FAQs | DEEP Measures

Digital measures can offer a whole new perspective and scientific insight. Not very different from what imaging technologies achieved - allowing us to see inside the human body. Digital measures allow us a new perspective into human lives, how diseases and interventions impact that day to day.

There are many other benefits too, for example they can be more sensitive, allowing us to measure the impact of interventions faster and more precisely than conventional measures.

Regulators are also starting to demand more patient-centric evidence about the value of our drugs and today we're quite limited in our ability to measure these things objectively.

Not at this time, but DEEP is keen to learn more about customer expectations with regards to this to better understand what export and archiving functionality would make the most sense. Some content, such as digital briefing books can be exported as PDFs for archiving and storing in internal systems (e.g. RIMS, etc.).
The catalog is structured according to latest best practices for defining digital measures and related solutions. The catalog helps assess the completeness and maturity of different solutions. Users can pull evidence and “components” from the catalog and build their own measures for specific contexts of use, leveraging what is already available. The catalog is directly connected to the collaboration platform, enabling easy working together with internal teams, external collaborators and stakeholders like regulatory authorities. The catalog saves time and cost and enables the harmonization and consistency of measures across contexts of use and a range of technologies.
DEEP can act as the matchmaker between parties to see if the content owner would be open to sharing the privately captured content under some conditions, e.g. in exchange for payment, sharing of other valuable content or other such scenario.

The DEEP catalog has a very different purpose, it helps users understand the completeness and maturity of specific measures and it directly feeds the digital measure builder to construct fit-for-purpose measures. The DEEP catalog mainly serves the purpose of developing measures.
DEEP is a technology and expert services company, whereas DiME is more in the business of developing best practices and thought leadership. DiME’s work has been very important in creating DEEP, but DEEP’s role is more to implement these recommendations in a pragmatic way through a technology platform and the DEEP model and support companies in developing measures directly in this way.
DEEP enables cloud-based submissions and regulatory interaction tools somewhat similar to Accumulus. However, DEEP’s focus is very specific to digital measures and is optimized to facilitate DHT-based measure submissions and reviews. Accumulus’s solution is intended to be much more broad. Accumulus mainly serves pharma companies, whereas DEEP is more open and welcomes any stakeholder actively involved in developing digital measures as a member.
The Critical Path institute is a thought leader in many important clinical domains and does a lot of important scientific work to advance these areas. The DEEP model and tools can be utilized to support this work, to connect results with other results coming from other areas, bringing efficiencies and benefits to all parties.

Currently information in the catalog is populated by DEEP’s Digital Health Analysts following structured checklists. In the future, more automation will be introduced with the necessary QC controls in place. The DEEP model is still evolving and the community as a whole is still in very active learning mode, so continuous adjustment of how content in the catalog is structured is expected for some time.

The DEEP model incorporates the latest thinking across a range of regulatory and expert sources in terms of what evidence is required for different purposes. The DEEP model then helps applicants to understand how complete and mature their evidence package is for a particular context of use or regulatory stakeholder. For example, the FDA and the EMA have a different approach to evaluating evidence and DEEP helps assess what you have and structure it in the optimal way for each type of engagement.
The DEEP model is a useful model to use in order to structure content for evidence dossiers and briefing book materials. The model helps standardize measures and solutions to capture the measure, independent of the specific technologies and versions, enabling more effective lifecycle management and consistent evidence generation. The DEEP model includes detailed checklists that incorporate the key regulatory requirements for acceptance as well as many other industry best practices (DiMe V3+, CTTI, etc.).

An algorithm in the DEEP model is a generic term that explains any data transformation from one kind to another, typically from “raw” sensor-derived format into a more human-understandable “health data”. For example, converting accelerometer readings into step counts. This includes any transformation, including human rating steps by an expert assessor for example. All relevant information that affects the performance of the algorithm is captured. The algorithm layer mainly focuses on these transformation steps, whereas the “Raw data / processing pipeline” layer describes the end-to-end data pipeline, from the capture of the raw data to how it is processed and stored in a clinical database. For example, some processing may take place in the sensor device, some may happen in a middleware layer and further processing may happen even later.

Building Digital Measures