Clinical Intake Hero banner

Achieve TMF inspection readiness and maintain oversight with AI

If outsourcing clinical trials is something you engage in regularly, then you’re probably familiar with the challenge of making the hundreds of thousands of critical documents ready for your TMF, which is a time-consuming, manual, and unavoidable process. 

But with the right partners and the right tool, you can become inspection-ready, ensure a complete TMF, and get a bird’s-eye view of your R&D landscape with a tenth of the effort. One of NNIT’s experts in the field shares his best advice for how to leverage AI and the Microsoft Power Platform to reach this goal.

The need for a better way

Large amounts of documents are commonplace in most companies in the biopharmaceutical sector. With documents often coming from a wide range of partners, CROs, acquisitions, and company mergers that rarely follow uniform standards, many companies in this industry struggle with creating significant overhead indexing and identifying and populating metadata, and when adding the risk of accidental unblinding and human errors, the need for a better way becomes apparent.

— For that, we envision a world where generative AI can not only classify documents and extract metadata, but also generate metadata that doesn’t necessarily exist verbatim in documents as well as reason and mimic human intuition in decision making and quality control. In fact, we’re currently working on a solution where we use such AI models to eliminate up to 90% of the manual work, says Tareef Roustom, Product Owner and AI Advisor at NNIT.

Convert urgent needs and good ideas into tools you can use to digitize procedures and create value for the business. That’s the concept behind the Microsoft Dynamics 365 Power Platform, designed to make it possible to develop a complete range of apps and IT solutions that are streamlined and secure. The benefits are efficient business development, rapid time-to-market, and relief for the IT Department, allowing various business units to develop their own solutions for their own needs.

Power Platform consists of nine different tools that can be used and combined according to requirements.

  1. Power BI: Gather knowledge and data to identify business opportunities at a detailed and high level

  2. Power Apps: Convert knowledge into apps and solutions

  3. Power Apps Mobile: Adapt your apps for mobile devices

  4. Power Apps Portals: Mix data from various apps (internal and external)

  5. AI Builder: Use artificial intelligence to develop pre-defined “point & click” models

  6. Power Automate: Bind your solutions together and automate business processes

  7. Power Virtual Agents: Build AI-powered bots for a range of support requests

  8. Power Pages: Design and publish websites able to run across a range of devices and browsers

  9. Microsoft Dataverse: Store and process your app data securely and in one location

Power Platform is what’s known as a low code tool, designed to enable non-IT Department personnel to use it. That shifts responsibility to the departments with a need, and frees up IT personnel to concentrate on the jobs only their expertise can tackle, without compromising security and compliance.

Efficient and scalable document processing

Many companies are currently lagging behind when it comes to performing quality controls on their clinical documents. That’s often because they use various document-sharing channels and systems, which are siloed and not built for purpose. This often leads to a variety of issues – issues that can be effectively resolved by combining enterprise-grade automation, artificial intelligence, and agile analyses.

— By building a cloud-native unified platform through which clinical trial documents are fed back to the sponsor, as well as leveraging a fully automated backend with multiple AI models working in harmony, life sciences companies can create a single highway, ensuring a consistent and almost real-time creation of TMF-ready documents, explains Roustom.

Tools like the Microsoft Power Platform, designed with scalability and built around collaborative teams, can easily be repurposed to accommodate the process of automating the creation of TMF-ready documents.

Computer and clinical trials

3 tips: How to get started

There are many benefits to reap when using practical AI and automation, but it’s also an approach that places demand on your business. Tareef Roustom offers his best advice on how to get the most out of it:

1. Assess your needs
AI and cloud-native platforms are designed to create TMF-ready documents at scale, and they require a stable internal TMF and established process in order to avoid the need for retraining the AI models and changing the automation flows:

— Even though most companies could benefit from automating the quality control of their clinical documents, the AI systems we use require stable internal TMFs and good-quality data to be able to achieve maximum potential. A helpful rule of thumb is to have an eTMF that has been in use for at least one year and around 50,000 documents covering the entire TMF to train the AI, as well as a monthly flow of a minimum of 10,000 documents to gain full benefit from the solution and make the solution economically viable, explains Roustom.

The platforms are therefore ideal and particularly beneficial for medium-sized and large companies.

2. Understand your processes and be prepared for change
It may seem overwhelming, but if you want to get the full benefit out of automation, it’s important that you review your current TMF processes. Established and well-defined processes are crucial for a successful transition and integration process:

— You must have full control of your processes. Our recommendation is to dedicate 70% of the effort associated with adopting AI in your TMF creation process to disambiguating processes and 30% to technology. This approach enables the identification of automation potential with the highest ROI, as well as the ease of integration with your downstream processes and systems.

3. Take a critical look at your digital maturity
To guarantee success, the business has to be digitally mature. Take a critical look at your business and decide whether you are ready to take the next step:

— What technologies do you currently use? And how disruption and digital-friendly is your IT landscape? Take a look at your Microsoft ecosystem and assess how the different systems work together. There are big synergy benefits to be gained in using the Power Platform along with Azure cloud and the services built into both, concludes Roustom.