SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Background

Guidelines published by the international gastroenterology societies establish standards of care and seek to improve patient outcomes.

Aim

We examined inflammatory bowel disease guidelines (IBD) for quality of evidence, methods of grading evidence and conflicts of interest (COI).

Methods

All 182 guidelines published by the American College of Gastroenterology, American Gastroenterological Association, British Society of Gastroenterology, Canadian Association of Gastroenterology, Crohn's and Colitis Foundation of America and European Crohn's and Colitis Organisation as of 27 September 2012 were reviewed. Nineteen IBD guidelines were found.

Results

Eighty-nine per cent (n = 17/19) of the guidelines graded the levels of evidence using seven different systems. Of the 1070 recommendations reviewed, 23% (n = 249) cited level A evidence; 28% (n = 302) level B; 36% (n = 383) level C and 13% (n = 136) level D. The mean age of the guidelines was 4.2 years. In addition, 61% (n = 11/19) of the guidelines failed to comment on COI. All eight articles commenting on COI had conflicts with 81% (n = 92/113) of authors reported an average 11.7 COI. Lastly, there were variations in the recommendations between societies.

Conclusions

Nearly half the IBD guideline recommendations are based on expert opinion or no evidence. Majority of the guidelines fail to disclose any COI, and when commenting, all have numerous COI. Furthermore, the guidelines are not updated frequently and there is a lack of consensus between societal guidelines. This study highlights the critical need to centralize and redesign the guidelines development process.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

In 1990 the Institute of Medicine (IOM) defined practice guidelines as ‘systematically developed statements to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances.[1]’ This IOM report describes five major ways that guidelines are utilized: assisting clinician decision making, educating individuals or groups, assessing and assuring quality of care, guiding allocation of resources and reducing the risk of legal liability for negligent care.[1] However, as noted in the Appraisal of Guidelines for Research & Evaluation (AGREE), the benefits of clinical guidelines are only as good as the underlying quality of evidence used to formulate the recommendations.[2, 3] It is therefore important that guidelines provide the most updated information based on strong evidence, as poor guidelines may be more harmful than helpful.[4] Unfortunately, to date, previous studies evaluating clinical guidelines have reported that the American cardiology, infectious disease and hepatology practice guidelines frequently base their recommendations on lower quality evidence or expert opinion.[5-7]

The key components to a useful clinical guideline include: the decision-making process (e.g. diagnosis and management), potential risks, review of relevant evidence, costs and presenting the information in a concise and easy-to-read format.[8] To standardize the guideline development process, multiple papers[8-10] and society statements[11-15] have been published providing guiding principles for the development of clinical guidelines.[8, 9] However, the process of this standardization varies greatly by society and by country.[11-15] Therefore, the AGREE statement was formulated as an international endeavour to improve the overall quality of practice guidelines.[2, 3]

The quality of clinical guidelines also may be limited by potential conflicts of interest (COI).[16] Particularly for guidelines in which recommendations are based on expert opinion or lower quality evidence, the presence of COI may result in bias.[4] Potential COI also may impact on the perception of the guidelines.[16-19] For example, a recent survey published in the New England Journal of Medicine noted that physicians were concerned about potential biases when studies were sponsored by pharmaceutical companies and therefore were less likely to believe the findings of the trials.[20]

As clinical guidelines are increasingly published by both national and international organizations, multiple guidelines often are written on the same topics. The World Health Organization (WHO) stipulates that if strong quality guidelines exist for the same topic by a recognized national leader, then duplicating this work is not appropriate.[21] Furthermore, it is not known if the recommendations are consistent across multiple societal guidelines.

We, therefore, performed a systematic review of major international gastroenterology and inflammatory bowel disease (IBD) society guidelines specifically on the topic of IBD that have been published on the professional websites from the American College of Gastroenterology (ACG), American Gastroenterological Association (AGA), British Society of Gastroenterology (BSG), Canadian Association of Gastroenterology (CAG), Crohn's and Colitis Foundation of America (CCFA) and European Crohn's and Colitis Organisation (ECCO). Our goal was to assess the overall quality of the evidence cited in formulating the recommendations, determine the methods used to grade the evidence, evaluate for differences in recommendations or grades of evidence, gauge potential conflicts of interest and highlight potential opportunities for improvement.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Guidelines

We reviewed the ACG, AGA, BSG, CAG, CCFA and ECCO societal websites[13, 15, 22-26] as of 27 September 2012 for published practice guidelines. We searched for IBD-focused clinical practice guidelines. These IBD-specific guidelines were further examined to determine if any grading system was used to assess the level of evidence for the recommendations. The quality of the level of evidence supporting the recommendations was evaluated for each individual guideline. In addition, the layouts of the guidelines were evaluated for consistency and easily identifiable recommendations. The guidelines and websites were also reviewed for any comment regarding planned updates to the current IBD-related guidelines. The guidelines were examined individually, in aggregate by society, between societies and in an overall analysis of all published IBD practice guidelines.

ACG (http://gi.org/clinical-guidelines/clinical-guidelines-sortable-list)

The ACG's published clinical practice guidelines[23] were reviewed for any IBD-specific guidelines. Individual recommendations were evaluated if the level of the quality of evidence supporting them was provided.

AGA (http://www.gastro.org/practice/medical-position-statements)

The AGA published practice guidelines were reviewed for any IBD-related guidelines.[22, 24] The AGA notes that both the medical position statements and technical reviews published in their archive section prior to 2003 were reviewed and reapproved by the Clinical Practice and Quality Committee on 19 September 2008,[22] and therefore, all their IBD-specific guidelines were evaluated. The AGA medical position statements were first reviewed for any grading of the quality of evidence. If no grading system was identified, then the technical review was analysed to determine if quality of evidence was assessed. Individual recommendations were analysed if the level of evidence supporting them was provided in either document.

BSG (http://www.bsg.org.uk/clinical/general/guidelines.html)

The BSG's published guidelines[13] were reviewed for grading of the evidence. Recommendations were further evaluated if the level of evidence supporting them was provided.

CAG (http://www.cag-acg.org/guidelines)

The CAG pub-lished guidelines[15] were reviewed for any IBD-related guidelines. Recommendations were evaluated if the level of evidence supporting them was provided.

CCFA (http://www.ccfa.org)

The CCFA website[26] was reviewed for IBD-specific published guidelines. Recommendations were evaluated if the level of evidence supporting them was provided.

ECCO (https://www.ecco-ibd.eu/publications/ecco-guidelines-publications.html)

The ECCO-published IBD guidelines[25] were reviewed. Recommendations were further evaluated if the level of evidence supporting them was provided.

Levels of evidence

Multiple systems were utilized to grade level of evidence by the different societies. To standardize the reporting of the levels of evidence, when feasible, we merged the grading systems into the standard ABC(D) grading system that has been utilized for most of the American Heart Association (AHA)/American College of Cardiology (ACC) guidelines and American Association for the Study of Liver Diseases:[5, 7]

Grade A Randomized controlled trials/meta-analysis

Grade B Single randomized control / nonrandomized

Grade C Expert opinion/case studies/standard of care

Grade D No grade of evidence provided for the recommendation

When assigning a Grade D level of evidence, only the guidelines that provided at least some grading of evidence for its recommendations were evaluated. If the guideline contained a clear recommendation as indicated by bullet point (or similar), or a recommendation was separated out from the remainder of the text and there was no accompanying grade or level of evidence noted, then a Grade D was assigned.

Table S1a, Table S1b and Table S2 describe the methods used to standardize the grading systems.

Comparison of recommendations

The societal recommendations for the therapeutic options for maintenance of remission in ulcerative colitis and treatment of a severe acute flare of ulcerative colitis (UC) were chosen as two representative topics to be reviewed for consistency and discrepancy across guidelines. If there were inconsistencies, the recommendations and level of evidence used to support the recommendations were further analysed.

Conflicts of interest

All guidelines were evaluated to determine if potential conflicts of interest (COI) were disclosed, if a comment was made that ‘no conflicts of interest reported’ or if there was no mention of conflicts of interest. If a potential COI was noted to be present, the guideline was further reviewed to determine the total number of authors with COI, number of COI for the first author and the number of conflicts recorded per author. Conflicts were subdivided into research awards/grants and other conflicts including: advisory board, speaker's bureau, consulting and industry sponsored continuing medical education activities. Government or nonprofit-based awards were not considered COI and were excluded from analysis. The conflicts of interest were assessed by individual society, between societies and in aggregate.

Review of the guidelines

All the guidelines were reviewed by two authors (JDF, AG) for the presence of grading the quality of evidence for the recommendations, the methods by which the evidence was graded, the clarity of the document layout, for conflicts of interest and for evaluating similarity or discrepancy of the UC recommendations. The merging of grading systems into the ABC(D) format was performed by two authors (JDF, ASC).

Exclusions

Guidelines that were not posted on the societies’ guidelines webpage were not considered in this analysis. If the evidence system used could not be consolidated into the ABC(D) format, then it was excluded from subgroup analysis. Lastly, if two versions of a guideline with the same title were published, the earlier guideline was excluded from further analysis.

See Figure 1 for inclusion and exclusion of guidelines.

image

Figure 1. Inclusion and exclusion of practice guidelines process. ACG, American College of Gastroenterology; AGA, American Gastroenterological Association; BSG, British Society of Gastroenterology; CAG, Canadian Society of Gastroenterology; CCFA, Crohn's and Colitis Foundation of America; ECCO, European Crohn's and Colitis Organisation; IBD, inflammatory bowel disease.

Download figure to PowerPoint

Data analysis

The Chi-squared test and Fisher's exact test were used for comparing proportions of graded evidence and COI reported between societal guidelines for small and large samples respectively. Subsequently, pairwise comparisons of societal guidelines were evaluated. Mean COI reported and mean years since publication were evaluated by one-way anova. Bonferroni corrections were used whenever multiple pairwise comparisons were made (for 10 pairwise comparison bonferroni-corrected P-value < 0.005), otherwise a P-value of 0.05 was considered significant. Analysis was done using Stata.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Guidelines grading of the quality of evidence

A total of 182 guidelines were reviewed as published on the societal websites of ACG, AGA, BSG, CAG and ECCO. Twenty (11%) of these guidelines specifically focused on IBD-related topics as indicated in the title of the guideline. One guideline by the CAG, ‘Infliximab for the treatment of Crohn's disease: review and indications for clinical use in Canada 2001[27]’ was excluded from analysis given a more recent version of the same titled guideline published in 2004 (see Figure 1).[28] Eighty-nine per cent (n = 17/19) of these guidelines graded the quality of evidence for their recommendations (Table 1 and Figure 2). Two guidelines, one from the AGA[29] and one from the BSG,[30] failed to provide any evidence for their recommendations and were excluded from further analysis. The CCFA website did not contain any published guidelines.

Table 1. Total number of practice guidelines with graded evidence and quality of evidence for recommendations
 CombinedACGAGABSGCAGECCOP-value
  1. ACG, American College of Gastroenterology; AGA, American Gastroenterological Association; BSG, British Society of Gastroenterology; CAG, Canadian Society of Gastroenterology; ECCO, European Crohn's and Colitis Organisation.

  2. a

     Duplicate guideline was excluded.

  3. b

     One guideline excluded due to inability to merge evidence.

Number of guidelines192333a8 
Guidelines with grades of evidence17 (89%)2 (100%)2 (67%)2 (67%)3 (100%)8 (100%) 
Recommendations10707161b18995654 
Recommendations with Grade A evidence23% (249)52% (37)26% (16)15% (29)43% (41)19% (126)<0.001
Recommendations with Grade B evidence28% (302)21% (15)31% (19)17% (33)16% (15)34% (220)<0.001
Recommendations with Grade C evidence36% (383)21% (15)30% (18)20% (38)26% (25)44% (287)<0.001
Recommendations with Grade D evidence13% (136)6% (4)13% (8)47% (89)15% (14)3% (21)<0.001
image

Figure 2. Comparison of quality of evidence of recommendations in 19 guidelines with graded evidence. ACG, American College of Gastroenterology; AGA, American Gastroenterological Association; BSG, British Society of Gastroenterology; CAG, Canadian Society of Gastroenterology; ECCO, European Crohn's and Colitis Organisation IBD IBD, Inflammatory Bowel Disease; UC, Ulcerative Colitis. * No evidence cited in guideline. ** Evidence cited but unable to merge into the ABCD level of evidence system.

Download figure to PowerPoint

Levels of evidence

Sixteen (of 17) guidelines that graded the level of evidence supporting its recommendations were able to be merged into the ABC(D) system. The AGA guideline ‘Colorectal Neoplasia in Inflammatory Bowel Disease, Diagnosis and Management’ was excluded from analysis because it used the U.S. Preventive Services Task Force (USPSTF) method of grading of evidence which could not be reconciled into the ABC(D) system.[31]

The remaining 16 guidelines had a total of 934 graded recommendations and 136 recommendations without any associated grade of evidence. Twenty-three per cent (n = 249) of recommendations were supported by level A evidence, whereas 28% (n = 302) were level B, 36% (n = 383) were level C and 13% (n = 136) were level D (Table 1 and Figure 2). The breakdown on grading of the level of evidence by society is shown in Table 1. The proportion of level A across societies differed (P-value < 0.01). The ACG and CAG reported more grade A evidence than the other societies. In post hoc analysis (bonferroni-corrected P-value < 0.005), ACG guidelines reported more grade A recommendations compared with AGA (P-value 0.003), BCG (P-value < 0.001) and ECCO (P-value < 0.001). Similarly, CAG guidelines reported more grade A recommendations compared with BSG (P-value < 0.001) and ECCO (P-value < 0.001), but not when compared with the ACG (P-value 0.27) and AGA (P-value 0.04). The ECCO reported more grade B evidence than BSG (P-value < 0.001) and CAG (P-value < 0.001); otherwise the remaining societies reported similar proportion of grade B evidence. ECCO reported more grade C guidelines than ACG (P-value < 0.001) and CAG (P-value < 0.001). BSG guidelines most frequently reported grade D guidelines when compared with the other four societies (P-value < 0.001 for all comparisons), whereas ECCO reported the least grade D guidelines except when compared with ACG guidelines (P-value 0.29).

Methods utilized to grade the evidence for recommendations and format

Seven different methods for grading quality of the evidence were utilized by the five societies (Table S1a and Table S1b). ECCO and the BSG used consistent methodology based on the Oxford Center for Evidence Based Medicine level of evidence system.[32] In contrast, ACG, AGA and CAG used multiple methods to grade the quality of evidence.

In addition, only ECCO used a consistent standardized format to present their recommendations in an easily identifiable section. The other guidelines utilized varying methods to indicate a recommendation and its associated grade of evidence (e.g. bullet points, italics, bolded text or incorporating the grade of evidence into the general text).

Age of guidelines and expected updates

The mean age of the guidelines was 4.2 years (s.d. 3.4). When analysed by society, the mean age of ACG guidelines was 2.5 years (s.d. 0.7), ECCO guidelines was 3 years (s.d. 0.9), CAG was 4 years (s.d. 3.6), AGA was 5.7 years (s.d. 3.5) and BSG was 7 years (s.d. 7.2). The mean age did not statistically differ across societies (P-value 0.43). The only guidelines noted to be in the process of being updated were the ECCO 2008 ulcerative colitis guidelines.

Table 2. Differences in recommendations
 ACGAGABSGCAGECCO
Maintenance on infliximab (level of evidence)Yes (level A)Yes (not graded)No (Level A)Yes (level A)Yes (Level A)
Maintenance with methotrexate (level of evidence)No (commented in the main text but not graded)No (Level B)Yes (Level C)No commentNo (commented in the main text but not graded)
Checking TPMT level (level of evidence)Commented on, but not recommended for or against testingYes (Level B)Yes (Level C)No commentCommented that no recommendation can be made regarding testing
Cessation of therapy (level of evidence)No commentNo commentYes (after 4 years) (Grade B)No commentNo comment
Rescue therapy with Tacrolimus (level of evidence)Commented on that evidence as insufficient to make recommendationNo commentCommented on that evidence as insufficient to make recommendationCommented on that evidence as insufficient to make recommendationYes (Grade A)

Comparison of recommendations

There were a number of notable differences in the societal recommendations regarding the therapeutic options for maintenance of remission in ulcerative colitis and treatment of a severe acute flare of ulcerative colitis. See Table 2 for the recommendations and the differing levels of evidence used to support the recommendations. The ACG, AGA, CAG and ECCO support the use of infliximab for maintenance of remission of ulcerative colitis, whereas the BSG recommends against the use of infliximab in this situation. In contrast, only the BSG recommends therapy with methotrexate for maintenance of remission in UC in those failing a thiopurine. Furthermore, only the BSG and AGA specifically recommend checking Thiopurine S-methyltransferase (TPMT) levels prior to initiating azathioprine or mercaptopurine. BSG alone comments that maintenance therapy for ulcerative colitis can be discontinued after 4 years, whereas the other societies do not comment on stopping therapy. Lastly, only ECCO recommends using tacrolimus for severe UC whereas the other societies cite a lack of evidence to support its use.

Table 3. Conflicts of interest in practice guidelines
 CombinedACGAGABSGCAGECCOP-value
  1. ACG, American College of Gastroenterology; AGA, American Gastroenterological Association; BSG, British Society of Gastroenterology; CAG, Canadian Society of Gastroenterology; ECCO, European Crohn's and Colitis Organisation; Avg, average; COI, conflict of interest; s.d., standard deviation.

  2. a

     Duplicate guideline was excluded.

Articles with NO COMMENT on COI11 (61%)01 (33%)2 (67%)08 (100%) 
Articles with NO COI present000000 
Articles with COI present8 (44%)2 (100%)2 (67%)1 (33%)3 (100%)0 
Total number of conflicts10701077180815aNA 
Mean number of conflicts/ article (s.d.)134.1 (134.1)53.5 (54.4)35.5 (19.1)80 (0)271.7 (123.5)NA0.12
Authors with COI present92/113 (81%)5 (100%)7/8 (88%)10/14 (71%)70/86 (81%)  
Avg number of conflicts/ first author (s.d.)13.7521 (9.9)19 (9.9)10 (0)6.7 (4.2)NA0.27
Avg number of conflicts/authors with conflicts (s.d.)11.8 (12.2)21.4 (14.7)10.1 (8.3)8 (2.4)11.8 (13.0)NA0.24

Conflicts of interest

Sixty-one per cent (11/19) of the IBD guidelines failed to disclose any information regarding COI. ECCO published 72% (8/11) of the guidelines that failed to comment on COI. The remaining three were published by the BSG and the AGA. The eight articles disclosing potential COI reported 1073 conflicts, of which 118 were research or grant and 955 were other COI. The average number of COI reported did not differ between societies that reported COI (P-value 0.12). All first authors disclosed COI with an average of 13.75 COI/author. The average COI per first author did not significantly differ between societies reporting COI (P-value 0.27). Similarly, there was an average of 11.7 COI for each author disclosing COI (n = 92). No difference was observed between societies in the average COI per author reporting COI (P-value 0.24). See Table 3 for breakdown of COI by societies. Only the CAG included disclosures of potential COI from all individuals partaking in the guideline development process in their guideline articles.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

The WHO and IOM comment that guidelines serve to advise physicians on treatments for their patients and to ultimately create a safer medical system.[21, 33] This goal is only achieved if guidelines are based on strong evidence, updated frequently and are succinct and easy to read.[4, 8] Ideally, the guidelines should be consistent across organizations,[4] and potential biases minimized. Unexpectedly, we noted that nearly 50% of IBD guidelines are based only on expert opinion and lack strong supporting evidence. We also noted that, when comparing recommendations for the treatment of ulcerative colitis, recommendations varied between societies with varying levels of supporting evidence. Furthermore, many guidelines failed to comment regarding any potential COI. In articles where COI was noted, all the articles reported multiple COI, with 81% of the authors disclosing potential COI.

The cost of developing new guidelines is quite substantial.[18, 34] The time commitment ranges on the order of months to years with costs in the hundreds of thousands of dollars to develop a guideline.[18, 34] The societies reviewed in this study have many overlapping guidelines. Although the methodologies and development processes differ between the societies, the overall content of the guidelines are fairly similar. ECCO published eight guidelines providing recommendations encompassing all areas in IBD care. Compared with ECCO, no other society included such comprehensive guidelines, and the topics and recommendations in the 11 guidelines from the other four societies were included in the ECCO guidelines.

Creating an international consensus working group would provide for more effective, consistent guidelines regardless of a country's geographic borders. Although there may be variations in recommendations between countries based on drug approvals or on cost-effective analysis, this can be included in an international guideline with qualifying statements specifically applicable to a certain region. The evidence, however, is unchanged regardless of the geographic location. In addition, at an approximate cost of $250 000 to $650 000 per guideline,[18, 34] the savings from avoiding duplicate guidelines amount to an average 4 million dollars and countless hours saved. The Guidelines International Network (GIN) is an international group that serves as a repository of clinical guidelines and has provided instructions that outline the minimum standard to develop high-quality clinical practice guidelines to improve patient care.[10, 35] Currently, 88 international organizations are members of the GIN with over 7000 guidelines published on their website.[35] Using a system like GIN or centralizing all IBD guidelines within an international group, like ECCO, could solve this problem.

As noted by Mendelson, potential conflicts of interest amongst authors of clinical practice guidelines may result in distrust of guidelines and lessen their value in establishing standards of care.[36] We found that over 50% of the IBD guidelines fail to comment on potential COI, and, when reported, the number of COI per guideline is substantial. Although it is common for clinical experts to have relationships with industry, transparency regarding the potential COI is critical.[17, 36] WHO and the National Institute for Clinical Excellence (NICE) in the United Kingdom have strict COI policy for disclosing potential COI. In addition, any panelist with an identified relevant COI may not participate in the relevant decision-making process.[14, 37] As guidelines are used to establish standard of care, transparency regarding COI and a system of checks and balances are necessary to avoid distrust in the guidelines.[36] We recommend the less stringent approach of the Council of Medical Specialty Societies (CMSS) to maintain the integrity of practice guidelines. The CMSS policy regarding COI has been adopted by multiple societies and does not entirely exclude clinical experts from participating in guideline development. Instead, it mandates that the guideline chair and majority of the panelists lack any relevant COI to the guideline being developed. In addition, it creates a conflict mitigation system that includes several reviewers and voters in the development process to minimize potential biases from experts with possible COI.[38] As the IOM notes, ‘the goal of conflict of interest policies in medicine is to protect the integrity of professional judgment and to preserve public trust rather than try to remediate problems with bias or mistrust after they occur.’ Ideally, an independent and international group that is transparent regarding any potential COI and is protected from external pressures by industry and politics should be tasked with developing clinical guidelines.[39] Ultimately, the purpose of transparency regarding potential biases is to protect the integrity of professional judgment.[16]

Although the majority of IBD guidelines grade the level of evidence, multiple different systems were utilized to accomplish this task. The use of multiple grading systems is confusing and makes it difficult for physicians to determine the validity of the recommendations. Ideally, guidelines should be user friendly with the recommendations clearly identifiable to the reader.[2, 3, 8] Although the GIN provides recommendations throughout the guideline development process, none discuss the actual format of the document.[10] In our study, only ECCO used a standardized format to present their recommendations in an easily identifiable section. Ideally, the guidelines should have a consistent layout with a table at the beginning of the document that summarizes all the recommendations, along with their grades of evidence. An effective way of standardizing the grading system is to use the GRADE system.[40-42] It rates the evidence as high, moderate, low and very low. A high quality indicates a strong evidence base and that further research is unlikely to change the confidence in the estimated effect. Moderate, low and very low quality denotes variable levels of certainty within which it is possible or likely that future studies will alter current practice.[40-42] This knowledge allows the physician to engage the patient in a shared decision-making process when deciding care based on varying levels of evidence. In addition, when the only available evidence results in a weak recommendation the guideline authors may opt to place the recommendation in a separate ‘best practice statement’. This would provide clinicians with a summary of the current knowledge on a specific topic, but would also alert the clinician and any other group using the guidelines that further studies are necessary which may alter the current statement. Such a process would further solidify the validity of a guideline with recommendations based on high-quality evidence while also summarizing current best practices based on the available literature.

A key feature of practice guidelines is to summarize the vast amount of literature published and to keep the physicians abreast of current standard of care. Although most societies adopt practices and recommend timely review of their practice guidelines and updates as needed,[11, 14, 43] some of the IBD guidelines are clearly dated. There have been many changes in the management of IBD in the last few years. For example, the AGA IBD guidelines fail to discuss FDA approved agents for Crohn's disease such as adalimumab (approved 2007), certolizumab pegol (approved 2008) or natalizumab (approved 2008).[44] When guidelines do not maintain updated recommendations consistent with current practice, they fail to serve the practicing physicians and patients. The utility of guidelines are diminished, and physicians are forced to seek alternate resources to clarify current options. When new diagnostic or therapeutic modalities are made available, practice guidelines must be updated with changing practice. Only actively updated guidelines will help create better and safer patient care.

Study limitations

Multiple systems were used to grade the quality of evidence by the various societies. To provide an adequate analysis of the level of evidence supporting the recommendations, we merged the grading systems into one system, ABC(D). To limit potential bias, this was performed by two authors. Only one guideline used a method too divergent to incorporate into the ABC(D) system. In addition, only the clinical practice guidelines published on the societal websites were reviewed. Guidelines or updates not available on the websites were not considered for this study. Furthermore, IBD guidelines published by other societies may exist, but were not considered in this analysis. Lastly, when assigning Grade D level of evidence to recommendations without any supporting evidence, only one Grade D was assigned even if multiple recommendations were included in one sentence. This potentially underestimates the total number of recommendations with ungraded evidence.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

The majority of IBD clinical practice guideline recommendations are based on low-quality evidence. Furthermore, the majority of guidelines fail to comment on potential COI, and when present, most of the authors report numerous COI. In addition, management recommendations vary between societies. This study highlights the need for improving the development of IBD practice guidelines. Recommendations would be improved by stronger supporting evidence, agreement between societies, up to date recommendations and transparency regarding all potential bias in the development process. Centralizing all IBD guidelines within an international group may help solve several of the issues. Given these concerns, clinicians must be cautious when utilizing the recommendations in published IBD practice guidelines.

Authorship

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Guarantor of the article: J. D. Feuerstein.

Author contributions: Dr Feuerstein and Dr Cheifetz had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Feuerstein and Cheifetz were involved in the study concept and design, drafting of the manuscript, and study supervision. Feuerstein, Gifford and Cullen were involved in acquisition of data. Feuerstein, Akbari and Cheifetz were involved in analysis and interpretation of data. Feuerstein, Cheifetz, Sheth and Leffler were involved in critical revision of the manu-script for important intellectual content. Akbari was involved in the statistical analysis. All authors approved the final version of the manuscript and contributed equally to the preparation.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Declaration of personal interests: Dr Adam Cheifetz has served as a consultant in the field of inflammatory bowel disease for Abbot Laboratories, Jansen Pharmaceuticals, Warner-Chilcott, Given Imaging and Prometheus Labs, and has received research funding from Pfizer. Dr Daniel Leffler has served as a consultant and/or received research support from Prometheus diagnostics, Alba Pharmaceuticals, Alvine Therapeutics and Shire Therapeutics.

Declaration of funding interests: None.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Authorship
  9. Acknowledgements
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
apt12290-sup-0001-TableS1-S2.docxWord document26K

Table S1. (a) Methods of merging evidence grading systems. (b) Evidence grading systems.

Table S2. Excluded evidence grading systems.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.