It’s Time for a Fresh Approach to Patient Data in Rare Disease

Asaf Evenhaim

Specialized therapies for smaller, highly targeted patient populations are fueling most growth in life sciences. This type of pipeline calls for new methods of finding and engaging the right patients and their healthcare providers (HCPs). To serve this growing patient population, companies need data-driven strategies that allow for more intentional patient targeting and a deeper understanding of the HCPs they visit.

Yet, with no meaningful innovation for nearly two decades, traditional patient data approaches aren’t meeting today’s commercial needs. Because of disconnected data sets and access limitations, companies struggle to fill the data gaps so they can truly understand rare disease patients’ path to diagnosis—and help to speed critical treatment.

Sales and marketing organizations need more comprehensive, actionable—but of course, still anonymous—patient data to quickly extract insights that inform high-impact commercial initiatives. To reach this nirvana, modern data teams should take a fresh look at their go-to-market approach with these four strategies: increase visibility across the full patient journey, include diverse data sources, leverage timely data for multiple use cases, and ensure efficient and accessible analysis.

1. Longitudinally complete visibility

In today’s highly specialized world, over-reliance on retail pharmacy-dispensed prescription data is no longer sufficient. To address more complex health conditions, data should include information about patient diagnoses, procedures, and office visits as well. This is especially important with rare diseases, since patient journeys are incredibly complex, and people often go undiagnosed (or misdiagnosed) for long periods of time. They may experience dozens of tests, scans, and prescriptions on their path to treatment.

To successfully commercialize specialty therapies with a patient-centric approach, data management teams require access to large-scale health data from multiple sources. This should include a longitudinal health history of a patient’s related—or even seemingly unrelated—medical activity, along with data about their treating and diagnosing specialists, referring healthcare providers, and treatment patterns, as well as elapsed time between interventions.

However, tracking a patient’s journey can often be challenging since patient tokens (an individual’s identity marker used for linking anonymized data sets) might change over time. In 2020, nearly 30 million people moved¹ and the majority who do so typically relocate to a different zip code.² Furthermore, one in four medical records contain an error in last name, first name, date of birth, or gender.³ Each of these data points is critical for creating an anonymized patient token to link information sources. Without a robust identity management strategy, companies end up with inaccurate or siloed data that makes it impossible to reveal the entire patient history surrounding rare diseases.

Once data is anonymized, critical identity information about that individual is lost. To solve these challenges while fully meeting privacy restrictions, teams need solutions built specifically to address them before de-identification. The most effective platforms track all versions of a person’s identity over time and use artificial intelligence (AI) to autonomously link and de-link data on its own—essentially providing self-driving data.

When a patient’s identity is properly managed throughout their healthcare journey, analytics teams can uncover risk markers, predictive indicators, and challenges to getting patients on a specialized therapy. They might also delve into which doctors tend to diagnose or misdiagnose patients with a particular condition, and which HCPs have innovative adoption profiles for new lines of treatment, or which practices might benefit from more education.

2. Diverse data sources that close key gaps

For too long, life sciences organizations have had to deal with large gaps in the patient data they license. This stems both from the complexity of the U.S. healthcare system, as well as from the antiquated data sourcing model still used by some data brokers. This often leads to incomplete insights or the inability to perform comprehensive analyses.

“Pharma has some of the best commercial data of any industry, however, typical data sets have several limits,” shared Eric Solis, director and lead data scientist at Takeda. Take, for example, products distributed solely through specialty pharmacies. Data about these products are usually “blocked” by the manufacturer, meaning that patient insights cannot be licensed to anyone. And since traditional data brokers only source patient data from pharmacies, companies have had a hard time tracking their sales.

This is where diversity plays a key role, particularly for marketing more specialized treatments. By expanding beyond pharmacy data to accurately combine patient data from switches, pharmacy benefit managers (PBMs), health plans, and other sources, commercial teams can gain visibility into previous blind spots. Compiling information from diverse sources and harmonizing it is challenging but worthwhile. A reimagined data strategy should involve continually adding information on prescription and medical benefit claims from among several under-utilized sources.

3. Timely data that’s ‘fit for use’

Flexible data strategies should also support a broad range of commercial use cases with anonymized patient data. To ensure agility, data management teams need to blend tailored, holistic data sets. They should look well beyond pre-determined patient groupings—also known as ‘market baskets’—that are based only on diagnostic codes, products, or procedures.

When teams are empowered to easily add in data from other categories, they are more likely to identify the misdiagnosed or under-diagnosed patient populations that are all-too common in the rare disease landscape. After all, the promise of machine learning and AI is its ability to identify unexpected predictive patterns in large data sets. That cannot be achieved if patient data is filtered down to just one pre-assumed market filter. As Solis noted, “To develop truly novel analytics, we need a lot more insight on patients and their healthcare journeys that goes beyond the myopic view of individual markets.”

By integrating various data sets, companies can rapidly adapt amidst shifting market dynamics and better address unanticipated questions that arise from stakeholders across the organization. For instance, when a data platform provides daily updates with incoming medical claims, unexpected insights often emerge. Especially in the early days of a new therapy’s launch, timely data helps pharmaceutical companies quickly identify the right specialists and other HCP targets.

4. More efficient and accessible analysis

Finally, data teams need ways to better collaborate with a variety of stakeholders, so they can contribute to more commercial initiatives. With this goal in mind, they should aim to license data via software designed from the ground up to be accessible for a variety of user types—from data scientists and analysts to business and IT users.

Continuous access to fresh data also helps analysts spot changes or answer new questions about niche patient populations. Given this, teams can benefit from solutions that simplify the process of scheduling recurring jobs, conducting ad hoc data exploration, and selecting new data pools when needed.

As life sciences businesses double down on data-driven operations, many turn to outside analytical experts to supplement their internal teams. They should ensure their licensed data is shareable without artificial limits or weeks of waiting for third-party access. With this kind of flexibility, data managers can greatly increase their analytical power.

Rethinking data-driven commercialization

“In order to tackle critical business questions we couldn't answer previously, the industry needs better solutions for patient data,” said Solis. “A fresh approach can go a long way in helping data science teams like ours tame the complexity of rich, unstructured data sets and deliver simple, easily-digestible answers to various stakeholders.”

Robust data informs more effective launch strategies through better market landscape analysis, customer profiling and segmentation, and tracking. Teams can map out the most typical journeys to diagnosis for rare disease patients. For example, they might notice comorbidities that signal a potential for earlier diagnosis and intervention, or pinpoint common specialty visits and care team dynamics.

Furthermore, efficient data access is a key enabler for the digital transformation that was jump-started overnight when COVID-19 hit last year. Life sciences reps are increasingly using digital channels as part of integrated omnichannel strategies to meet HCPs wherever they are.⁴ When reps know which doctors high-risk patients are visiting and what those HCPs tend to prescribe, they can provide information that is most relevant for them and their patients.

Innovative life sciences companies are thinking differently about patient-level information and finding better ways to support commercial operations, strategic planning, AI, and go-to-market execution. By doing so, they’re able to find high-risk patient populations sooner and help improve rare disease outcomes.

References

1. U.S. Census Bureau, December 2020.
2. “US Moving Statistics for 2019,” Moving.com, October 2019.
3. “How to Maintain a Clean Master Patient Index,” IronMountain.com
4. “Opening Digital Doors: Remote Selling During COVID-19,” Pharmaceutical Executive, December 2020.

Asaf Evenhaim, CEO, Veeva Crossix