data oversight

One of the issues that may affect data quality oversight is data sources. Nowadays, many clinical trials collect and process clinical data that is not included in the CRF. Import and management of external data provided by other vendors (such as central laboratories) is a challenge that aims to maintain data accuracy and integrity and includes the reconciliation between CRF data and data from external sources using key variables to match subject, visit and sample with each sample record.

How do we guarantee data quality oversight?

For each variable included in the data transfer, it is important to define name and label, length and type (e.g. numeric, character, date, etc.). The number of transferred data files and format should also be defined (e.g. xls, csv, SAS, etc.) along with the frequency for providing data transfers and the type of the data transfer if cumulative or incremental.

Efficient communication with vendors is fundamental to define the data transfer process and to ensure timely receipt of data. The Data Manager should agree with the vendor the structure and format in which external data is to be provided and imported; the Data Transfer and Import Specifications document is then developed to collect all required information including transmission details (e.g. email, cd/dvd, etc.) and data protection (e.g. ZIP file with password).

The goal is to receive a data transfer that matches the agreed specifications, to correctly import the external data without compromising quality and the study timelines.

It is preferable that the data to be transferred is provided in the same format as it is expected to be when imported file (i.e. usually SAS format so that no programming will be required to import the data in SAS); otherwise, a specific import program is needed. The data to be transferred should not contain extra information (e.g. headers or other data that is not to be imported) and should not contain formatting, colors etc. What is important is that the data file should not be manipulated in order to be imported, data should be received in the agreed format allowing be imported to the required format without any data manipulation.

Once the data is imported, the reconciliation process will take place. Specific computerized checks will be developed to identify discrepancies between external data and CRF data reducing reconciliation time and increasing data quality. These reconciliation checks should be agreed with the Sponsor during the study start up, together with the data cleaning processes and discrepancy management.

Vendors can provide several types of clinical data collected during a clinical study and these usually play a key role (e.g. being the primary endpoint of the study). If the data is not adequately managed, this could affect the study outcome. The quality of deliverables is therefore crucial, although in some cases it might be difficult to achieve. The poor quality of data or missing deliverable, along with a bad data cleaning, could result in the delay of study closure or even in an inability to properly analyse the data, thus compromising overall study results.

Getting Support from a Biometrics Vendor to Ensure Oversight

As a global, data-driven CRO, CROS NT supports companies in implementing and executing quality biometrics processes. How can we support when it comes to data management?

  • Managerial Consulting: gap analysis and review of SOPs and biometrics processes
  • Strategic Consulting: organization of biometrics processes and implementing necessary SOPs and documentation
  • Expert resources either for project-based studies (centralized biometrics) or Micro FSP for flexible resourcing without the need of setting up an internal biometrics department
  • Oversight support in compliance with ICH GCP E6(R2) in terms of consultancy and technology solutions (comprehensive EDC portfolio, BI and Centralized Statistical Intelligence tools, Data Warehousing)