This is the End, Beautiful Friend

So, about the title… No, it’s not the end of the xPRESS; this is only our second installment! Quitting before beginning? Never! Okay, so then what’s up with that title? Well, we’re talking about “the end” this month.?

The end of what??

Last month, I spoke about how we’re now in the middle innings of the real-world data and real-world evidence ballgame. If the middle innings look like this, what will “the end” look like??

Stakeholders may all have varying views on what that state may look like, but at xCures, for instance, we have a vision.?

Imagine a world where your dataset has no holes, gaps, or continuity errors. Where you know that each patient in the dataset has a completely longitudinal journey documented from their diagnosis (ideally, pre-diagnosis), until the present day (with verified dates of death, where relevant). Just imagine!

Imagine a world where your dataset has data from every single medical encounter on each patient’s care journey. All providers. All sites. Inpatient and outpatient. Rural, community, and academic settings. Imaging centers, radiology reports, raw DICOMs. Pathology reports and data from commercial NGS labs. Pharmacy data on prescriptions filled. All comorbid conditions and concomitant medications. The full 360-degree view of the patient’s care journey. Just imagine!

But in the current US healthcare environment, as fragmented as it is, we’re currently left with cross-sections of patient journeys, rather than a fully longitudinal one, due to the nature of the legal framework that enables RWD companies to access data from only very specific providers, health systems, and/or EMRs. And since there are so many scenarios where individual patient data is disparately spread amongst multiple sites, providers, EHRs, it leaves us with limited solutions.?

One solution taking shape is tokenization, though early results are suggesting that tokenization to create fully longitudinal journeys is roughly only successful 10% of the time.?

At xCures, we believe there is one paramount solution, and that is to form a relationship directly with the patient, rather than with individual sites, providers, health systems, EMR vendors, etc. By partnering with the patient, we can get fully longitudinal journeys from directly consented patients. To us, it’ll be as simple as that, in the end.?

Speaking of the end…

When working with a large real-world dataset, how does one know when an individual patient has reached the end, i.e., final outcome, of their care journey?

When working with the current legacy RWD sets, it is not uncommon to be uncertain about dates of death. Nobody wants to think about patients passing away, but in serious diseases like cancer, it is inevitable for many. And being able to understand the impact of a product on overall survival (OS) is critically important. Ultimately, mortality/survival is the most important outcome for all stakeholders involved, but especially patients. Thus, understanding mortality at the individual patient level is essential.?

The current RWD paradigm makes it near impossible to determine overall survival in a real-world treatment setting. The data is already de-identified. Patients constantly switch providers, health systems & sites. Patients move. Patients are “lost to follow up,” as we say in the RWD biz.?

Unless death is explicitly stated in the medical record, which, frankly, doesn’t happen too often, we’re left wondering, “what happened to this patient?” The way we deal with it now, such as Kaplan-Meier Curves, are necessarily evils, until we can solve the problem of understanding the full longitudinal journey of each patient in a dataset.?

But imagine an end-state where we’ve partnered with the patient. They’ve provided their informed consent. We know who they are before de-identification. When we know who the patient is, we can cross reference against the National Death Index, published obituaries, and most excitingly, identify verification services like LexisNexis, who – like Santa – know when you are sleeping, know when you’re awake, and they also know when you die.?

When you’ve partnered with the patient, and you can use identify verification to track them prospectively as they flow through the healthcare system, and you know if, when, and where they pass away, it makes analyses for real-world Overall Survival much more accurate and brings an entirely new dimension to leveraging real-world data.??

Seem farfetched? Hardly! It is innovations like these that we have brought to xCures. To us, the end state does not seem far away. In fact, we have brought this model to fruition for more than 50,000 patients and counting. We have the full 360-degree view of these patients, and via identify verification, we will be able to follow them prospectively until final outcome, no matter where they go.

Quite the end to the story, huh? Far from it, friend. In fact, it’s just the beginning.?

Max??

xPRESS Volume 2: The End (Spotify Playlist)