Inspiring future progress in rare diseases will require an increased focus on real-world data and AI

RwHealth’s Mike Hughes Director, Real World Evidence speaks on the value of data-driven innovation within Rare Diseases

AI-driven initiatives hold great promise in augmenting rare diseases research and prediction of real-world outcomes. 

The emergence of AI and machine learning technologies have begun to transform healthcare and medicine. In a rare diseases context, AI-driven platforms like RwHealth’s Data Science Platform (DSP) work intimately with input data across multiple medical datasets to improve the diagnosis, prognosis and treatments for rare diseases. 

Due to next generation technological advances, health providers can now leverage AI and real-world data initiatives to share, combine and link critical information to enhance treatments. Tools that work intimately with predictive analytics, like the DSP; allowing organisations to access previously untapped data, patient-level data, as part of their rare disease programmes. By providing both an extensive knowledge base and insight tool for rare diseases, the DSP bolsters further advancements in disease epidemiology, associated diagnostic and treatment pathways, as well as patient outcomes. As Mike Hughes, Director of RWE at RwHealth, states:

‘AI and Machine Learning platforms, like the DSP, allow for healthcare professionals to interact more closely with rare diseases data. As a highly centralised resource, the DSP combines best-in-class data technologies and expertise to optimise treatments for patients with rare diseases.’ 

The Rare Diseases Context

The most prominent challenge in rare disease research is accessing and, then analysing, relevant clinical data. 

In times prior, impeded access to data posed an insurmountable challenge to rare disease research. Historically, accessing curated data from a small and widely dispersed patient population has proven to be more logistically difficult for health organisations. As rare diseases are largely underrepresented in the published literature, there is a paucity of validated outcome data to inform further treatments. Due to the frequency of rare diseases and scarcity of patient data, clinical trials are often too small to identify subgroups or specific phenotypes with particular unmet need or different treatment response. Moreover, they rarely reflect the identification, treatment and outcomes for these patients in a real world setting.

Integrating findings from the small patient populations with rare diseases, and the subsequent sharing of genetic, drug discovery, treatment and outcome data was a difficult task for health and research organisations. Historically, the small numbers of patients involved may have impeded the development of novel treatments and improved diagnosis – to such an extent that 95% of rare diseases currently have no cure. 

In some cases, the lack of knowledge regarding disease pathophysiology, molecular pathways needed to develop therapeutic targets, prevent patients from receiving appropriate care. This has historically limited the timely care and possibility of patient enrolment in recent clinical trials. In observing the current research landscape, Mike Hughes sees the lack of AI-driven tools that integrate and streamline data sources as a particular obstruction for true rare diseases innovation:

‘Many currently available real world data sets only capture a particular fragment of the patient journey, which limits the ability to conduct certain types of studies. Such data sets do not offer the possibility of conducting the longitudinal tracking of patients, which is crucial for understanding the diagnostic and treatment pathways, and prognoses, of rare diseases patients. Epidemiological research relies on continuity and integration to succeed, which, unfortunately, are lacking in general’

This dearth of knowledge may have contributed to delayed or inaccurate diagnoses, thus further limiting the availability of data. 

Despite this, research into rare diseases has been transformed in recent years, driven in part by dedicated approval or reimbursement programs such as orphan drug designations, as well as by advances in molecular biology and its application to treating the genetic defects that underly so many rare diseases.

The Need for Intelligent Data Platforms 

In this context, the need to engage with networked and AI-optimised data is abundantly clear. Without platforms that accumulate insights from clinical data exchanges, interacting with basic biological access knowledge, research and tools will continue to blight the lives of rare disease patients. 

Thanks to emerging data-driven platforms like the DSP, healthcare organisations can now overcome these challenges and hone focus into (previously unmanaged) rare diseases. As Mike Hughes states, using ‘linked datasets and AI-driven insights through the DSP can expedite the development of rare diseases innovations: 

‘We need to look more deeply into how curated data can benefit the rare diseases domain. By using AI and predictive analytics to aggregate granular insights from small data sets, health organisations can rely on the DSP to accelerate diagnoses, identify new rare diseases and improve patient care’

The development of technologies like the DSP provides a great opportunity for progressing research and development in rare diseases. Through predictive analysis, the DSP integrates data from multiple epidemiological and laboratory datasets, as well as electronic health records, to assist in diagnosis and treatment. By integrating findings from a comprehensive data landscape, the DSP accelerates drug discovery, development and clinical trial target selection for rare diseases. 

The DSP: Leveraging AI to Inform Rare Disease Innovation 

The rise of data-driven platforms such as DSP permits insight generation from across an end-to-end view of the rare disease patient. As Mike Hughes states, the DSP’s predictive analytics capacities allow health organisations to accrue sufficient data to build innovative treatments:

‘The DSP allows companies to expedite the process for developing particularly innovative treatments for rare genetic diseases. With curated data, the DSP allows for the improved forecasting of prognoses, the identification of unmet needs in a patient population by current therapies, and so on’ 

The DSP’s harnessing of AI and machine learning allow for healthcare organisations to empower their rare disease initiatives, through data-driven decisions. By unlocking the power of data, the DSP uses predictive analytics to catalyse the identification and design of highly efficient and much-needed therapies. In applying AI to derive crucial insights, the DSP provides the pharmaceutical industry with the opportunity to diversify rare diseases pipelines and enable more competitive strategies – while addressing gaps in rare disease research and other areas of unmet need.

To Conclude

The use of data-driven platforms like the DSP holds tangible benefits for rare diseases. Applicable to a wide spectrum of disease pathways, its leveraging of AI allows for patients, clinicians and manufacturers to benefit from quicker diagnosis and patient identification, more efficient drug discovery, and ultimately, to better patient outcomes. 

The DSP’s capacity in modelling patient data through predictive analytics allows for further diagnosis support and innovation in rare diseases. RwHealth’s insights are gathered using our DSP. Built with the intention of supporting providers in their endeavours to provide innovative therapies, our DSP helps you to curate and execute on real-time analytics necessary for your operational success. 

We invite any of you interested in trying the DSP for yourselves to access a free trial of the platform by contacting a member of our team.