Analytics in payer-oncologist partnerships

Reducing unwanted care variation is critical to driving quality cancer care while managing costs. But as the number of available oncology treatments has exploded, research has identified a wide variation in treatment provided. One large study, for example, found that only about half of patients with advanced lung cancer and a specific mutation received the appropriate treatment targeted at that mutation.³

For every cancer patient who doesn't receive an appropriate targeted treatment, others receive unnecessary treatment that fails to improve their outcomes. In fact, a recent systematic review identified overuse of 154 cancer-related services, including imaging, procedures and therapeutics.⁴ This increases cancer patients' risk of financial toxicity or hardship, which can severely compromise their quality of life.⁵

The oncology field needs a comprehensive strategy to synthesize evidence from multiple sources, determine which interventions are most associated with favorable outcomes, and generally improve the care experience and clinical outcomes.

Payers can offer this system by creating point-of-care decision support tools and evidence-based data analytics services that guide providers to choose high-value, clinically appropriate therapies and supportive care. These services can promote everything from proven clinical pathways to correct dosage amounts, while reducing the documented variation that exists in treatments for common cancers such as prostate⁶ and breast.⁷

Clinical trials have traditionally offered providers the data needed to guide oncology treatment. However, in recent years, researchers and physicians have increasingly embraced real-world data (RWD) to unlock additional insights beyond what is available from a controlled sample population. This data helps researchers and providers better understand a new therapy's safety, efficacy and performance relative to other options.

Payers sit in a unique position when it comes to RWD: They have not only cost and claims data for hundreds of thousands of cancer patients, but also de-identified EHR data linked to medical and pharmacy claims. This data captures information across a long period of time, from patients' pre-cancer clinical appointments to their diagnostics, labs, post-surgical care and treatment encounters. This trove of information offers the breadth and depth providers need to understand treatment efficacy, dosing complexities, short- and long-term outcomes, complications and whether a specific regimen is appropriate for individual patients.

RWD offers great promise, but harnessing additional, extremely granular elements can improve its relevance to precision oncology. Specifically, deeper details — such as a patient's cancer stage, grade, histology, genetic mutations, other blood biomarkers, lines of therapy and their response to treatments — are invaluable to providers navigating clinically specific treatment options.

Collecting patient-reported outcomes through electronic capture tools and case manager discussions can help monitor a patient in between visits and reinforce the provider’s care plan. And by actively monitoring through electronic patient-reported outcomes (ePROs) and evaluating the latest changes in oncology trends, payers can support a highly adaptable and configurable benefit management solution for an ever-evolving oncology landscape.

With precision oncology as a core driver of therapy decisions, payers can ensure an automated prior authorization platform delivers patient, indication and tumor-specific therapy recommendations and prior authorization determinations that incorporate the rapid changes in literature that reflect oncology practice today.

The statistics are sobering: Each U.S. resident faces roughly a 40% chance of developing some form of cancer in their lifetime.⁸ The disease kills more than 600,000 Americans each year and another 1.9 million are diagnosed annually.⁹

Diagnosing cancer at an early stage has the greatest impact on improving overall survival. It provides time to find appropriate, personalized treatments and allows doctors to stop the disease before it can spread. With cancer treatment costs being nearly 3 times greater at a later stage versus an earlier one,¹⁰ early detection can also help ease these overwhelming costs. Unfortunately, many forms of cancer have no recommended early-detection screening.

Using this valuable oncology data, payers can partner with providers to create and deploy powerful prediction models that can potentially identify patients with cancer sooner, when it's easier and less expensive to treat. These models can combine EHR data with other imaging or biomarker screening tools to target high-risk patient populations to accurately identify early-stage patients.

In one study using de-identified payer EHR data, researchers created a prediction model for pancreatic cancer based on nearly 30,000 cases of the disease over 8 years. The model included more than 18,000 features, from diagnoses and procedures to clinical notes and medications, that could detect 58% of late-stage pancreatic cancers a median of 24 months before their actual diagnosis.¹¹

These models can make a real difference in patients' lives. Payers can use them to drive real-world interventions, like proactively sending at-home colon cancer screening kits or assisting in helping individuals access care. Predictive models can also improve quality of care for existing cancer patients by helping providers better understand their risk for certain outcomes and tailor their care accordingly with individualized treatment plans.

Access to information is the foundation for more effective oncology care. By bridging gaps in the system and ensuring data can be leveraged to deliver on this promise, payers can collaborate with providers to usher in a new era of effective precision cancer treatment.