Summary

Real-world data (RWD) – information on patient care and outcomes collected outside of clinical trials – are an important source of information within oncology. One source of RWD is an electronic health record (EHR), which contains a person’s demographic characteristics and medical history. Collecting and merging information from EHRs can be difficult due to incompatible EHR systems and inconsistent data reporting across oncology practices.1 The American Society of Clinical Oncology (ASCO) introduced the Cancer Learning Intelligence Network for Quality (CancerLinQ®) to overcome the barriers associated with collecting data from EHRs. CancerLinQ® collects and summarises RWD directly from EHRs and other sources in real time (i.e. a rapid learning health system). It is a subscription-based platform that functions as a quality-monitoring tool for clinical practice and a research database.2

 

Problem

Only 3% of adults with cancer in the United States (US) are treated in clinical trials. Trial participants tend to be younger, fitter and less ethnically diverse than the general population with cancer.3-5 Therefore, guidelines and recommendations based solely on trial findings may misrepresent the needs of the remaining 97% of adults with cancer1 – as well as those of children, who are usually excluded from clinical trials. Real-world data (RWD) – information on patient care and outcomes collected outside of clinical trials – offer complementary information to create a more representative overview of people with cancer in the US.

One potential source of RWD is an electronic health record (EHR), which contains a person’s demographic characteristics and medical history (e.g. medication/treatments, diagnoses, imaging and laboratory test results, and consultation notes).6 However, collecting EHR data is difficult; EHR systems are often incompatible, practices may report different outcomes or use different terminologies, and EHRs contain identifiable sensitive information that cannot be shared.1 These barriers make it difficult for healthcare professionals to maintain an updated overview of oncology beyond their own practices, thereby hindering clinical progress and multidisciplinary collaboration.7 8

Solution

The Cancer Learning Intelligence Network for Quality (CancerLinQ®) was introduced by the American Society of Clinical Oncology (ASCO) in 2015 to facilitate the use of RWD in clinical practice and research. It now runs as a non-profit wholly owned subsidiary of ASCO (CancerLinQ® LLC).7

CancerLinQ® is a rapid learning health system that summarises large amounts of RWD in real time. It collects both structured and unstructured data – such as diagnosis codes and consultation notes – through an automatic daily feed from EHRs and other sources (e.g. tumour and death registries) without the need for additional input from healthcare professionals. All subscribing practices (i.e. practices that upload their data to CancerLinQ®) can access their own identifiable data through the online platform. Additionally, aggregated de-identified data are made available to all subscribing practices and to researchers.1 2 7

Healthcare professionals can use CancerLinQ® as a tool to:

  • inform clinical decisions (e.g. matching patient characteristics with existing clinical guidelines and data from similar patients)
  • identify areas for improvement in their practice (e.g. comparing practice data with existing clinical guidelines)
  • fulfil mandatory data reporting requirements at the federal, state and quality programme levels
  • visualise and analyse patient data to identify trends in specific patient populations.1 9

Online access to these tools ensures that real-time care information reaches healthcare professionals directly, with the aim of giving people with cancer up-to-date, high-quality cancer care regardless of where they live or receive treatment.9

 

What has it achieved?

As of 2019, the CancerLinQ® platform includes data from over 1.5 million patients (~1.05 million with a primary malignant tumour diagnosis), representing more than 100 oncology practices across the US.10 It has launched several initiatives to improve the use of RWD in oncology research and clinical practice.

 

CancerLinQ Discovery®

CancerLinQ Discovery®, the research database of the CancerLinQ® platform, was introduced to increase RWD usage in oncology research.2 Fit-for-purpose de-identified datasets can be requested by:

  • members of the research community (e.g. academic researchers or public health agencies)
  • life sciences companies and other commercial entities (through Tempus and Concerto HealthAI).10

CancerLinQ Discovery® also has an ongoing partnership with the US Food and Drug Administration (FDA) to monitor the performance of new cancer treatments in practice using RWD.1 11

 

The mCODE initiative

The Minimal Common Oncology Data Elements (mCODE) initiative aims to improve the compatibility, quality and consistency of EHR data to facilitate multidisciplinary research and collaboration. The initiative published a standardised set of important EHR data elements and an open-source data model to connect different EHR systems. These tools are publicly available via the mCODE website.12

 

Partnership with the National Cancer Institute (NCI)

CancerLinQ® launched a partnership with the Surveillance Epidemiology and End Results (SEER) programme at the National Cancer Institute (NCI) in 2017 to bring population-level cancer data directly to oncologists and to improve cancer surveillance through the use of RWD.13 This bidirectional data-flow approach, currently being piloted across the state of Utah, has two components:

  • SEER data will be incorporated into the CancerLinQ® platform to improve quality monitoring and clinical decision-making. Practices will be able to compare their data against regional and national trends using the SEER data.
  • Practices will be able to upload their data directly to the SEER programme through the CancerLinQ® portal10 13

 

Next steps

CancerLinQ® will continue recruiting a range of oncology practices (e.g. community-based and academic centres) to provide an even more representative overview of real-world cancer care. The database will continue to grow through improved extraction processes for both structured and unstructured data, and the inclusion of data sources such as financial and administrative data, claims data, genomic/molecular data and drug inventory data.8 10

 

Further information

  • The CancerLinQ® website outlining its uses in research and clinical practice
  • The CancerLinQ Discovery® website, including a form to submit a data access request
  • The mCODETM initiative website
  • The SEER programme website
  • CancerLinQ® certification for EHR systems
  • An example of how CancerLinQ Discovery® data can be used to monitor new treatment approaches (abstract from the 2019 ASCO Annual Meeting)

References:

  1. Shah A, Stewart AK, Kolacevski A, et al. 2016. Building a rapid learning health care system for oncology: why CancerLinQ collects identifiable health information to achieve its vision. Journal of Clinical Oncology 34(7): 756-63
  2. Miller RS, Wong JL. 2017. Using oncology real-world evidence for quality improvement and discovery: the case for ASCO's CancerLinQ. Future Oncology 14(1): 5-8
  3. Lewis JH, Kilgore ML, Goldman DP, et al. 2003. Participation of patients 65 years of age or older in cancer clinical trials. Journal of Clinical Oncology 21(7): 1383-9
  4. Murthy VH, Krumholz HM, Gross CP. 2004. Participation in cancer clinical trials: race-, sex-, and age-based disparities. JAMA 291(22): 2720-26
  5. Al-Refaie WB, Vickers SM, Zhong W, et al. 2011. Cancer trials versus the real world in the United States. Annals of Surgery 254(3): 438-43
  6. Centers for Medicare and Medicaid Services. 2012. Electronic health records. Available here: https://www.cms.gov/medicare/e-health/ehealthrecords/index.html [accessed: June 2019]
  7. Rubinstein SM, Warner JL. 2018. CancerLinQ: origins, implementation, and future directions. JCO Clinical Cancer Informatics 2(1): 1-7
  8. Miller RS. 2016. CancerLinQ update. Journal of Oncology Practice 12(10): 835-37
  9. CancerLinQ®. Practice tools. Available here: https://cancerlinq.org/practice-tools [accessed: June 2019]
  10. Miller RS. 2019. Interview with Marissa Mes at The Health Policy Partnership [telephone]. 18/06/19
  11. Tallent A. 2017. CancerLinQ partners with FDA to study real-world use of newly approved cancer treatments [online]. ASCO News Releases. Available here: https://www.asco.org/about-asco/press-center/news-releases/cancerlinq-partners-fda-study-real-world-use-newly-approved [accessed: June 2019]
  12. American Society of Clinical Oncology. mCODETM: Minimal Common Oncology Data Elements. Available here: https://mcodeinitiative.org/ [accessed: June 2019]
  13. Tallent A. 2017. CancerLinQ and National Cancer Institute announce partnership to enhanced patient care and strengthen cancer surveillance. ASCO News Releases. Available here: https://www.asco.org/about-asco/press-center/news-releases/cancerlinq-and-national-cancer-institute-announce-partnership [accessed: June 2019]