Using new diagnostic tools to detect patients earlier means working in heterogenous environments. To this end, Volv has developed a technical capability and framework that we can use to deploy models in any data setting.
Tooling must effectively address within the following complexity:
Healthcare systems vary country by country and in some cases region by region: clinician terms are different, clincial coding can be different, and even the way disease is thought of and treated can be different.
The data that we need to work with varies by healthcare providers or by research dataset. The skills available on the ground also vary greatly. Volv has created a simple methodology to work with these systems in efficient ways to reduce the barriers to effective data access and analysis.
Over and above the technical variance, each country has different population characteristics and the algorithms Volv develops need to be generalisable within these different contexts. We can adapt the methods to suit any large changes in population demographics should they be needed, and in this case, the system can become even more generalisable.
To that end, Volv has developed a technical capability and framework that we can use to deploy shown in the architecture above, across any country in the world.
The deployment of predictive disease models is impacted by many internal and external factors which we work through with our customers. We then build a collaborative target operating model for deployment and maintenance.
