The Top 4 Ways De-Identified Data Improves Research
I’ve spent the better part of my career connecting researchers with data. For 10 years at a large academic medical center, my job was to support researchers with various types of data. Through my experiences there—working first on a genomics project, then on building and managing the research portion of the center’s enterprise data warehouse (EDW), I’ve become absolutely convinced of the immense value data warehousing holds for efficient, effective research that ultimately benefits patients.
All patients and providers want better care—research helps determine what we mean by better care. Research is a critical part of creating a better healthcare system, so we owe it to patients to make the research process more efficient. This article describes research roadblocks and the top four ways de-identified data removes these roadblocks and improves each step of the research process.
Research Process Roadblocks
Although procedures and protocols differ somewhat from person to person and facility to facility, most clinical researchers follow four steps before starting their research project:
- Hypothesis generation.
- Cohort exploration (determine if enough patients match study criteria).
- Grant application (apply for funding to pursue research).
- Institutional review board (IRB) application.
Only after these steps are completed can researchers truly start to answer their research questions. Although this seems fairly straightforward, anyone involved in research knows how difficult and time-consuming these four steps in the research process can be. These initial steps, which are absolutely fundamental, can be hampered by significant roadblocks:
Hypothesis Generation Roadblock
Researchers may come up with a very intriguing hypothesis, but it has to be sufficiently backed up by previous research and current data. Taking a hypothesis to the next level—cohort exploration and grant application—requires significant time and resources. Without exploratory tools to help assess the validity of the hypothesis from the outset, researchers might invest too much in a weak hypothesis.
Cohort Exploration Roadblock
Insufficient recruitment is one of the main reasons research projects fail. Researchers must determine whether or not they have access to enough patients to deliver the statistical power needed to generate conclusions from their proposed study. Making this determination is harder than it seems. Typically, researchers have to turn to operational analysts for this information. Their data requests go into a report queue, where they are typically prioritized much lower than operational data requests. In short, it can take a long time for researchers to make this basic determination.
Grant Application Roadblock
Once researchers decide they have a good hypothesis and enough patients to inform an effective study, they often apply for assistance in the form of a grant. Without a doubt, grant applications backed up with strong data are far more powerful than those without. However, trying to collect this data for a grant application can be challenging and frustrating for researchers for the reasons mentioned above (hypothesis generation and cohort explanation roadblocks). In particular, funders want to see applications that show thorough cohort exploration. Grant applications are strengthened when research organizations have strong histories of success in recruiting patients for studies (and can prove it with data).
IRB Application Roadblock
Submitting an application to the IRB is an important but often cumbersome and time-consuming process. The IRB process is regulated by HIPAA to ensure that all research projects respect patient privacy rules. HIPAA is very specific about how researchers can use data if it contains any kind of patient identifiers. The IRB application requires specific information about the cohort being studied:
- Cohort size.
- Protocols researchers will follow.
- Specific data points researchers will need to access throughout the process.
Researchers must follow each of these steps (representing months of work) before the research project can even begin. A lack of solid information at any point in this process can result in significant roadblocks and wasted time and resources.
The Top Four Ways De-Identified Data Improves Research
They keys to removing research roadblocks is de-identified data and self-service tools. Self-service tools allow researchers to explore data with much lower regulatory barriers while maintaining patient privacy. De-identified data has the power to remove many—if not most—roadblocks leading up to a research project and increases researchers’ odds for success in each step of the research process:
- Hypothesis generation: de-identified data enables early discovery and exploration to test the validity of the hypothesis before committing time and resources.
- Cohort exploration: de-identified data enables researchers to explore and refine their cohorts and determine whether the patient population can support the scope of the project.
- Grant application: de-identified data enables researchers to put together strong grant applications without having to tax the resources of enterprise data analysts—and without having to wait for analysts to answer relatively straightforward questions.
- IRB application: de-identified data enables researchers to come to the IRB with a strong, fully supported application. A data-driven research process ensures that both researchers and IRB reviewers don’t have to waste their time on projects that may not be viable.
Learn More about the Healthcare Research Analytics Adoption Model
Level 1 of the Healthcare Analytics Adoption Model (a framework for analytics implementation that applies lessons learned from the HIMSS EMR Adoption Model) is providing de-identified data marts and self-service tools. I’m working on adapting this model to address the needs of research organizations—to outline steps organizations can take to build a robust analytics infrastructure to support research. Specifically, the Healthcare Analytics Adoption Model provides the following:
- A roadmap for health systems to measure their progress with analytics adoption.
- A framework for evaluating vendors’ products.
- A framework for evaluating the healthcare industry’s adoption of analytics.
Learn more about subsequent steps in the Research Analytics Adoption Model from this complementary webinar: Powering Medical Research with Data: The Research Analytics Adoption Model.