By Ellen Derrico | February 13, 2023
Here at Proscia, it seems like we have a front row seat to the increasingly complex challenges facing R&D in the life sciences. For our customers, the processes, systems, and data models used in research are more sophisticated than ever. More stakeholders mean more requirements. More compliance policies mean more reporting. And when it comes to digital pathology, more images mean more surface area for messy data.
If this is the reality for life sciences R&D, then maintaining data integrity and data quality is vital – because bad data can lead to bad science.
What’s the difference? Data integrity vs quality
Though often referred to interchangeably, data integrity and data quality are not the same thing. Data integrity refers to the accuracy and consistency of data in a database or system. It is the measure of the completeness, accuracy, and reliability of data, as well as the ability to protect data from unauthorized modification or deletion. For pathology R&D, data integrity is about making sure individual whole slide images (WSIs) are stored in the correct study folder and remain associated with the correct metadata and image analysis results.
Data quality, on the other hand, refers to the overall excellence of data. It is a measure of how well data meets the needs of its intended audience and how fit it is for its intended use. Data quality includes factors such as accuracy, completeness, timeliness, and consistency, as well as more subjective factors such as relevance and usefulness. When it comes to pathology R&D, maintaining image data quality is particularly important. A poorly scanned image from a scanner, for example, may contain bad data that gets incorporated into future analysis. This is one way bad data can lead to bad science.
What’s standing in the way? The obstacles to data integrity and quality
It’ll come as little surprise that there are significant obstacles to maintaining data integrity and quality – particularly in the face of ever-growing research data volumes. It doesn’t help, for instance, that workflows and data are spread across multiple platforms and disparate systems (Excel, eSM, HALO, etc.). As data sources proliferate, few organizations are able to centralize data views for the day-to-day work of pathology R&D. This slows down research.
Image scanning is another obstacle. Image quality can vary dramatically based on a variety of variables – such as the quality of the viewer used, issues with resolution or clarity, and the speed at which images are scanned. The need to rescan faulty or sub-par slides only acts as another break on the steady flow of research.
The points at which data is entered, stored, and transferred can also lead to issues. With so many partners (CROs, internal collaborators, outside consultants) creating and sharing data, there are more opportunities than ever for the inconsistent labeling of the data. And let’s not forget that the potential for human error is only increased when users fail to follow established data management protocols – which, for many organizations, happens frequently.
Concentriq for Research data model
Key capabilities that researchers depend on – such as querying data, filtering information, and analyzing WSIs – are only possible with good data integrity and quality. But as sources of data grow along with data volume and diversity, organizations need to prioritize, standardize, and optimize the process of data collection and management across various technology platforms to ensure data integrity and quality.
Concentriq for Research can help. Our data model is purpose-built for scientists and pathologists working with images, annotations, analysis, and more. Concentriq for Research captures data in a way that is accessible, interoperable, and reusable. You’ll be able to:
- Streamline data capture: Pull in data from spreadsheets and LIMS in a standardized manner that sidesteps the risk of error associated with manual input.
- Manage data diversity: Deploy sophisticated data models that support image and non-image data in a unified, accessible hub.
- Ensure data governance: Use configurable roles and permissions, data access policies, and audit logs to enforce data governance for all information brought into the lab.
- Automate image quality control: Tap our built-in AI to identify and reject low-quality images from impacting results.
- Create SOPs and protocols: Standardize experiments and results capture across teams and improve data quality and compliance.
- Get FAIR: Maintain a FAIR (Findable, Accessible, Interoperable, Usable) database that meets the needs of research.
- Find what you need: Easily access and filter data with our advanced search-ability functionality.
At a time when data integrity and quality can mean the difference between good science and bad, it’s important to tackle the challenge head on with a powerful platform like Concentriq for Research that is built to meet the needs of pathology R&D for even the largest life sciences firms.
Solutions for Life Sciences