Study objectives

The primary aim of the APOLLO study is to describe the operative and nonoperative management of emergency CRC presentations in an international cohort. Secondary aims will be to describe 30- and 90-day management outcomes, identify the risk factors for intraoperative, 30-, and 90-day mortality and ostomy rates in patients deemed for active management (i.e., not for palliative management), and develop a mortality and ostomy risk prediction model for patients undergoing active management for CRC. This study will also aim to validate risk criteria of LBO in patients with previously known CRC undergoing neoadjuvant chemotherapy or awaiting elective surgery.

Study design

An international, multicentre, prospective observational study of patients presenting acutely with CRC will be delivered by the EuroSurg Collaborative using the trainee- and student-led collaborative model. EuroSurg is an international collaborative group supported by national collaborative groups including Student Audit and Research in Surgery (STARSurg, UK), the Portuguese Surgical Research Collaborative (PTSurg, Portugal), Student-Initiated German Medical Audit (SIGMA, Germany), the Italian Surgical Research Group (ItSURG, Italy), and Trials and Audit in Surgery by Medical students in Australia and New Zealand (TASMAN, Australia and New Zealand).

“Mini-teams” of collaborators will participate at each hospital. Team members will include medical students, junior doctors, trainees, registrars, and supervising consultants (Figure 1).

Study setting

The study is open to any secondary or tertiary hospital across the world with a general or colorectal surgery department performing major CRC surgery. All participating centres will be required to register the study according to local regulations prior to the commencement of data collection from each respective site. As this is an audit of practice, limited to using data obtained as part of usual care, no change to normal patient management is required.

At each hospital, one mini-team per period will collect data on eligible patients admitted over any consecutive 6-week period between January and June 2023. Patients will be followed up at 30- and 90-days after admission. Future dates and sites able to participate may be added at the discretion of the steering committee.

Eligibility criteria

Patients should be included regardless of cancer metastatic extent. Patients with known CRC presenting with progression of disease, that is, new acute obstruction, and not for side effects of treatment (chemotherapy/radiotherapy), are also included.

Patients should be excluded if they present acutely for the side effects of chemotherapy or radiotherapy. Nonadmitted patients, patients without adenocarcinoma, or patients with noncolorectal primary cancers that have metastasised to the colon or rectum (e.g., melanoma or lymphoma) should be excluded. Patients with primary neuroendocrine tumours, gastrointestinal lymphomas, gastrointestinal stromal tumours, and primary colorectal squamous cell carcinoma should also be excluded.

Outcomes

The primary outcome is 90-day mortality from the day of presentation. Secondary outcomes include (a) mortality at other timepoints (operative, in-hospital, 30-day), (b) rates of primary anastomosis, (c) rates of stoma formation and reversal, (d) rate and grade of surgical complications measured by Clavien Dindo Classification [10], (e) rate of colonic stenting and their associated complications, (f) representation, (g) rates of radiological assessment on admission, and (h) the proportion of patients presenting acutely with LBO with known CRC who had high risk criteria for obstruction (as defined by the FOXTroT obstruction criteria) at index assessment.

Audit standards

Other variables

Additional variables for risk-adjustment of analyses will be collected (Table S1). These include demographics, comorbidities, frailty score [23], reason for presentation, cancer type and location, cancer TNM and mutation status, CRC screening status, relevant investigations, management strategies, surgical details, and surgeon/proceduralist skill levels.

Data collection and management

Data will be collected and stored online via the Research Electronic Data Capture (REDCap) web application [12, 13], hosted and managed by the Birmingham Surgical Trials Consortium (BiSTC) REDCap system hosted at the University of Birmingham, UK.

REDCap allows collaborators to enter and store data in a system that is encrypted and compliant with HIPAA-Security Guidelines in the United States. The security of the study database system is governed by the policies of the University of Birmingham. Data management and data security within the BiSTC REDCap will abide by the requirements of the General Data Protection Regulations and any subsequent amendments. Collaborators will be given secure REDCap project server login details, allowing secure data storage on the REDCap system. No identifiable patient data will be uploaded or stored on the REDCap database without prior local permissions.

Patients will be followed up for 90-days after presentation. No change in routine follow-up will take place.

Study recruitment and power calculation

Patients will be included if their hospital admission starts (defined as “date of hospital admission”) within the time during the data collection periods as specified above.

Teams should collect data on consecutive eligible patients at their hospital during the data collection period. Strategies to identify consecutive patients could include daily review of theatre lists, handover meetings/sheets, and ward lists, new inpatient referrals to surgical teams, CRC multidisciplinary team meetings, and daily review of emergency department admissions.

Based on previous EuroSurg studies, APOLLO is anticipated to include 150 centres in the UK and 150 centres in Europe and other countries [14]. We anticipate that on average, around five patients will present per 6-week data collection period at each participating site giving a minimum sample size of 1500.

Statistical analysis

The primary descriptive analysis will be investigating emergency CRC presentation, their respective management, and short-term outcomes. Frequency of different presenting symptoms such as obstruction, perforation and bleeding will be described. Comparisons will be made between patients with different management approaches such as resection, primary anastomosis, end-stoma formation, stents, and palliation.

Planned secondary analyses include the development of a mortality risk prediction tool in emergency CRC surgery with curative intent as per the TRIPOD statement [15]. If the recruitment numbers allow, this aims to let surgeons preoperatively predict short term morbidity and mortality, in turn determining which patients may benefit from alternative bridging stents, and which may benefit from a primary anastomosis.

For model development, a mixed effects multivariable regression model will be built with candidate variables selected based on clinical plausibility. Relative model fits will be assessed using Akaike information criterion to obtain the most parsimonious model. Area under the receiver operating characteristic curve will be used to assess discrimination. Validation of the model will be performed as the methods of Bonsdorff et al. [16]. Statistical analysis will be completed using R (R Foundation Statistical Program).

Study deliverance and quality assurance

This protocol has been written with guidance from an international expert cross-speciality advisory group and with the contribution of patient representatives from ACPGBI. A data dictionary will be developed to help collaborators in collecting data and patient inclusion. E-learning materials will be available on the EuroSurg website (eurosurg.org).

Following data collection, only data sets with >95% data completeness will be accepted for pooled analysis. To emphasise the importance of data completeness to collaborators, centres with >5% missing data points will be excluded from the study and collaborators from those centres withdrawn from the published list of citable collaborators. Data validation will be performed to ensure all eligible cases were recruited.

Data governance and ethics

Ethical approval according to the requirements of each participating country will be completed. The study will be submitted for consideration as a clinical audit at eligible centres. Ethical approval documentation will be required from collaborators at each centre and country prior to the commencement of data collection. Data including ethical approval documentation will be stored in the REDCap secure online database provided by BiSTC.

Any changes to protocols will be mediated by the steering committee and communicated to all relevant ethical boards through local hospital teams.