LAUSR

Referred to as a comic opera by Richard Horton of The Lancet in 1996, surgical research has frequently been criticized for poor standards of study design and reporting1,2. The critics have usually overlooked the significant challenges that face evaluation of surgical innovations, including problems with learning curves, randomization, blinding, equipoise, standardization of techniques and publication bias3. Surgical procedures are complex and initially evolve through rapid iterative changes as operators learn from early experience4. This invalidates the concept of comparison against a control treatment until the new procedure is stabilized. Surgeons typically develop their own variants of a standard technique and their own ideas about patient selection. However, consensus is needed regarding which variants and which patients are acceptable before a collaborative trial can be considered. Quality of delivery has a major impact on outcomes in surgery. If surgeons vary widely in their ability to perform the new operation, the statistical noise generated may obscure a real benefit, whilst the learning curve may affect outcomes adversely in the experimental group, introducing systematic performance bias into any randomized trial. The IDEAL (Idea, Development, Exploration, Assessment, Long-term study) Collaboration was founded in 2009 with the aim of improving the quality of the surgical evidence base5. Through its Framework and Recommendations, it established an integrated stepwise model for evaluating and reporting the evolution of complex interventions that recognized the challenges listed above and proposed specific study formats to deal with them6. On review of early publications that attempted to use IDEAL, however, it was clear that many of them were reported inaccurately7. To guide authors and journals better, the IDEAL Collaboration set out to develop reporting guidelines for IDEAL studies. Reporting guidelines serve to guide researchers when planning a research project, and specify the minimum necessary information to be included when writing a manuscript8. A multitude of reporting guidelines currently exist on the EQUATOR (Enhancing the QUAlity and Transparency Of health Research) Network (http:// www.equator-network.org), but none adequately addresses all stages of surgical innovation. The Annals of Surgery recently published an editorial9 calling for the development of comprehensive IDEAL reporting guidelines, and these have now been published in the same journal10. The IDEAL reporting guidelines were developed through a formal Delphi study, for which the protocol was published in 201811. The need for guidelines was considered separately for each stage of IDEAL, and the specific items needed at each stage were then examined by 61 individuals from the IDEAL Collaboration. They determined that there was a need for reporting guidelines for IDEAL stages 1, 2a, 2b and 4, but that IDEAL stage 3 appeared adequately covered by CONSORT12 and its extensions. Through two rounds of a Delphi survey and a final conference call, draft checklists for each stage were modified and refined. The resulting IDEAL statement paper presents reporting checklists for stages 1, 2a, 2b and 4 (Appendices S1–S4, supporting information), designed to help authors comply with the principles set out in the original IDEAL publications when designing studies and crafting manuscripts. The checklists will also help journal editors and reviewers to verify whether papers have in fact complied with the relevant IDEAL recommendations. Used in this way, the IDEAL reporting guidelines should help improve the quality, transparency and use of research in surgery and other complex interventions. The potential of IDEAL to raise the quality of scientific evaluation of surgery is clear, but the recommendations are not holy writ, having been produced by the rather traditional and fallible process of expert consensus (evidence level 5 on most of the pyramids constructed by teachers of evidence-based medicine). To avoid becoming dogma and to retain its value as a scientific tool, IDEAL needs to be tested, refined and validated through use. The authors therefore hope that the guidelines will stimulate more authors to plan and report their studies using IDEAL. In future years, meta-analysis of studies using the new IDEAL guidelines will determine whether they work as intended or need revision.