One proprietary methodology powers all of Volv Global’s solutions and adapts to any difficult-to-diagnose disease. Each new disease is modelled from first principles, not adapted from an existing one.
This methodology detects patterns in real-world data that show a disease is present, developing, or likely to respond to a specific treatment, depending on what a clinical or commercial team most needs to know.
What is Volv Global's methodology?
Volv Global’s methodology looks beyond a single diagnosis code. It builds a more complete, personalised picture of each patient’s health journey from the real-world data available, rather than treating everyone who shares a label as the same.
That picture gives researchers and pharmaceutical teams a stronger evidence base on how a disease presents: the groundwork for a shorter path to the right care.
Volv Global's methodology learns the biomarkers of a disease directly from real-world data, building prediction models for difficult-to-diagnose conditions where unmet need is greatest.
Our methodology continuously turns real-world data into fresh insight along the whole patient journey, helping to establish how a disease progresses and what its natural history looks like.
The methodology separates patient sub-groups by how their disease presents and progresses, informing trial design and evidence strategy.
Volv Global's insights help reduce the cost of late or incorrect diagnosis: the ineffective treatment, the long-term care, and the decline in health that follows.
Insights at your fingertips
inVolv is Volv Global’s secure platform for delivering the insights generated by its methodology directly to partners and clients. It supports data-driven decisions with tailored reporting and analytics for each stakeholder, including interactive filters and visualisations such as charts and heatmaps.
Volv Global started in rare and difficult-to-diagnose diseases
Rare diseases affect fewer than one person in several thousand. Clinicians see them less often, and developers of therapies and payers have fewer patients to draw understanding from, which can leave people without a clear path to diagnosis or care.
Depending on how ‘rare’ is defined, more than 300 million people, between 3.5 and 5.9% of the world’s population, live with one of an estimated 6,000 to 10,000 rare diseases. Volv Global’s methodology began here, and now reaches far beyond it.
Navigating complexities
That same challenge is not unique to rare disease. Many other difficult-to-diagnose conditions share the same small, heterogeneous patient populations, and building a true understanding of their prevalence, progression, patient journey and outcomes is just as complex.
So Volv Global builds machine learning algorithms tailored to each disease, able to learn for themselves how to characterise it.
Of the thousands of rare diseases, only around 5% have an approved treatment, leaving the remaining 95% unaddressed.
For many of these diseases, the underlying disease biology is still poorly understood.
Small, heterogeneous patient populations present challenges for clinical development.
'Pipeline herding' means development efforts are unevenly distributed, leaving some diseases with a persistent gap in unmet need.
The cost of not treating rare diseases, in healthcare spend and in quality of life, is higher than most people realise.
Understanding every patient, across every disease.
Growing the real-world evidence on a disease, and a clearer view of how many people it affects.
Studying the early biomarker patterns that precede a condition, sometimes years before conventional signs appear.
Modelling how disease progresses across diagnosed populations, using real-world evidence.
Sizing the diagnosed population not yet on treatment, and showing where the gap to care opens.
Volv Global powers solutions across stakeholders
De-risking clinical trial planning by learning the biomarkers of a disease directly from real-world data, supporting more accurate patient search and recruitment.
Building natural histories of disease and generating the real-world evidence regulators and payers need for approval and reimbursement.
Turning real-world evidence into insight on disease progression, outcomes and endpoints, supporting clinical conversations with treating physicians.
Profiling the populations most likely to benefit from a therapy, sharpening launch decisions.
Bridging technology and treatment
Volv Global’s proprietary methodology extracts meaningful insight from sparse, unstructured or inconsistent data in electronic health records and claims data, learning the signature of each disease directly from population-scale data.
This methodology builds proprietary computational models that detect biomarkers, phenotypes and other clinical features directly from that population-scale data. For diseases with especially heterogeneous populations, it also clusters patients into meaningful sub-groups, reflecting the different ways a disease actually presents.
Volv Global's methodology is customised for each project. It learns from the data available rather than from existing knowledge or guidelines, building a mathematical representation of the disease from first principles as its starting point.
Volv Global's models combine computational rigour with expert human judgement: each model is reviewed by disease experts, and results have been presented at ISPOR, ERS and ATS, including the largest machine learning study of Pompe disease to date (3,549 patients).
Volv Global does not hold patient data. No identifiable patient data is obtained at any point; all work is conducted through trusted data partners operating under the privacy and regulatory frameworks that apply in each market.
The methodology has been applied to claims and electronic health record data across geographies and data partners. It is designed to learn once and then be adapted across different coding systems and ontologies. Most machine learning models need retraining for each new setting; this one does not.
Volv Global's methodology integrates with the healthcare systems already in place. Because coding systems and clinical practice differ by country, it learns its own model for each one, adapting to local language, data structure and clinical thinking.
