Using Real-World Data to Improve Cancer Care Delivery, Precision Medicine

THE ADVANTAGES OF REAL-WORLD DATA

Real-world data is particularly useful for researchers like Rocque, as it provides some advantages that other types of data don’t. Rocque’s research is heavily focused quality of care delivery, which is assessed in many clinical trials. But these assessments, through which data insights are gleaned, can be impacted by patient demographics.

In a recent project, Rocque and her team utilized real-world data to understand outcomes for those typically underrepresented or not represented in clinical trials, such as individuals at both ends of the age spectrum and people of color. Researchers also looked at outcomes for patients whose abnormal lab results or diagnosis of a second cancer would typically exclude them from clinical trials.

Rocque noted that though these patients are not well represented in clinical trials, clinicians see them often in the clinical setting.

“So, we want to have information on what to tell our patients about expected outcomes and how well different medications work. [In the study], one of the key findings was that for those patients that are excluded typically from trials based on their abnormal lab results… they really had inferior survival outcomes,” Rocque stated. “That really highlights the need to do this kind of work using real-world data for populations that aren't included, and to start to think about ‘What are the drivers of those differential outcomes?’ and ‘How can we provide better care to our patient population across the board?’ [instead of] just those patients who are physically fit, oftentimes younger, and white.”

Because clinical trial data is unavailable for these patients, real-world data is one of the only ways to gain clinically useful insights. Many oncology data platforms, such as CancerLinQ which Rocque used for this study, utilize EHRs to generate real-world datasets. This can also give researchers access to more comprehensive information about patients.

In Rocque’s study, lab values were key to determining why some patients were excluded from clinical trials. When researchers must rely on other types of data, such as information pulled from a claims database, rather than real-world data, some crucial information, such as lab results, are often limited. For example, researchers may have access to diagnosis codes in claims data, but not lab values indicating severity of disease. In this way, real-world data allows researchers to look deeper into clinical variables, such as severity of disease and comorbidities, that could significantly improve research quality.

Real-world data also helps researchers who are interested in less common conditions that aren’t as likely to be studied in clinical trials. With access to thousands of patients’ data, lack of clinical trials becomes less of a barrier for researchers interested in rare diseases.

REAL-WORLD DATA AND PERSONALIZED CANCER CARE

Personalized cancer care is becoming more popular as innovations in genomics continue to develop.

Patients with particular biomarkers or genetic patterns for certain types of disease make up a smaller subset of cancer patients. Because of this real-world data can be used in a way similar to how researchers use it to evaluate patients who are underrepresented in trials.

In both cases, real-world data allows researchers to examine these subsets of patients on a larger scale by aggregating their data, rather than leaving researchers to rely on information from clinical trials that may or may not have occurred.

“I think being able to do some of that more complex analysis will help guide us towards what is the correct treatment for individuals with particular patterns, from a genetic perspective,” Rocque noted. “In addition, that may lead you to then test this formally, and should lead you to test this formally, in randomized control trials. So, sometimes what we see from this big data-based analysis is actually a signal that gives us a direction to say, ‘We should be developing studies to test that particular drug in this particular setting because we saw that signal in the real-world data.’”

By analyzing these subsets of patients, researchers can dive deeper into how these patients experienced care and what kind of treatment works for them, which can be used to improve cancer care delivery, Rocque stated.

DATA HARMONIZATION AND PATIENT-REPORTED DATA

Real-world data is extremely useful for researchers as it increases the volume of high-quality patient data, which is key for generalizability. More generalizable findings are desirable in clinical settings because those findings can be used to treat various patient populations, which can lead to better health outcomes and help providers address issues like health inequity.

However, real-world data is not without limits. According to Rocque, data harmonization is a crucial factor for health systems to consider when capturing and sharing data. Data harmonization refers to the effort to combine data from disparate sources and integrate it into a more unified, easy-to-understand format. It is similar to interoperability, but is more specifically concerned with standardizing how data is captured and shared to ensure data accuracy.

In Rocque’s view, data harmonization and interoperability across health systems will help to achieve more accurate data use in clinical research. But she also highlighted the importance of patient-reported data.

“We are in an era where patient-reported data is becoming increasingly important,” Rocque stated. “So, one of the places where I think we also have a huge opportunity is thinking about how we aggregate that patient-reported data, patient-reported outcomes, patient-reported preferences, [and] all of the different things that come directly from patients. There is an increasing movement toward asking patients for direct feedback, and I think that's going to be another important real-world data source as we move forward.”