approximately twice per month in a virtual space to develop research questions, screen studies, analyze evidence, vote on and grade conclusion statements, and develop and discuss recommendations.

Guideline Focus

Based on the results of the scoping review, it was clear that recent evidence-based nutrition guidelines for individuals with celiac disease do not focus on medical nutrition therapy or guidelines to help registered dietitian nutritionists primarily working in the United States. Therefore, in this guideline, the expert workgroup panel focused on the effectiveness of nutrition interventions like FODMAP diet, gluten-free diet, prebiotics/probiotics, oats, and supplements. For nutrition topics outside of the scope of the guideline, external evidence-based guidelines were reviewed using the AGREE II tool and individual graded recommendations were voted on by expert workgroup panel members in order to provide practitioners with a comprehensive guide to celiac disease nutrition care. 

Systematic Review Process

Question Development, Literature Search and Study Selection 
This guideline followed the Academy of Nutrition and Dietetics systematic review methodology. During the initial teleconference calls, the expert workgroup panel developed a list of questions that were deemed important for clinicians and patients (Table 1). The expert workgroup panel developed the a priori inclusion and exclusion criteria as listed in the search plan (Table 2). The PICO questions and search plan for this systematic review were registered a priori on the PROSPERO database (#CRD42020169998)6

A comprehensive search of literature was conducted by an information specialist using MEDLINE (Ovid), EMBASE (Ovid), Cochrane CENTRAL (Ovid), CINAHL (Ebsco), Web of Science, and PsycINFO search engines. The literature search was conducted to identify studies addressing nutrition intervention questions in individuals with celiac disease. The inclusion criteria included: humans of all ages with celiac disease and were published between 2007 and December 2016. The search terms included terms to identify relevant nutrition interventions for patients (e.g. celiac, nutrition therapy, diet, dietitian, prebiotics, etc). The literature search focused on intervention questions and identified 5,294 potential studies.  The PRISMA diagram illustrating the study selection process is presented in Figure 1.  

After the search was completed, studies were systematically screened based on a priori inclusion/ exclusion criteria. For the intervention questions, randomized controlled trials, non-randomized trials and observational studies were included. The list of titles and abstracts were independently reviewed and marked for inclusion or exclusion (along with the reason) and any differences were resolved by discussion with a third reviewer. Full texts of articles meeting inclusion criteria were ordered and reviewed for inclusion. A total of 83 studies met the inclusion criteria for intervention questions. A list of excluded articles with reasons for exclusion was also created to maintain transparency (available at www.andeal.org/celiac).

Data Extraction and Study Quality Assessment

Relevant data was extracted from the included articles using a standardized online data extraction tool. Key information extracted from each study included: Authors information; year of publication; type of study design; details of intervention: type of intervention, duration of the intervention, who delivered the intervention, setting, number of centers; Participants: sample size, mean age, age range, gender, study inclusion and exclusion criteria, comorbidities; Interventions: intervention details, comparison group details, medication use; Outcomes: reported primary and secondary outcomes, time points of reported outcomes; other details such as funding source.

All included studies were critically appraised for risk of bias. Two independent reviewers assessed the quality or studies using the Academy’s online risk of bias tool, the Quality Criteria Checklist (QCC).4  The questions of the QCC are based on quality constructs and risk of bias domains identified by the Cochrane Collaboration and the Agency for Healthcare Research ad Quality (AHRQ). Questions examine sampling bias, performance bias, detection bias, attrition bias, and reporting bias.  Any discrepancies between the two reviewers were resolved by consensus or by a third reviewer.

Data Synthesis and Grading the Evidence

Descriptive synthesis of evidence was conducted for all identified outcomes for which studies were included. Meta-analysis was considered for the RCTs examining effect of nutrition therapy on primary outcomes, but data was insufficient for meta-analysis for all PICO questions/outcomes. 

After completion of the data extraction and data synthesis, systematic review results were provided in the following formats for the expert panel to review, edit, and approve: 1) Evidence summary: a narrative summary of all included trials for each identified outcome was drafted for each research question in the systematic review. A conclusion statement was developed for each proposed question /outcome. The conclusion statement is a clear, simple and to the point answer to the proposed questions.; 2) Study characteristics table: provided information regarding study characteristics, sample size, population, intervention details and quality of each included study; 3) Quality of evidence (strength of evidence):  Each of the conclusion statements were assigned a GRADE (reference) to reflect the quality of studies, inconsistency of results, imprecision, indirectness of the evidence, and publication bias.  Using this method, the evidence for each outcome of interest was graded as A (high), B (moderate), C (low), or D (very low). A GRADE table was generated using GradePro7 and demonstrated how the strength of evidence (GRADE) was derived for each outcome of interest.

Guideline Development

This guideline followed the Academy’s Evidence Analysis Center’s process for guideline development. For each nutrition topic investigated for which evidence was available, 2-4 workgroup members completed GRADE’s Evidence-to-Decision framework8,9, which guides review of the balance of benefits and harms, certainty of evidence, outcome importance, resource use, equity, patient values, acceptability and feasibility based on available evidence and clinical expertise in order to develop recommendations. When no or very little evidence was available to answer the systematic review questions posed, workgroup members discussed if, even in the absence of included evidence, recommendations were still needed to guide practice. If so, the workgroup drafted consensus recommendations based on: clinical expertise, literature outside of the systematic review; and nutrition principles and growth goals for the general population, with specifications that all practice decisions should be individualized according to the client. All consensus recommendations were discussed and approved unanimously by the workgroup. The workgroup members drafted comprehensive recommendations for nutrition counseling and care for individuals with celiac disease. During this phase, the role of the expert panel members was to translate the available evidence into action statements that were clear, concise, and ready to be implemented by practitioners. The workgroup and staff rated recommendations based on strength of evidence/confidence in findings and clinical experience. Strong recommendations use the terminology “recommend” and “should”, which means that this course of action should be applied to most people and practitioners can have confidence that implementing this recommendation has more benefit than risk. Weak recommendations use the terminology “suggest” and “may”. Terminology for Fair or consensus recommendations were at the discretion of workgroup members. The GRADE method involves two major components: a rating for quality of evidence (described above) and rating the strength of recommendations. The evidence grades are reported at the end of the recommendation statements (e.g A, B, C, or D) and reflect the confidence in the estimated effects (Table 3).

When providing the level for the strength of the recommendation, a number of factors besides the quality of evidence are taken into consideration, including patient values and preferences, quality of evidence, benefits and harms, cost/resources to implement the recommendation, acceptability, feasibility, and health equity.  In addition to evidence-based recommendations, in certain scenarios “Consensus” statements were developed. These statements were developed when there was not enough evidence or evidence was too low of quality to write a graded recommendation, but the workgroup determined it was important to provide some guidance to patients and practitioners. These recommendations are ungraded, and usually refer to general or routine practice.  

Once the full draft of recommendation statements was ready, it was reviewed and edited multiple times by all the workgroup members and the staff. The expert panel participated in a final blinded vote of recommendation statements, and a majority of votes approving the statement was necessary for each statement to be accepted into the final guideline. Each recommendation was approved unanimously by the workgroup members. 

For nutrition topics outside of the scope of this guideline, external evidence-based guidelines were reviewed using the AGREE II tool and individual graded recommendations were voted on by workgroup members in order to provide practitioners with a comprehensive guide to celiac disease nutrition care (see Overview of Nutrition Topics in Celiac Disease).

Draft Report with Supporting Rationale

Once the recommendation statements were developed, the work group members drafted a guideline manuscript based on the evidence and evidence to decision framework components, including: potential risks and harms, conditions of application, costs, recommendation narrative/rationale and rationale for the recommendation rating. In these sections, the workgroup members also cited additional references important to the respective topic, including discussion of studies published after our search dates or other systematic reviews on the topic. 

Peer Review Process

These guidelines underwent a systematic peer review process. An external review was conducted by 9 topic experts. The AGREE II tool (Appraisal of Guidelines for Research and Evaluation) criteria was used to assess the quality of guideline reporting. An additional external content review was conducted by the Celiac Disease Foundation in order to ensure feedback from a variety of stakeholders in the celiac community. Reviewer comments from all phases were collated by staff and sent to workgroup members for discussion and possible edits. The workgroup chair coordinated the final revision of the guideline document based on review comments and the final guideline manuscript will be submitted for publication. 


  1. Institute of Medicine Committee on Standards for Developing Trustworthy Clinical Practice G. In: Graham R, Mancher M, Miller Wolman D, Greenfield S, Steinberg E, eds. Clinical Practice Guidelines We Can Trust. Washington (DC): National Academies Press (US) Copyright 2011 by the National Academy of Sciences. All rights reserved.; 2011.
  2. Handu D, Moloney L, Wolfram T, Ziegler P, Acosta A, Steiber A. Academy of Nutrition and Dietetics Methodology for Conducting Systematic Reviews for the Evidence Analysis Library. J Acad Nutr Diet. 2016;116(2):311-318.
  3. Papoutsakis C, Moloney L, Sinley RC, Acosta A, Handu D, Steiber AL. Academy of Nutrition and Dietetics Methodology for Developing Evidence-Based Nutrition Practice Guidelines. J Acad Nutr Diet. 2017;117(5):794-804.
  4. Academy of Nutrition and Dietetics Evidence Analysis Library. Evidence Analysis Manual: Steps in the Academy Evidence Analysis Process: a Systematic Review and Guideline Manual 2016.
  5. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383-394.
  6. Feon Cheng, Deepa Handu, Rachel Sinley, Angie Almond, Marilyn Geller, Emily Germer, Joann McDermid, Kristen Roberts, Mary Sharrett, Theresa Taylor. Nutrition Interventions in Patients with Celiac Disease. PROSPERO 2020 CRD42020169998 Available from: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020169998
  7. GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. [computer program]. McMaster University; 2015.
  8. Moberg J, Oxman AD, Rosenbaum S, et al. The GRADE Evidence to Decision (EtD) framework for health system and public health decisions. Health research policy and systems. 2018;16(1):45.
  9. Alonso-Coello P, Schunemann HJ, Moberg J, et al. GRADE Evidence to Decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 1: Introduction. Bmj. 2016;353:i2016.



Proceed to References Return to CD home page