With deep expertise modernizing, integrating, and enhancing health information technology (IT) systems, Halfaker, an SAIC company, delivers extensive health data services to ensure care providers, researchers, and policymakers can securely access an ever-expanding pool of valuable health data to facilitate life-saving research and improve care outcomes. Since 2018, we have been working with one of our federal health clients to enhance and support a robust health informatics and computing infrastructure that uses leading-edge analytics tools and techniques to support health services research, epidemiology, decision support, and business intelligence capabilities for clinicians and researchers to assess and improve healthcare. As a key resource in our client’s cooperative scientific community, the platform offers comprehensive data services to research projects and assists in providing custom data sets tailored to inform each partner’s business or research needs. With thousands of health researchers accessing the health data of millions of patients, much of which is unstructured, our client requires innovative technologies to maintain data availability, speed of access, and data accuracy. 

The challenge

Every day, our client runs multiple major health programs that rely on a centralized health informatics platform that captures the data of millions of patients across thousands of healthcare facilities for research and analysis. With the intake of data growing almost 10 TB annually, our client requires increased enhancement and sustainment support for interface management, data set creation, and user services to enable clinicians and researchers to derive value from these massive datasets.

Our solution

To handle the growing number of users and data sources, Halfaker offers our client comprehensive data management, provisioning, visualization, and analysis services across thousands of research projects. While our primary focus is on providing various datasets to researchers, we also provide a robust health computing infrastructure, combined with popular analytical tools such as SAS, Stata MP, and R, to help researchers explore, analyze, and visualize those datasets to quickly extract valuable insights. Using proven big data tools like Hadoop, IBM Machine Learning, and Tensor Flow, our team maintains a health informatics platform for researchers to load and perform data-intensive computing on large and complex datasets to enhance research capabilities. To further boost data exploration capabilities, we also develop and optimize NLP modules using oNLP, iKnow, and SOLR/Lucene to extract concepts from unstructured text (e.g., clinical notes), show relationships, and map concepts to terms. By incorporating NLP capabilities, we enhance the accuracy and completeness of health information, which in turn enables researchers to broaden their understanding of patient needs and improve diagnostic and treatment decision-making. 

Halfaker’s data exploration tools and solutions make data available through a highly secure and user-friendly web portal that offers interactive search and filtering capabilities to create customized data sets. Our team created researcher and operational topic-specific data sets and a comprehensive suite of databases, models, business practices and file structures to ensure rapid responsiveness and data quality in population health analytics. With our understanding of the client’s business practices, data combinations, and data access, we create, populate, and maintain data cubes using a data warehouse to fulfill researchers’ and operational stakeholders’ requests for data sets. We tailor datasets, database schemas, or database views to specific project needs and deliver current and accurate data models and metadata as defined by our client’s terminology and functions. 

Our data managers amass massive data sets for researchers, allowing researchers and operational stakeholders to combine data sets with partner entities. Where feasible, we implement data standardization and exchange initiatives to expand data integration and sharing and increase the value and demand of data for operational purposes. We create customized scripts for database creation, table creation and data extraction, and data view creation/alteration to boost database query efficiency, automate data capture and transformation, improve data quality, and save database and server processing resources. As part of our data set delivery, we develop data reports and data profiling information, provide data mining and BI support, and continuously research and recommend ways to analyze data more efficiently. Halfaker works with researchers and operational stakeholders to ensure the data sets provided are statistically solvent and that the results can be reproduced independently and withstand a scientific peer-review as needed. 

Realized benefits

Halfaker’s enhancement and sustainment support for our client’s health informatics infrastructure, data, and tools offer comprehensive solutions for their growing research community users and data needs. Our health data management expertise and tools implementation have enabled:

• Time-saving, efficient research and analysis of domain datasets
• Significant reduction in exploratory queries with unknown answers
• Increased database and server resource savings