[Paper Review] Report on Data Quality in Biobanks: Problems, Issues, State-of-the-Art

Suneth Ranasinghe, Horst Pichler+1 more|arXiv (Cornell University)|Dec 17, 2018

Ethics in Clinical ResearchMedicine17 references3 citations

TL;DR

This paper proposes a comprehensive data quality management system for biobanks, emphasizing metadata quality as critical for enabling efficient biomedical research. It outlines key dimensions of data quality, defines best practices for documentation, and positions biobanks as data brokers where high-quality metadata ensures reliable sample and data retrieval for clinical studies.

ABSTRACT

This report discusses the issues of data quality in biobanks. It presents the state-of-the-art in data quality: the definition of data quality, the dimensions of data quality, and the quality management system for achieving or describing the aspired data quality characteristics and we present and discuss all elements for such a data quality management system. In depth, we discuss the requirements and the context of data quality for biobanks in particular, where we argue that biobanks can be seen as data brokers, where the indented use of the data is to support the search for suitable material and data in the preparation of medical studies. For such an intended use, the quality of the metadata is of high importance and biobanks have to emphasize to strive for adequate documentation of the quality of the data annotating the samples.

Motivation & Objective

• Address the critical challenge of data quality in biobanks, where poor metadata undermines biomedical research efficiency.
• Identify biobanks as data brokers whose primary function is to support the discovery of suitable biological samples and associated data.
• Establish a structured data quality management system tailored to biobank operations and research use cases.
• Emphasize metadata quality as a central factor in ensuring data usability and reproducibility in medical studies.
• Provide a state-of-the-art framework for defining, measuring, and improving data quality in biobank environments.

Proposed method

• Define data quality through established dimensions such as accuracy, completeness, consistency, and timeliness.
• Propose a data quality management system integrating policies, processes, and tools for monitoring and improving data quality.
• Frame biobanks as data brokers, where the primary output is discoverable, well-documented data and samples.
• Stress the importance of standardized documentation practices for sample origin, handling, and associated phenotypic data.
• Integrate quality management principles from information systems into biobank-specific workflows.
• Use case analysis to illustrate how metadata quality directly impacts the success of medical research preparation.

Experimental results

Research questions

• RQ1 What are the key dimensions of data quality relevant to biobanks, and how do they differ from general data quality frameworks?
• RQ2 How can biobanks function effectively as data brokers when the primary output is not data but discoverable, high-quality metadata?
• RQ3 What are the core components of a data quality management system tailored for biobank operations?
• RQ4 Why is metadata quality more critical than raw data quality in biobank contexts?
• RQ5 How can biobanks ensure consistent, standardized documentation to support long-term research usability?

Key findings

• Metadata quality is the most critical factor in enabling effective search and retrieval of biological samples and associated data in biobanks.
• Biobanks function as data brokers, where the primary value lies in the discoverability and reliability of documented data, not just the physical samples.
• A structured data quality management system must include defined quality dimensions, governance, and documentation standards to ensure usability.
• The state-of-the-art in data quality for biobanks emphasizes completeness, consistency, and timeliness of metadata as essential for research reproducibility.
• Standardized documentation of data quality characteristics enhances trust and facilitates integration into larger biomedical research initiatives.
• The paper establishes a foundation for future work by defining a comprehensive framework applicable to biobank data quality management.