Predictors of 30-Day Mortality Among Dutch Patients Undergoing Colorectal Cancer Surgery, 2011-2016.

Importance: Quality improvement programs for colorectal cancer surgery have been introduced with benchmarking based on quality indicators, such as mortality. Detailed (pre)operative characteristics may offer relevant information for proper case-mix correction. Objective: To investigate the added value of machine learning to predict quality indicators for colorectal cancer surgery and identify previously unrecognized predictors of 30-day mortality based on a large, nationwide colorectal cancer registry that collected extensive data on comorbidities. Design, Setting, and Participants: All patients who underwent resection for primary colorectal cancer registered in the Dutch ColoRectal Audit between January 1, 2011, and December 31, 2016, were included. Multiple machine learning models (multivariable logistic regression, elastic net regression, support vector machine, random forest, and gradient boosting) were made to predict quality indicators. Model performance was compared with conventionally used scores. Risk factors were identified by logistic regression analyses and Shapley additive explanations (ie, SHAP values). Statistical analysis was performed between March 1 and September 30, 2020. Main Outcomes and Measures: The primary outcome of this cohort study was 30-day mortality. Prediction models were trained on a training set by performing 5-fold cross-validation, and outcomes were measured by the area under the receiver operating characteristic curve on the test set. Machine learning was further used to identify risk factors, measured by odds ratios and SHAP values. Results: This cohort study included 62 501 records, most patients were male (35 116 [56.2%]), were aged 61 to 80 years (41 560 [66.5%]), and had an American Society of Anesthesiology score of II (35 679 [57.1%]). A 30-day mortality rate of 2.7% (n = 1693) was found. The area under the curve of the best machine learning model for 30-day mortality (0.82; 95% CI, 0.79-0.85) was significantly higher than the American Society of Anesthesiology score (0.74; 95% CI, 0.71-0.77; P < .001), Charlson Comorbidity Index (0.66; 95% CI, 0.63-0.70; P < .001), and preoperative score to predict postoperative mortality (0.73; 95% CI, 0.70-0.77; P < .001). Hypertension, myocardial infarction, chronic obstructive pulmonary disease, and asthma were comorbidities with a high risk for increased mortality. Machine learning identified specific risk factors for a complicated course, intensive care unit admission, prolonged hospital stay, and readmission. Laparoscopic surgery was associated with a decreased risk for all adverse outcomes. Conclusions and Relevance: This study found that machine learning methods outperformed conventional scores to predict 30-day mortality after colorectal cancer surgery, identified specific patient groups at risk for adverse outcomes, and provided directions to optimize benchmarking in clinical audits.

Dutch Gastrointestinal Endoscopy Audit: automated extraction of colonoscopy data for quality assessment and improvement.

BACKGROUND AND AIMS: The importance of having quality assessment, assurance, and improvement tools in health care is increasingly recognized. However, the additional associated administration burden progressively interferes with the structural implementation and adoption of such tools, especially when it concerns high-volume procedures such as colonoscopies. The development of the Dutch Gastrointestinal Endoscopy Audit (DGEA), a registry with automated extraction of colonoscopy quality data, and its first results are described. METHODS: In close cooperation with commercial endoscopy reporting systems and a national histopathology database, healthcare professionals performing colonoscopies initiated a quality registry that extracts data from its core hospital resource or histology database without manual interference of the healthcare providers. Data extracted consisted of patient age, gender, indication of the colonoscopy, American Society of Anesthesiologists score, Boston Bowel Preparation Score, and cecal intubation; for the colonoscopy after a positive fecal immunochemical test in the colorectal cancer screening program, other data were polyp detection rate, which was available for all 48 hospitals or endoscopy centers, and adenoma detection rate, which was available for 26 hospitals or endoscopy centers. RESULTS: Between January 1, 2016 and March 31, 2019, 48 hospitals or endoscopy centers voluntarily participated in the DGEA, and 275,017 unique patients with 313,511 colonoscopies were registered. Overall missing values were limited to <1%. CONCLUSIONS: The results of this study demonstrate that it is feasible to deploy a quality registry collecting uniform data without additional administration burden for healthcare professionals.

Nationwide analysis of hospital variation in preoperative radiotherapy use for rectal cancer following guideline revision.

INTRODUCTION: The revised Dutch colorectal cancer guideline (2014), led to an overall decrease in preoperative radiotherapy (RT) use. This study evaluates hospital variation in RT use for resectable rectal cancer and the influence of guideline revision, including the nationwide impact of changing RT application on short term outcomes. METHODS: Data of surgically resected rectal cancer patients registered in the Dutch ColoRectal Audit were extracted between 2011 and 2017. Patients were divided into groups based on time of guideline revision (<2014 and ≥ 2014). Primary outcome was guideline adherence at hospital level regarding RT application, stratified for three stage groups. Secondary outcomes included positive circumferential resection (CRM+) and 30-day complicated postoperative course. RESULTS: The groups consisted of 7364 and 12,057 patients, respectively. In total, 6772 patients did not receive RT (17.6% (<2014) vs. 45.7% (≥2014), p < 0.001). The largest increase of surgery alone was observed for cT1-2N0 stage rectal cancer (35.1% vs. 91.8%, p < 0.001), with a substantial decrease in hospital variation (IQR 22.2-50.0% vs. IQR 87.6-98.0%). For cT1-3N1MRF- stage rectal cancer, a substantial amount of hospital variation in short course RT remained after guideline revision (IQR 26.8-54.1% vs. IQR 26.2-50.0%). A significant decrease in CRM+ (5.8% vs. 4.2%, p < 0.001) and complicated course (22.5% vs. 18.5%, p < 0.001) was observed. CONCLUSIONS: Radiotherapy for early-stage rectal cancer was uniformly abandoned after guideline revision, while substantial hospital variation remained for intermediate risk resectable rectal cancer in the Netherlands. The substantial nationwide decrease in the use of RT for rectal cancer treatment did not negatively impact CRM involvement.

Outcomes of surgical resections for benign colon polyps: a systematic review.

BACKGROUND: Not all benign colonic polyps are suitable for endoscopic resection, although criteria for endoscopic non-resectability vary worldwide. Clinical decision-making largely depends on endoscopic treatment options, as well as postoperative risks after surgical resection. This systematic review aimed to determine postoperative outcomes and the characteristics of surgically resected benign colonic polyps. METHODS: MEDLINE, EMBASE, and the Cochrane Library were searched for studies investigating the outcomes of surgical resection for benign colonic polyps since 1980. Studies were considered eligible when at least one postoperative outcome (morbidity and/or mortality) was reported. Meta-analyses were conducted for the primary outcome measures (morbidity and mortality) for studies that included patients only after the year 2000. RESULTS: Of the 4210 studies retrieved, 26 studies describing 139 897 patients were included. The most common indications for surgical resection were polyp location in the right-sided colon, non-pedunculated morphology, and large polyp size. The pooled 1-month complication and mortality rates of studies that included patients after the year 2000 were 24 % (95 % confidence interval [CI] 15 % - 36 %) and 0.7 % (95 %CI 0.6 % - 0.8 %), respectively. CONCLUSION: The postoperative morbidity and mortality after colonic resection for benign polyps are substantial. Referral to an advanced interventional endoscopist should be considered before referral for surgery to evaluate the possibilities for endoscopic treatment of large, non-pedunculated, and/or colonic polyps in difficult locations without suspicion of submucosal malignant invasion.

Definition of the Rectum: An International, Expert-based Delphi Consensus.

BACKGROUND: The wide global variation in the definition of the rectum has led to significant inconsistencies in trial recruitment, clinical management, and outcomes. Surgical technique and use of preoperative treatment for a cancer of the rectum and sigmoid colon are radically different and dependent on the local definitions employed by the clinical team. A consensus definition of the rectum is needed to standardise treatment. METHODS: The consensus was conducted using the Delphi technique with multidisciplinary colorectal experts from October, 2017 to April, 2018. RESULTS: Eleven different definitions for the rectum were used by participants in the consensus. Magnetic resonance imaging (MRI) was the most frequent modality used to define the rectum (67%), and the preferred modality for 72% of participants. The most agreed consensus landmark (56%) was "the sigmoid take-off," an anatomic, image-based definition of the junction of the mesorectum and mesocolon. In the second round, 81% of participants agreed that the sigmoid take-off as seen on computed tomography or MRI achieved consensus, and that it could be implemented in their institution. Also, 87% were satisfied with the sigmoid take-off as the consensus landmark. CONCLUSION: An international consensus definition for the rectum is the point of the sigmoid take-off as visualized on imaging. The sigmoid take-off can be identified as the mesocolon elongates as the ventral and horizontal course of the sigmoid on axial and sagittal views respectively on cross-sectional imaging. Routine application of this landmark during multidisciplinary team discussion for all patients will enable greater consistency in tumour localisation.

Postoperative Outcomes of Screen-Detected vs Non-Screen-Detected Colorectal Cancer in the Netherlands.

Importance: The nationwide fecal immunochemical test-based screening program has influenced surgical care for patients with colorectal cancer (CRC) in the Netherlands, although these implications have not been studied in much detail so far. Objective: To compare surgical outcomes of patients diagnosed as having CRC through the fecal immunochemical test-based screening program (screen detected) and patients with non-screen-detected CRC. Design, Setting, and Participants: This was a population-based comparative cohort study using the Dutch ColoRectal Audit and analyzed all Dutch hospitals performing CRC resections. Patients who underwent elective resection for CRC between January 2011 to December 2016 were included. Interventions: Colorectal cancer surgery. Main Outcomes and Measures: Postoperative nonsurgical complications, postoperative surgical complications, postoperative 30-day or in-hospital mortality, and complicated course (postoperative complication resulting in a hospital stay >14 days and/or a reintervention and/or mortality). A risk-stratified comparison was made for different postoperative outcomes based on screening status (screen detected vs not screen detected), cancer stage (I-IV), and for cancer stage I to III also on age (aged ≤70 years and >70 years) and American Society of Anesthesiologists score (I-II and III-IV). To determine any residual case-mix-corrected differences in outcomes between patients with screen-detected and non-screen-detected cancer, univariable and multivariable logistic regression analyses were performed. Results: In total, 36 242 patients with colon cancer and 17 416 patients with rectal cancer were included for analysis. Compared with patients with non-screen-detected CRC, screen-detected patients were younger (mean [SD] age, 68 [5] vs 70 [11] years), more often men (3777 [60%] vs 13 506 [57%]), and had lower American Society of Anesthesiologists score (American Society of Anesthesiologists score III+: 838 [13%] vs 5529 [23%]). Patients with stage I to III colon cancer who were screen detected had a significantly lower mortality and complicated course rate compared with non-screen-detected patients. For patients with rectal cancer, only a significant difference was found in mortality rate in patients with a cancer stage IV disease, which was higher in the screen-detected group. Compared with non-screen-detected colon cancer, an independent association was found for screen-detected colon cancer on nonsurgical complications (adjusted odds ratio, 0.81; 95% CI, 0.73-0.91), surgical complications (adjusted odds ratio, 0.80; 95% CI, 0.72-0.89), and complicated course (adjusted odds ratio, 0.80; 95% CI, 0.71-0.90). Screen-detected rectal cancer had significantly higher odds on mortality. Conclusions and Relevance: Postoperative outcomes were significantly better for patients with colon cancer referred through the fecal immunochemical test-based screening program compared with non-screen-detected patients. These differences were not found in patients with rectal cancer. The outcomes of patients with screen-detected colon cancer were still better after an extensive case-mix correction, implying additional underlying factors favoring patients referred for surgery through the screening program.

Achievements in colorectal cancer care during 8 years of auditing in The Netherlands.

INTRODUCTION: The efficacy of auditing is still a subject of debate and concerns exist whether auditing promotes risk averse behaviour of physicians. This study evaluates the achievements made in colorectal cancer surgery since the start of a national clinical audit and assesses potential signs of risk averse behaviour. METHODS: Data were extracted from the Dutch ColoRectal Audit (2009-2016). Trends in outcomes were evaluated by uni and multivariable analyses. Patients were stratified according to operative risks and changes in outcomes were expressed as absolute (ARR) and relative risk reduction (RRR). To assess signs of risk averse behaviour, trends in stoma construction in rectal cancer were analysed. RESULTS: Postoperative mortality decreased from 3.4% to 1.8% in colon cancer and from 2.3% to 1% in rectal cancer. Surgical and non-surgical complications increased, but with less reintervention. For colon cancer, the high-risk elderly patients had the largest ARR for complicated postoperative course (6.4%) and mortality (5.9%). The proportion of patients receiving a diverting stoma or end colostomy after a (L)AR decreased 11% and 7%, respectively. In low rectal cancer, patients increasingly received a non-diverted primary anastomosis (5.4% in 2011 and 14.4% in 2016). CONCLUSIONS: No signs of risk averse behaviour was found since the start of the audit. Especially the high-risk elderly patients seem to have benefitted from improvements made in colon cancer treatment in the past 8 years. For rectal cancer, trends towards the construction of more primary anastomoses are seen. Future quality improvement measures should focus on reducing surgical and non-surgical complications.

Laparoscopic conversion in colorectal cancer surgery; is there any improvement over time at a population level?

Conversion of laparoscopic colorectal cancer resection has been associated with worse outcome, but this might have been related to a learning curve effect. This study aimed to evaluate incidence, predictive factors and outcomes of laparoscopic conversion after the implementation phase of laparoscopic surgery at a population level. Patients undergoing elective resection of non-locally advanced, non-metastatic colorectal cancer between 2011 and 2015 were included. Data were extracted from the Dutch Surgical Colorectal Audit. Patients were grouped as laparoscopic completed (LR), laparoscopic converted (CONV) with further specification of timing (within or after 30 min) as registered in the DSCA, and open resection (OR). Uni- and multi-variate analyses were used to determine predictors of conversion and outcome (complicated course and mortality), with evaluation of trends over time. A total of 23,044 patients with colon cancer and 11,324 with rectal cancer were included. Between 2011 and 2015, use of laparoscopy increased from 55 to 84% in colon cancer, and from 49 to 89% in rectal cancer. Conversion rates decreased from 11.8 to 8.6% and from 13 to 8.0%, respectively. Laparoscopic hospital volume was independently associated with conversion rate. Only for colon cancer, the rate of complicated course was significantly higher after CONV compared to OR (adjusted odds ratio 1.486; 95% CI 1.298-1.702), and significantly higher after late (> 30 min) compared to early conversion (adjusted odds ratio 1.341; 1.046-1.719). There was no impact of CONV on mortality in both colon and rectal cancer. The use of laparoscopic colorectal cancer surgery increased to more than 80% at a national level, accompanied by a decrease in conversion which is significantly related to the laparoscopic hospital volume. Conversion was only associated with complicated course in colon cancer, especially when the reason for conversion consisted of an intra-operative complication, without affecting mortality.