OICR-CCO Health Services Research Network
The Health Services Research Network is a collaboration between OICR and Cancer Care Ontario (CCO) to provide the knowledge needed to optimize the delivery of cancer services today and to ensure appropriate dissemination of health service innovations and well-evaluated technologies.
Six priority areas
Assessing the Real-World Clinical and Economic Outcomes of Emerging Innovative Technologies in Oncology: The Cases of Biosimilars and CAR T-cells
Kelvin Chan, Sunnybrook Research Institute
Scott Gavura, Cancer Care Ontario
Wanrudee Isaranuwatchai, St Michael’s Hospital
Biosimilars and CAR T-cell therapy are two types of emerging innovations that have the potential to significantly impact care of cancer patients. For different reasons, both biosimilars and CAR T-cells have limited comparative clinical data to help evaluate effectiveness and safety against current treatment approaches. Given the limited Canadian experience and the substantial budget impact associated with adopting emerging technologies, incorporating these innovations into the health system requires thoughtful assessment of the expected health and resource impacts.
This project will examine the real-world uptake, safety, effectiveness and economic impact of the implementation of biosimilar bevacizumab in advanced colorectal cancer and also evaluate the real-world health outcomes and economic impact of CAR T cell therapy in Ontario.
The evidence generated by these evaluations can be used to make informed decisions regarding the real-world impact of these new technologies, and to help develop proper policies and guidelines to ensure the best evidence-based care for people with cancer.
Does integration of the Predict 2.1 algorithm into routine clinical practice improve the value of multigene assays in early stage breast cancer patients? A prospective cohort study with a pre- versus post-comparison
John Hilton, The Ottawa Hospital/University of Ottawa
Mark Clemons, The Ottawa Hospital/University of Ottawa
A number of prognostic and predictive tools are available for patients with newly diagnosed breast cancer. These range from free and publicly available mathematical algorithms (e.g. Predict 2.1) to expensive gene array tests (e.g. Oncotype DX). Given that patients with early stage breast cancer represent a significant burden of cancer and indiscriminate use of Oncotype DX provides poor value, it is imperative to rationalize the use of Oncotype DX. Predict 2.1 may represent a helpful tool in selecting which patients should undergo additional Oncotype DX testing.
Evaluate whether incorporating Predict 2.1 results as standard of care affects ordering of additional Oncotype DX tests, improves the patients’ cancer care experience and the value of Oncotype DX testing.
Appropriate use of Oncotype DX could reduce costs to the province by millions of dollars a year and increase the value of Oncotype DX testing.
Symptom Burden Among Patients with Stage IV Lung Cancer: Analysis of Province-Wide Patient-Reported Outcomes
Natalie Coburn, Sunnybrook HSC
Alexander Louie, Sunnybrook HSC
In an effort to improve symptom management for cancer patients, CCO implemented a program whereby patients report their symptoms via the Edmonton Symptom Assessment System (ESAS) during cancer clinic visits. ESAS is a tool for assessing the severity of nine common cancer-associated symptoms. The goal of the program is for physicians to use patient-reported symptom information to address patient symptom concerns. However, it is not clear what interventions occur in response to a high score, or the impact on subsequent symptom burden or resource utilization. In order to improve the symptom-screening program more information is needed on which patients are most at risk for high symptom burden, what supportive care interventions are given for high scores, and how healthcare utilization is impacted by these interventions.
In a population of lung cancer patients, this project will describe symptom trajectories and identify predictors of high ESAS scores; determine relationship between high ESAS scores and use of supportive care and therapies; and compare healthcare utilization and cost of patients receiving supportive care for ESAS vs those not receiving care.
Establishment of symptom burden and interventions is the first critical step to improve proactive patient-centred supportive care for this population.
Building an Artificial Intelligent System to Enhance Online Support Group in Cancer
Yvonne Leung, UHN/ University of Toronto
Mary Jane Esplen, UHN/ University of Toronto
Cancer Chat Canada (CCC) offers therapist-led text-based online support groups (OSG) to address patients’ cancer-related distress with positive results. However, therapists often feel challenged to lead OSG while addressing individual group member’s distress /needs in the absence of visual cues. Recent advances in artificial intelligence (AI) may offer novel solutions. AI can be applied to address the challenges that therapists face attending to group members’ emotional reactions, by capturing real-time patient experiences during the text-based group session and predicting patterns of issues well ahead of time.
This project will develop and evaluate an AI system entitled ‘Artificial intelligence Co-Facilitator (AICF)’. AICF will serve as a “co-facilitator” by alerting and providing real-time actionable analytics based on the OSG discussion texts.
AI will enhance patient experience of CCC by providing therapists individualized information, enabling identification of issues amendable to treatment, and direct referrals for more specialized care in a timely manner. This project will improve online psychosocial oncology services currently being delivered by CCC.
Using real-world data and iterative economic evaluation to prioritize resource allocation for care and research in patients with relapsed/refractory B-cell acute lymphoblastic leukemia or diffuse large B-cell lymphoma
Kednapa Thavorn, Ottawa Hospital Research Institute
Natasha Kekre, The Ottawa Hospital/University of Ottawa
Harry Atkins, The Ottawa Hospital/University of Ottawa
The CAR T-cell therapy has just been approved in Canada for advanced leukemia but does not yet have a funding mechanism in Ontario. The therapy relies on a complicated manufacturing process that involves collecting a patient’s cells and then reengineering them to attack cancer cells. This process can be long and costly. To gain a better understanding of the factors contributing to the cost of this expensive treatment, the potential clinical benefits and what the health system can afford, a continuous economic evaluation of CAR T-cell therapy is needed.
To assess the cost-effectiveness of CAR T-cell therapy among adults with r/r B-cell ALL; assess whether it is feasible to integrate real-world data, including Ontario’s health administrative and registry data to inform the economic evaluation of CAR T-cell therapies; and to develop a framework to engage stakeholders in economic evaluations.
The results of this study will help facilitate realistic commercial valuation of CAR T-cell therapy and assist public payers in setting a price for CAR T-cell therapy based on the additional benefit gained and the other associated costs of the treatment.
Long-term healthcare dependency outcomes of older adults undergoing cancer surgery
Julie Hallet, Sunnybrook Research Institute
Older adults (OA) who experience cancer want information about their ability to live independently when making choices about care. When surgery is an option, there is currently no information available on the likelihood of living independently at home after one undergoes treatment. There is no information to support conversations between both patients and physicians, and patients and their families. These limited discussions can cost the health care system and patients.
This project will assess the likelihood of returning home, requiring home care services or being admitted into a long-term care facility following cancer surgery. A web-based tool will be developed that can be used by patients and healthcare providers to estimate their likelihood of experiencing those outcomes following cancer surgery. The tool will support counselling, decision-making, and preparation for surgery.
This study will address long-term healthcare dependency for OA undergoing cancer surgery. By providing needed information, it will contribute to enhancing patient-centred care for OA facing treatment choices for cancer surgery, and to optimizing healthcare resource utilization, sustainability, and system planning.
A “real-world” evaluation of healthcare utilization and the economic burden associated with azacitidine therapy for higher-risk myelodysplastic syndromes
Rena Buckstein, Sunnybrook Odette Cancer Centre
Lee Mozessohn, Sunnybrook Odette Cancer Centre
Myelodysplastic syndromes (MDS) are stem cell disorders characterized by progressive bone marrow failure and increased production of immature cells (blasts) with the development of acute myeloid leukemia (AML). Current standard of practice for patients with higher-risk MDS/low-blast count AML is treatment with the drug azacitidine (AZA). This was largely based on a pivotal clinical trial demonstrating a survival advantage that did not include an economic component to their analysis. Despite the survival advantage, “real-world” data on healthcare resource utilization and the economic impact of this patient population is lacking.
To evaluate this, this project will link the largest registry of higher-risk/low-blast count AML patients treated with AZA to administrative databases at IC/ES. In so doing, this project will reveal important information on resource utilization, cost and predictors of healthcare usage including comorbidity and disease characteristics to better understand their impact on survival in the “real-world”.
Knowledge gained from this study will have a significant impact on patients in Ontario and beyond with higher-risk MDS/low-blast count AML treated with AZA. It will inform patients and healthcare providers on the expected “real-world” risks and costs of AZA therapy.
Multi-gene panel testing for tailored treatment of advanced cancer in Ontario: What is the cost and benefit?
Timothy Hanna, Queen’s University Cancer Research Institute
Multi-gene panel testing with next generation sequencing (NGS) is a powerful technique that can be used to identify specific cancer mutations to tailor drug treatments for individual patients. This technique is funded for patients in Ontario with hard-to-treat cancers through provincial initiatives (IMPACT/COMPACT and OCTANE). The utility of NGS when used in routine practice is not well-defined. The proportion of patients who receive a targeted treatment identified by NGS in Ontario is unknown, as is the impact of NGS on overall survival. Building an understanding of the impact on treatment and survival is required to define the cost-effectiveness of this innovation.
Evaluate the impact of NGS testing on subsequent systemic treatment; compare health resource utilization and cost based on use of NGS; evaluate the impact of NGS testing on overall survival; and describe the cost-effectiveness of NGS in Ontario.
By quantifying outcome and cost of NGS, this study will improve the health care system by providing policy makers with the information they need to ensure value-based investment in cancer care. The methods utilized in this study will also inform the development of a conceptual framework for health technology assessment for innovations in cancer control in Ontario, further ensuring optimization of outcomes and health spending in Ontario.
Analytics and machine learning for disease pathway concordance
Timothy Chan, University of Toronto
Claire Holloway, Cancer Care Ontario
Disease pathway concordance (DPC) refers to the degree of alignment between the actual care patients receive and ideal care described in a clinical pathway. Previous research confirmed the feasibility of developing a population-based measure of DPC. However, the approach used “weighted” each activity in a pathway equally; e.g., missing treatment was considered equally disadvantageous as repeated imaging. A more refined concordance measure that ascribes different weights to different pathway elements is required to improve accuracy and generalizability. Recently, this research group demonstrated feasibility of a data-driven approach to generating importance-based weights for elements of the colon cancer pathway using inverse optimization, a state-of-the-art machine learning method.
Develop a general methodology for measuring DPC based on inverse optimization, applicable to all cancers; and apply the concordance metric to real patient survival data for validation and for exploring drivers of concordance and its variation.
By improving DPC measurement, it will be possible to evaluate complex practice patterns at the population level and describe clinical (survival) and system (cost) outcomes in relation to variation in care. These data could provide decision support to clinicians and health system administrators.
Small area geographic inequalities in cancer incidence and survival in Ontario: an advanced spatial analysis
Prithwish De, Cancer Care Ontario
Todd Norwood, Cancer Care Ontario
Cancer affects many Ontarians, and there are large differences in the cancer rates across the province. It is important to understand where these differences occur, how they have changed over time, and explore possible reasons for these differences.
This study will analyze nine cancer types and explore how social and economic factors, urban-rural location and other factors may affect both cancer rates and cancer survival at the community level.
The results of this study will inform advocacy, policy, clinical care and educational sectors to better address geographical disparities in the burden of cancer.
Dr. Christine Williams
Deputy Director and Interim Head, Clinical Translation
Find out more about what’s happening in Health Services Research at OICR News.
Dr. Olusola Dokun