Ontario is Canada’s most populous province (estimated population of 14.6 million in 2019) and is a large and diverse region. The province includes large metropolitan areas, as well as sparsely populated rural communities. The demographics of Ontario residents include a wide range of socioeconomic backgrounds, a large number of first- and second-generation Canadians, and indigenous populations [5].

Residents of Ontario are eligible for a government-run, universal, single-payer health insurance plan (ie, the Ontario Health Insurance Plan). This plan covers the costs of physician visits, diagnostic investigations, hospital stays, surgical services, and cancer services (including radiation therapy and chemotherapy). Ontario routinely collects clinical and demographic data, as well as other health care–related data, within a number of linked administrative databases.

Individuals who were diagnosed with rectal cancer between 2010 and 2019 were identified through the Ontario Cancer Registry (OCR) by using the International Classification of Diseases, Tenth Revision (ICD-10), codes for rectal cancer (ICD-10 20.9). The OCR captures incident cancer cases through mandatory reporting and is estimated to include >95% of newly diagnosed cancer cases, in which >94% of colorectal cancers were microscopically confirmed [6,7]. Individuals with a histopathology code that was consistent with the adenocarcinoma of the rectum were included. Those aged <18 years at time of diagnosis, those who died on or before the date of diagnosis, and those who were ineligible for the Ontario Health Insurance Plan were excluded. We also excluded those who were diagnosed with cancer at another site (ie, breast cancer, prostate cancer, lung cancer, etc) within the preceding 3 years.

The OntaReCC combines data from a number of provincially maintained administrative data sets with data from histopathology reports that are made available through the OCR. A full listing of the administrative data sets can be found in Textbox 1. These data sets include health administrative data that are generated whenever a health care service is delivered to an individual, which are utilized for billing, registration, transactions, or record keeping by health care providers. The heath administrative data sets are linked to other provincially maintained data sets that provide vital statistics, census information, and other demographic data.

These data sets provide comprehensive data on individual characteristics (ie, demographics, comorbidities, and socioeconomic status), diagnostic details (the date of diagnosis, histologic variants, and cancer sites), and treatment details. In addition, health care provider details and institution details are captured within these data sets. By using the abovementioned data sets, we will be able to derive important factors or characteristics, including the completeness of preoperative investigations; the type, extent, and completeness of treatments (including surgery, radiation, and chemotherapy); treating clinician characteristics (volumes and practice setting) and institution characteristics (volumes, the level of care, and the presence of a multidisciplinary cancer center); and cancer-related outcomes (overall and cancer-specific survival, the treatment of metastatic disease, the use of palliative treatments, and symptom burden).

Histopathology reports were requested from the OCR for individuals who were undergoing biopsy, local excision/polypectomy, or major surgical excision for rectal cancer during the study dates. Cancer Care and Epidemiology at the Queen’s Cancer Research Institute has established processes for abstracting comprehensive details from these types of reports [8-10]. These processes allow for reliable and reproducible data abstraction. A data abstraction manual was created for abstraction. Trained histopathology data abstractors were utilized for data abstraction. Regular meetings for reviewing inconsistencies within and between reports were completed to ensure the validity and reliability of the pathology data abstraction. The data elements are summarized in Textbox 2.

The proposed cohort will contain a number of exposure and outcome variables that are pertinent to a wide range of research questions. Potential exposures include those related to individuals, the circumstances of the cancer diagnosis, health care providers, the institutions where care was delivered, the regions, and the nature and extent of treatments, as summarized in Table 1. A number of outcomes will also be examined and will include the intent and extent of treatments, adherence to recommended care, and short- and long-term outcomes. In addition, symptom burden will be examined, which is a unique feature of the proposed database.

Ethics approval for this research was obtained through the Queen’s University Health Research Ethics Board (approved on October 20, 2020; reference number: 6019418). A number of potential issues were addressed, including how personal health information will be used, the necessity for the linkage of personal health information with other information, and how the linkage would be completed. In addition, we identified potential harms and how they would be mitigated by using previously developed data confidentiality and privacy protocols, via appropriate data handling and curating, via the use of encryption, and via storage on a restricted server behind a firewall. The study will take place at the Queen’s Cancer Research Institute – Cancer Care Epidemiology, which has undergone previous external audits pertaining to data curation and security.

The overall objective of the OntaReCC is to provide a comprehensive data set of individuals diagnosed with rectal cancer in Ontario, and it will include details on the complete cancer journey (from diagnosis to death). This data set will allow investigators to assess the real-world, population-level outcomes of these individuals and allow for an assessment of the performance of the cancer system. A number of research themes will be explored by using the OntaReCC database, including (1) the regional variability in the delivery of care and outcomes, (2) the predictors and impact of adherence to recommended care, and (3) assessments of other understudied areas of rectal cancer care.

Regional variability in survival following the treatment of colorectal cancer has been reported by the Cancer System Quality Index [11]. This report demonstrated a 10% absolute difference in 5-year mortality between the best- and worst-performing regions within Ontario. Other publications identified similar themes [2], demonstrating large variations in 5-year survival between regions and between hospitals despite them having a similar case mix. Members of our group have explored the variation in the delivery of care for individuals with colon cancer. Within this population, significant variations were observed in the use of preoperative investigations [12], surgical quality and pathologic assessments [13,14], and treatment delivery to specific populations [8,9,15,16]. We anticipate similar themes for individuals with rectal cancer, and due to the increased complexity of care, we expect that the magnitude of difference will be greater. The proposed cohort will allow for the assessment of variability in the delivery of care based on geography, hospital type (ie, the presence of a multidisciplinary cancer center), and hospital volume. Importantly, we will use this comprehensive data set to identify predictors of variations in care and outcomes for potential knowledge translation and for interventions at the hospital and regional levels.

Cancer Care Ontario has developed recommendations for the care of individuals with rectal cancer [17]. These recommendations are consistent with those proposed in other jurisdictions within Canada and elsewhere [18]. Despite the wide availability of clinical practice guidelines and recommendations, significant variability in adherence has been demonstrated in a number of studies [19-24]. The OntaReCC will be utilized to assess the predictors of adherence to recommended care. In addition, we will be able to assess the impact of guideline-adherent care on patient outcomes.

Practice patterns and outcomes of understudied aspects of rectal cancer care will be assessed, including the treatment of early rectal cancer and end-of-life care.

As described above, the OntaReCC database will have a near-complete capture of incident rectal cancer cases in Ontario during the study period. Unlike the Surveillance, Epidemiology, and End Results program (around 30% capture) [30] and the National Cancer Database (around 70% capture) [31], the OCR has a near-complete capture of incident colorectal cancer cases [6,7]. As treatment for colorectal cancer is almost exclusively provided through the publicly funded health care system, the linked administrative databases provide a complete description of an individual’s cancer care journey. Low losses to follow-up are expected, regardless of the location or institution in which an individual receives care. By using detailed histopathology data, a comprehensive assessment can be undertaken, which is often not possible with other large population-level data sets or registries. The described cohort will also allow for an assessment of the symptoms that an individual experiences throughout their treatments and follow-ups, which is another unique aspect of the OntaReCC.

Although the proposed cohort will allow for a comprehensive assessment of individuals with rectal cancer, limitations to such population health research exist. First, the reasons behind treatment decisions and patient preferences are unknown. Nonadherence or variations in care may be partially explained by an individual refusing aspects of care or preferring nonstandard approaches, which will not be captured. Second, we will be unable to capture the reasons behind delays in treatment or care, which could be related to nonmodifiable factors. Third, we will not have access to detailed diagnostic radiology reports. This will limit our ability to assess the pretreatment stage of disease (which is primarily determined through magnetic resonance imaging) and the ability to fully capture the development of metastatic disease. Our data set will identify metastatic disease that is either treated (ie, via resection or other local therapies) or documented within applicable linked data sets. Finally, a number of potentially pertinent patient characteristics are not captured well within the available data sets, including individuals’ level of education, employment status, race/ethnicity, obesity, and alcohol or cigarette use.

The OntaReCC will provide a comprehensive and complete assessment of the care and outcomes of individuals diagnosed with rectal cancer in Ontario. The data set will be an important resource in evaluating the real-world management and outcomes of individuals with rectal cancer. In doing so, we will be able to identify opportunities for improving the delivery of cancer care to this group of individuals.