Addressing Patient Dispersal: Challenging the Hierarchy to Solve the Rare Disease RWD Puzzle

In my last blog, I outlined the significant challenges in generating real-world data (RWD) for rare diseases, introducing what I termed the "data puzzle." A key piece of this puzzle is patient dispersal—a factor that complicates efforts to collect the data necessary for advancing our understanding and treatment of these conditions. Today, I’ll delve deeper into this issue, exploring why patient dispersal is such a critical challenge in rare disease RWD generation and how it impacts the development of effective treatments.

Understanding Patient Dispersal

Although over 7,000 rare diseases were identified globally to date, the number of individuals affected by each condition is minimal. This rarity means that patients are not only few in number but also dispersed across vast geographic areas—sometimes across entire countries or continents. This dispersion poses substantial obstacles to both patient care and data collection.

Imagine a patient diagnosed with a rare condition like cystinosis, a metabolic disorder that requires highly specialized care. The patient might live in a small town, where primary care is available but lacks the expertise for such a complex condition. To manage their illness, the patient may need to travel to different healthcare settings: primary care for routine check-ups, a specialist hospital in a nearby city for regular treatments, and possible a major urban center in another country for advanced care. Each of these healthcare settings may be geographically separated, adding layers of complexity to the patient’s journey. The travel burden on patients cannot be overlooked. It disrupts their lives and limits their ability to participate in clinical studies or contribute to long-term data collection efforts.

More importantly, patient dispersal in rare diseases will lead to significant variability in outcomes due to inconsistent access to specialized care, variability in treatment approaches, challenges in coordinating multidisciplinary care, and delays in diagnosis. In addition to potentially varying levels of quality of care, the difficulty in coordinating care across dispersed healthcare systems can lead to fragmented treatment plans and poorer health outcomes. Consequently, RWD are fragmented across various institutions, each with its own record-keeping systems and standards. Integrating these disparate data sources into a coherent and comprehensive dataset that accurately reflects the patient’s health status is no small task.

Patient dispersal is therefore not just about physical distance; it’s about the fragmentation of care and data that comes with it. This dispersal forces us to confront the reality that without innovative approaches to data collection and integration, the data puzzle of rare diseases will remain incomplete.

The Impact on RWD Generation

Differences in Treatment Approaches and Outcomes

Patient dispersal leads to significant variability in treatment approaches and the outcomes measured. Clinicians will also be spread across diverse geographies, each within different healthcare systems, using varied approaches to treatment and patient care. This disparity can be due to differences in local medical guidelines, available healthcare resources, or the specific expertise of the healthcare providers.

For instance, in treating a rare metabolic disorder like Fabry disease, clinicians in one country might have access to advanced enzyme replacement therapies, while those in another region may rely on symptomatic treatment due to resource constraints. These differences lead to varying patient outcomes.? This variability complicates efforts to compare data across different populations; inconsistent treatment methods and outcome measures make it challenging to draw meaningful conclusions about the efficacy of treatments or the natural progression of the disease. ?

Variability and Delays in Data Collection

The dispersal of patients leads to variability in how data are collected. Different regions and healthcare systems may use different standards and protocols for recording patient data. In one country, a patient's medical history might be meticulously documented in an electronic health record (EHR), while in another, records might be kept in less accessible paper files.? For instance, in a study on a rare autoimmune disease like systemic sclerosis, data from patients in urban centres might be digital and detailed, while those from rural areas might be sparse and fragmented. This variability complicates the creation of a consistent dataset, hindering comprehensive analysis and the ability to draw meaningful conclusions.

Finally, due to basic logistics of traditional methods of data generation, patient dispersal will delay data collection. ?If a registry protocol mandates in-person visits to specialized centres, recruitment for such studies will inevitably be slower, and the risk of patient attrition significantly higher.

Expanding the Solution: Rethinking the Hierarchy

For drug developers planning to generate RWD in rare diseases to support regulatory discussions, overcoming the challenges posed by patient dispersal is no longer optional. Recent FDA guidance on the need to provide common data elements and formats (CODEFs) when using RWD makes it a necessity. But how do we tackle this challenge?

International collaboration leading to standardization is a necessary step. Initiatives like the European Reference Networks (ERNs) exemplify how cross-border collaboration can help address rare diseases. However, internationalization and standardization alone are not sufficient. Decentralization of data generation must follow. Virtual and decentralized observational studies are becoming more common, and super-sites—hubs that collect data from dispersed patients—are already the norm. And we cannot stop there. We must include patient-generated health data, information from local health networks and clinics, and logistical arrangements supporting localized and convenient data generation, such as mobile clinics.

This democratization of data generation will inevitably bring with it concerns about data quality and sources of bias. If we adhere strictly to traditional hierarchies of evidence, we risk dismissing valuable data that could provide critical insights into rare diseases. This is the crux of our next "a-ha" moment in solving the data puzzle: to minimize the challenge of patient dispersal in rare diseases, we must discard the rigid notion of a hierarchy of data sources. We need to be open to evaluating all data sources on their merits, without preconceived notions about their quality.

Drawing Parallels: The Evolution of Quality in Journalism

This shift in thinking parallels a transformation seen in journalism. Traditionally, the hierarchy of sources in journalism placed government reports, academic studies, and official press releases at the top, with social media posts and citizen journalism at the bottom, often disregarded. However, the advent of digital media and the rise of social media platforms have forced journalists and news organizations to reassess this hierarchy. In many instances, the new data sources have provided crucial, on-the-ground insights during breaking news events. While these sources were once dismissed as unverified or unreliable, they have increasingly become a vital part of the journalistic toolkit, often providing the first and most immediate reports on unfolding events.

The key has been not to reject these sources outright but to develop methods for evaluating and corroborating them, integrating them into the broader fabric of reporting. Similarly, in rare disease research, we must learn to integrate and evaluate a broader spectrum of data sources. Patient-generated health data, information from local health networks, and even anecdotal reports from patient communities can offer invaluable insights when traditional data sources are scarce.        

For instance, due to the founder effect, certain rare diseases may be geographically clustered. However, the true extent of this clustering—and the potential under-diagnosis within these clusters—cannot be fully understood without broadening our real-world data generation methods. By incorporating non-traditional data sources, we can uncover patterns and gaps that traditional routes may overlook. The challenge then lies in developing robust methodologies to assess the validity and relevance of these diverse data sources, ensuring they are not dismissed due to their position in the traditional hierarchy of evidence.

Embracing a New Perspective

By adopting this more inclusive approach to data sources, we can unlock new opportunities for understanding and treating rare diseases. Just as journalism has evolved to include and validate new forms of information, so too must the field of rare disease research. Addressing the challenge of patient dispersal in rare diseases requires a fundamental shift in how we perceive and evaluate data sources. We need to embrace this broader perspective on data quality, ensuring that we do not overlook critical insights that can only be gained from non-traditional sources.? Stay tuned for my next post (due August 20th), where I will focus on how broader ecosystem initiatives, such as European Health Data Space, will contribute to the shift in how we think of RWD generation in rare diseases.