Challenges Associated With Using Large Data Sets for Quality Assessment and Research in Clinical Settings

Area of Science:

• Health Informatics
• Clinical Research Data Management
• Big Data in Healthcare

Background:

• Electronic health records (EHRs) generate vast amounts of data for research.
• Challenges include data access, security, quality, consistency, and specialized staff.
• Existing data infrastructure often requires significant adaptation for research purposes.

Purpose of the Study:

• To describe the experience of creating a data-mart for clinical research.
• To detail the process of integrating data from multiple electronic sources.
• To identify and address challenges in utilizing large-scale clinical datasets.

Main Methods:

• Data integration from admission-discharge-transfer, cost accounting, EHR, clinical data warehouse, and departmental systems.
• Development of a domain-specific data-mart within a single hospital network.
• Utilized the National Institutes of Health Big Data to Knowledge framework for analysis.

Main Results:

• Successfully assembled a data-mart containing over 760,000 discharges (2006-2012).
• Encountered and documented issues related to data quality, standardization, and access.
• Identified specific obstacles in data curation and manipulation for research.

Conclusions:

• Building data-marts from diverse electronic health data is feasible but complex.
• Proactive strategies are needed to overcome data quality and integration challenges.
• This approach supports large-scale clinical, epidemiological, and cost-effectiveness research.