Non-CRF Data in Clinical Trials: Comprehensive Aspects

Clinical research heavily relies on effective clinical data management, which can significantly influence the success or failure of a study. The study database consists of two primary components: Direct Data Capture (Paper CRF or eCRF) and Non-CRF data. Unlike Direct Data Capture, digital data collection in clinical trials deviates from EDC architectural principles due to continuous data collection, unmappable data lacking clinical significance, and potential volume exceedance of the EDC system's storage capacity. Understanding the impact of dealing with Non-CRF data is crucial. 1

What is Non-CRF data:

Understanding the impact of dealing with Non-CRF data is crucial. Non-CRF data, also known as a third-party vendor data, is/are collected through alternative channels. The integrity and quality of data from external sources play a critical role in clinical trial data management and study success. Non-CRF data encompasses central and core laboratory data, central imaging (including medical images), subject diaries such as patient questionnaires related to quality of life, pharmacokinetics and pharmacodynamics data, safety laboratory data, genetic data, ECG, drug accountability data, biomarkers, device data, and randomization data etc. Its like many rivers merging in to the ocean.

Collecting Non-eCRF data:

It’s pre-defined that how the non-eCRF data will be collected. DTA- Data transfer Agreement drives modus operandi for transferring data between Vendor and EDC (majorly but not the only) of CRO/Sponsor. Usually it comes in a form of separate file and demands manual reconciliation, which is having its own challenges and limitations.

04 Major Sources of Non-eCRF data collections:

1. Vendor Data: CORE Lab, central imaging (any type of medical images), Central lab, Clinical/Bio analysis lab, PK/PD lab , safety laboratory data, genetic data, biomarkers
2. Direct from Patient: Patient data like PDC which includes patient-oriented tools such as questionnaires pertaining to the quality of life  
3. Health technologies/ IoT: Wearables/Sensors, Medical Devices equipped with remote data transmission, personal health record, Interactive health tool, health mobile application, Microchip / smart tattoo
4. EHR: Electronic Health Record System

Current Practice for data transfer:

1. Secure FTP transfer
2. Physical media such as DVD, storage devices

Challenges:

Manual reconciliation process involves risk of missing out on errors, missing data across datasets, identifying duplicate records and so on. Any lapse in non-eCRF data can have significant implications, directly impacting trial participant and study safety.

While working with non-eCRF data poses unique challenges, it also offers valuable insights to enhance the safety and effectiveness of new drugs and treatments. Some common challenges and considerations when dealing with non-eCRF data include:

1. Standardization: Ensuring that non-eCRF data is standardized is crucial for accurate analysis and comparison across different sources. This may involve developing data collection guidelines and implementing standardized formats.
2. Timely and accurate collection: Collecting non-eCRF data in a timely and accurate manner can be challenging, especially when relying on manual processes or data input from multiple sources. Implementing strategies to streamline data collection and improve data quality is essential.
3. Data security: Protecting non-eCRF data from unauthorized access is crucial, especially when dealing with sensitive patient information. Ensuring appropriate data encryption, access controls, and compliance with data protection regulations is important.
4. Integration into the overall dataset: Incorporating non-eCRF data into the overall clinical trial dataset can be complex. It may involve data mapping, transformation, and validation to ensure compatibility and consistency with other data sources.
5. Data quality: Non-eCRF data can be prone to lower quality compared to eCRF data due to factors such as human error, incomplete data, and inconsistent formatting. Implementing data quality checks and validation processes can help identify and address issues. Here, thorough drafting and review of DTA comes in picture which describes database structure, definitions of data and timelines for data exchange etc. for ensuring overall data quality.

Overall, working with non-eCRF data requires careful planning, standardization efforts, and robust data management practices. It can provide valuable insights and enhance the understanding of treatment effectiveness and patient outcomes in clinical trials.

Several underlying reasons contribute to the challenges faced:

• Adequacy of DTA across data sources
• Lack of QC before data intake from sources
• Lack of standardization, work flow, triggers, oversights
• Manual processing (Manual query, reconciliation review, report etc.)
• Post EDC Go-live changes like specification/edit check
• Inadequate communication / governance between stakeholders

Factors affecting data variations across sources can be both controllable (like through standardization) and non-controllable:

• Variation in data formats (Not aligned with pre-defined standard expectations)
• Technology/instruments (Technology variation like Device variation, screen size, operating systems, applications, mid trial upgrades etc.)
• Timeline variation 

What Next: 

Soon be practical for the industry to use customizable platforms that leverage the DTA, accommodate with vendor data management system (like LIMS for Lab, ITR for randomization etc.) to upload and manage their data seamlessly, ensuring traceability, quick turnaround and robust data quality with minimal human efforts.

The utilization of LOINC (Logical Observation Identifiers Names and Codes) serves as a pre-coordinated laboratory coding system within healthcare IT systems, facilitating interoperability and data exchange. It includes lab tests, clinical measures, HIPAA documents and standardized survey instruments. It also contains terms for human clinical research but its scope goes beyond research use. Interestingly, LOINC is used in over 170 countries and is mandated in 30. LOINC is maintained by Regenstrief Institute, INC and is supported by the US National Library of Medicine (NLM). But there are many types of non-laboratory measurements categorized as non-CRF data but not addressed by LOINC (like Imaging Tumor Measurements, Sleep Measurements/ Polysomnography). LOINC working group has addressed concerns on using LOINC in clinical trials and FDA-CDISC will streamline it for further use. 2,3

Industry, Regulators and Experts aligning to make this even flawless, accurate and practical utility for Clinical Trials with such multiple data sources. This discussion welcomes your view point and valuable suggestions.

Bhavik Patel Nimesh Parekh Dr. Jalak Patel Dr. Shariq Anwar Kondari Srinivasa Rao Anurag Parihar Rushiraj Vaghasia Vaibhav J.