Sep 01, 2020
By Katherine H. Crawford, MA, PMP, ASCO Communications
"There is a future in which our information systems will help us deliver higher quality care with greater efficiency,” said JCO Clinical Cancer Informatics editor in chief Debra A. Patt, MD, MPH, MBA, in a 2017 editorial about the launch of the then-new publication. She continued, “With this optimal use of informatics, we will learn from all patients, not just the 3% of adult patients with cancer who accrue to prospective clinical trials.”
CancerLinQ®, a real-world oncology data platform developed by ASCO to rapidly improve patient care and accelerate discovery, helps to bring this future to the present by securely compiling, harmonizing, analyzing, and de-identifying vast amounts of information on patient characteristics (e.g., molecular profiles, comorbidities), treatments, and long-term side effects. By using data from more than 1.6 million patients in near real time, CancerLinQ can identify trends and associations between myriad variables, thereby enabling physicians to generate new hypotheses and apply those conclusions to improve care in real-world settings.
In the past few months, the CancerLinQ LLC team has been busy rolling out new products and onboarding practices to help bring this future to fruition through various extensions of CancerLinQ, the only oncology database developed by oncologists and for oncologists. The platform is among the largest in oncology and reflects the true landscape of cancer care, with ever-increasing diversity in geography, practice settings, patient demographics, and cancer types.
SmartLinQ: Enabling Faster, More Efficient Quality Reporting Through Automation
The CancerLinQ team introduced a new service in June for practices that currently hold Quality Oncology Practice Initiative (QOPI®) Certification: automated quality reporting for re-certification and maintenance submission using CancerLinQ’s SmartLinQ™ quality platform.
QOPI is a voluntary self-assessment and improvement program launched by ASCO in 2006 to help hematology-oncology and medical oncology practices assess the quality of the care they provide to patients. ASCO launched the QOPI Certification Program in January 2010—the first program of its kind for oncology in the United States, and now internationally recognized—to certify that practices have scored above the threshold on key QOPI quality measures and have met chemotherapy safety standards established by ASCO and the Oncology Nursing Society. More than 300 practices currently hold QOPI Certification, and the QOPI Certification Program has certified practices in six countries.
With its complementary mission of facilitating quality improvement, the QOPI Certification Program is a perfect program for CancerLinQ to support through electronic reporting. The SmartLinQ platform helps practices and researchers improve health care operations and transform patient care through quality automation tools, comprehensive data sets, and actionable insights.
“We were delighted when CancerLinQ evolved to the point where they were able to electronically abstract data,” said Barbara L. McAneny, MD, MACP, FASCO, chief executive officer of New Mexico Cancer Center (NMCC) in Albuquerque, which participated in the pilot SmartLinQ program along with three other CancerLinQ practices. “That is a major hurdle that has been overcome.”
For NMCC, participating in the pilot program was an opportunity to pursue QOPI Re-Certification in a new way. The center first achieved QOPI Certification in 2013 as part of its pursuit of clinical and operational excellence. While QOPI Certification lasts 3 years, the manual documentation to recertify each renewal period required nearly 100 hours of staff overtime. Practice leaders desired a more efficient way to pursue QOPI Re-Certification.
CancerLinQ’s ability to electronically extract quality measure data from the electronic record allowed NMCC to automate its QOPI Re-Certification reporting, saving 80 hours of staff overtime. In addition, CancerLinQ helped NMCC’s practice leaders quickly correct data entry errors that may have otherwise been overlooked, and to prioritize areas for staff training for future quality improvement efforts.
“We felt very comfortable with all of the CancerLinQ team, and being able to look at the dashboard and see all of the results in real time was very helpful,” said Marta Bauman, RN, an NMCC nursing manager. “Number one, it saved me a lot of time. Number two, it saved our practice a lot of money. I would absolutely say that it was a game-changer.”
“ASCO has been the leader in quality improvement for more than 20 years, and the range of its programs designed to support practices in this area reflects that,” said CancerLinQ chief executive officer and ASCO executive vice president Cory Wiegert. “With CancerLinQ, we wanted to help build on this success by automating the process and giving practices more time to focus solely on patient care.”
CancerLinQ’s goal as it further develops the SmartLinQ application is for all participating CancerLinQ practices to be able to automate their quality management and reporting, identify real-time opportunities for quality improvement, and benchmark quality performance against the broader CancerLinQ network.
Another practice that piloted electronic reporting was Avera Cancer Institute in Sioux Falls, SD, which enrolled in the program to pursue QOPI Certification for the first time.
“With electronic reporting, things are seamless,” said Luis A. Rojas, MD, FACOG, clinical vice president of cancer services at Avera Cancer Institute. “It’s actually gratifying to the physicians when we see our numbers. We’ve been in quality meetings where physicians are sitting around the table and our quality officer comes in and pulls up CancerLinQ and we’ll see our scores and the number. So that’s a plus for us.”
Capturing Data on COVID-19 Infections in Patients With Cancer
In April, ASCO launched its Survey on COVID-19 in Oncology Registry, which aims to help the cancer community learn about symptom patterns and COVID-19 severity among patients with cancer, as well as how COVID-19 infections impact the delivery of cancer care and patient outcomes.
ASCO’s COVID-19 registry, which is open to all U.S. oncology practices, utilizes the CancerLinQ platform to collect and securely store data. Participating practices complete a baseline data capture form on each patient with cancer who has a confirmed diagnosis of COVID-19, and subsequent follow-up information on status, treatment, and outcomes. Limited patient identifying data—including ZIP code, date of birth, gender, race, ethnicity, type of cancer, and comorbidities—is collected in a secure way to make longitudinal analysis possible.
“Each week our research team reviews our institution’s list of COVID-positive patients to identify patients with cancer to be sure that they are captured for entry into the COVID-19 in Oncology Registry,” said Randall A. Oyer, MD, medical director at the Penn Medicine Lancaster General Health’s Ann B. Barshinger Cancer Institute and a member of ASCO’s COVID-19 Research and Analysis Steering Group. “We have enrolled a number of patients. The case logs are easy to use, and each case entry takes under 15 minutes to complete. Our research and physician teams are excited about participating and helping to build the strong database that will lead to greater understanding. We are also pleased that, as study participants, we will be receiving periodic updates on what has been learned so far. This is important to us.”
The registry will collect both baseline and follow-up data into 2021 with the objective of capturing longitudinal data on how the virus impacts cancer care and outcomes during the COVID-19 pandemic. As data becomes available and is analyzed, ASCO will issue reports on findings, such as characteristics of patients most impacted by COVID-19; estimates of disease severity; treatment modifications or delays; use of telemedicine in the cancer treatment setting; and clinical outcomes among patients related to both COVID-19 and cancer. Ultimately, ASCO staff plan to develop peer-reviewed manuscripts based on the data provided.
“The registry was very easy to open, and the information collected makes sense to us,” Dr. Oyer said. “It will help provide the answers we need to know about how and which patients with cancer are affected, what treatments might work and which might be harmful, and what happens to cancer outcomes as usual cancer care is interrupted.”
Separate from the ASCO COVID-19 registry, practices that have signed on to the CancerLinQ platform already are providing data directly about COVID-19 infections and cancer outcomes among their patients to help improve the quality of care. This data is being automatically collected from the practice electronic health record and added to the practice’s overall CancerLinQ data. The CancerLinQ team is producing dashboards and regular reports related to COVID-19 in CancerLinQ practices, using de-identified data.
Powering Research With CancerLinQ Discovery
In May, CancerLinQ LLC upgraded the CancerLinQ Discovery® platform, making the curated sets of cancer data pulled from CancerLinQ for practitioners and researchers more accessible through an improved and user-friendly interface.
CancerLinQ Discovery, which was rolled out in November 2016, offers de-identified custom data sets derived from the vast pool of real-world electronic health record data contributed by participating practices across the U.S. to the CancerLinQ database. These curated data sets are used by academic researchers, government agencies, major cancer centers, and others in the oncology community to generate practical knowledge that improves cancer care, as well new hypotheses for clinical research.
The new CancerLinQ Discovery Research Platform offers researchers an easy-to-use, self-service program of de-identified real-world cancer care data available for:
- Accessing five disease areas: breast, lung, ovarian, and prostate cancers, as well as chronic lymphocytic leukemia/small lymphocytic lymphoma; more data sets are scheduled to be available later this year
- Reviewing key summary statistics about each data set prior to requesting access
- Accessing a secure AWS workspace for completing analytic work with scalable cloud computing resources like Apache Spark, RStudio, and Jupyter Notebooks, with additional upgrades and tools available upon request
The platform also offers streamlined data request, approval, and access processes.
Judith A. Paice, PhD, RN, director of the Cancer Pain Program in the Hematology-Oncology Division at Northwestern University’s Feinberg School of Medicine, has been using CancerLinQ Discovery data to study patients with metastatic non-small cell lung cancer. “We know that this patient population often has pain associated with their advanced disease, so we are focusing on oncologists’ prescriptions of opioids for pain management,” Dr. Paice said. “We are looking at how opioid prescribing patterns change over time and whether they differ by patients’ clinical characteristics, such as age, race and ethnicity, presence and types of metastases, pain scores, etc.”
Because the CancerLinQ Discovery data set distinguishes which type of opioid was described, Dr. Paice and her group can see the changes in use for each type over their study period of 2010 to 2017. “We analyzed the data in light of what we know about changes to federal and state regulations and payer and institutional policies regarding opioid prescribing,” she said. “We saw the expected overall decrease in opioid prescribing and shifts between different drugs based on Drug Enforcement Administration schedule changes.”
De-identified data from CancerLinQ Discovery powered 11 abstracts accepted and presented at the ASCO20 Virtual Scientific Program, including:
- An examination of the concordance of death dates from CancerLinQ Discovery and an external source, as well as the effect of incorporating external death data on estimates of overall survival (Abstract 7064).
- A retrospective, observational analysis of opioid use among patients with metastatic non-small cell lung cancer, finding that overall opioid prescriptions had declined since 2012 with a slight uptick in lower-potency ones (Abstract 2076).
- A review of ECOG performance status scores derived using machine learning techniques developed using real-world data from CancerLinQ Discovery (Abstract e19318).
- A study of the demographics and practice patterns of real-world immune checkpoint inhibitor usage in older patients with lung cancer (Abstract e24016).
- A validation of a natural language processing algorithm built using a CancerLinQ Discovery data set of patients with breast cancer to capture metastatic status, which is often in physician notes rather than structured data in an electronic medical record (Abstract e14056).
- A review of an artificial intelligence model that was developed from CancerLinQ Discovery data to dynamically predict recurrence of metastatic breast cancer (Abstract e13078).
- An analysis of the characteristics of genetic testing for genetic mutations related to ovarian cancer (Abstract 6041).
“CancerLinQ Discovery was created as tool for practitioners and researchers to allow them to learn from the experiences of patients throughout the CancerLinQ network,” said CancerLinQ medical director Robert S. Miller, MD, FACP, FASCO. “The data is pulled from a diverse network of rural, community, academic, and health system care settings, and this inclusiveness enables research results to be generalizable to large segments of the patient population.”
CancerLinQ Discovery data “provides rich information for exploring research questions, but requires understanding the clinical context of data that will be visible from an oncology practice’s electronic health records and transparency about assumptions applied to and limitations of analysis and interpretation,” Dr. Paice said. “It is vital to be aware that the lack of data does not translate to the absence of the clinical characteristic or circumstance.”
Other abstracts accepted to the ASCO20 Virtual Scientific Program that used CancerLinQ Discovery data included studies on:
- Creating a machine learning model to predict potential cardiovascular events in patients with advanced cancer who received PD-(L)1 therapy (Abstract e15075).
- Developing an artificial intelligence model to predict slow progression for advanced non-small cell lung cancer (Abstract e21596).
- Identifying transgender people with cancer in electronic health records to study disparities in care (Abstract e19046).
Furthering Interoperability With mCODE
Version 1.0 of mCODE, the Minimal Common Oncology Data Elements standard, was approved by Health Level Seven International (HL7) as a Standard for Trial Use (STU) in March. The mCODE elements, which are available for use free of charge, are a core set of structured data elements to facilitate cancer data interoperability. They were developed in a collaboration between ASCO, CancerLinQ LLC, the MITRE Corporation, the Alliance for Clinical Trials in Oncology Foundation, and the American Society for Radiation Oncology.
The elements have been designed using standard, widely available medical terminologies to enable searchability. If oncology care team members search different EHRs that are all using mCODE, they should find similar results—the data can be found, accessed, and made available. The data elements are grouped into six main areas: patient, disease, genomics, labs/vitals, treatment, and outcomes. Some of the elements are being used in CancerLinQ, as well as ASCO’s Survey on COVID-19 in Oncology Registry. ASCO and CancerLinQ LLC continue to support interoperability and quality of data as espoused by mCODE and are working to include all mCODE elements within CancerLinQ.
Earlier this year, the mCODE Council was established with 20 members from academic, government, industry, and nonprofit organizations. The council provides thought leadership, advises on new and amended data elements, suggests and advises on use cases, and promotes the initiative. mCODE staff are actively seeking additional members for the council; interested parties can fill out a brief application at mcodeinitiative.org to be considered for membership.
Putting CancerLinQ’s Promise Into Practice
ASCO, its members, and the greater oncology community have long known that transformational improvement in patient outcomes can only be fully realized through an oncology rapid learning health system, and this was the driving force behind CancerLinQ’s creation. Seven years after the development of CancerLinQ first began, the platform is coming into its own and continuing to expand on its promise as a hub where routine patient care data seamlessly informs scientific discovery, and, reciprocally, research informs practice.
As Dr. Patt noted in her introductory editorial for JCO Clinical Cancer Informatics, “It’s a thrilling time to be in this field, since its influence and potential are increasing exponentially as biomedical informatics becomes more integrated into all aspects of health care. Understanding this growing wealth of informatics capabilities and applying them to patient care in oncology is something that will help us all improve oncology care.”
Acknowledgment: Aaron Tallent contributed reporting on SmartLinQ.