The intention of this article is to widen the practical use of IDEAL by clarifying the rationale for and practical details of the Recommendations. Additional research based on the experience of implementing these Recommendations is needed to further improve them.

The quality of clinical research in surgery has long attracted criticism. High-quality randomised trials have proved difficult to undertake in surgery, and many surgical treatments have therefore been adopted without adequate supporting evidence of efficacy and safety. This evidence deficit can adversely affect research funding and reimbursement decisions, lead to slow adoption of innovations, and permit widespread adoption of procedures that offer no benefit, or cause harm. Improvement in the quality of surgical evidence would therefore be valuable. The Idea, Development, Exploration, Assessment, and Long-term Follow-up (IDEAL) Framework and Recommendations specify desirable qualities for surgical studies, and outline an integrated evaluation pathway for surgery, and similar complex interventions. We used the IDEAL Recommendations to assess methodological progress in surgical research over time, assessed the uptake and influence of IDEAL, and identified the challenges to further methodological progress. Comparing studies from the periods 2000-04 and 2010-14, we noted apparent improvement in the use of standard outcome measures, adoption of Consolidated Standards of Reporting Trials (CONSORT) standards, and assessment of the quality of surgery and of learning curves, but no progress in the use of qualitative research or reporting of modifications during procedure development. Better education about research, integration of rigorous evaluation into routine practice and training, and linkage of such work to awards systems could foster further improvements in surgical evidence. IDEAL has probably contributed only slightly to the improvements described to date, but its uptake is accelerating rapidly. The need for the integrated evaluation template IDEAL offers for surgery and other complex treatments is becoming more widely accepted.

Ductal carcinoma in situ (DCIS), the noninvasive form of breast cancer (BC), comprises just over 20% of breast cancer cases diagnosed each year in the USA. Most patients are treated with local excision of the disease followed by whole breast radiation therapy. Total mastectomy is not an uncommon approach, and total mastectomy with a contralateral risk-reducing mastectomy has been on the rise in the past decade. In estrogen receptor-positive disease, patients are often offered endocrine ablative therapy with a selective estrogen receptor modulator or an aromatase inhibitor as both treatment and prevention. Local regional treatment options have no impact upon ultimate overall survival. Long-term survival rates are higher in patients with DCIS than with any other form of the disease. Are these strikingly high success rates a testament to effective treatment strategies or is there a significant subset of DCIS that was unlikely to ever progress to invasive ductal carcinoma? DCIS was not seen in the US prior to the advent of screening mammography. When compared to other countries, the USA has the highest utilization of screening mammography and the incidence rate of DCIS. Other lines of evidence include autopsy series examining the breast tissue of women who died of other causes, missed-diagnosis series and current retrospective reviews of DCIS, all align in support of the concept of DCIS as indolent in the majority of cases [3-14]. The evidence suggests that both patient and physician misconceptions about DCIS have led to overdiagnosis and over-treatment of DCIS. Recently, a gene expression profiling tool (12 gene assay, Oncotype DCIS) has emerged that shows considerable promise in predicting class in DCIS patients.

The Supported Champions model allowed all surgical teams to reduce delay for septic patients by more than 50%, using distinct Quality Improvement strategies to address local issues. Improvement was implemented in 4 diverse settings with a quarter of the level of expert input previously used in a single hospital.