ETL vs ELT: How Your Data Is Loaded in a Database Makes All the Difference in Legacy EHR Data Archiving Solutions

Extract, Transform, Load (ETL)

The difference between the two strategies depends on when the data transformation takes place. Here are the steps In the ETL method that MediQuant uses:

1. Extract: Data is extracted from the target source system(s).
2. Transform: The extracted data is transformed into a standard format based on the data type being migrated. This may involve data cleaning, filtering, aggregating, verifying, and applying business rules.
3. Load: The transformed data is loaded into DataArk where users can readily access information within normal workflows and across data types.

Transforming extracted data before loading it into an archive makes all the difference. It may take a little more time and cost a little more money, but the result is a robust archive of uniform data that can be accessed individually or in conjunction with current operational systems.

For example, clinicians can look at current and historic data at a glance in one view.

Advantages of ETL include:

• Assurance that only clean and validated data is stored in the data warehouse
• Reduced load on the data warehouse during queries
• Assurance that data privacy and compliance requirements are met before data is loaded

Extract, Load, Transform (ELT)

The typical ELT engagement includes these steps:

1. Extract: Data is extracted from various source system(s).
2. Load: The extracted data is loaded directly into the target data warehouse.
3. Transform: Transformations are performed on the loaded data within the data warehouse environment.

While it is quicker to move information to a data warehouse before the transformation process, it still ties health systems to legacy software because the data is likely in a proprietary format. In these cases, APIs are required to access a particular data type for a particular software system. Multiply the number of archived systems by the number of current systems, and IT staff could be maintaining hundreds, if not thousands, of APIs. Given the current state of cyberattacks, health systems should be reducing the number of technology connections — not increasing them.

Why ETL is the Way to Go for Legacy Data Archiving Solutions

Data is the new currency in healthcare, both clinically and operationally. With the rise of artificial intelligence, machine learning, and other advanced data analysis methods, your health system’s EHR archive data continues to have value long after the diagnosis is made or the patient discharged.

A few short years ago, health data retention was based on health data compliance needs that can vary by state. Now, however, the difference is how usable data can remain long after a particular legacy system has been retired.

Here are five reasons why MediQuant’s unique approach to data normalization and access benefits clinicians, IT staff, and the C-suite:

1. Maintaining the data schema of each legacy system continues to lock your organization to the old legacy system, requiring APIs to provide access that varies in quality and completeness. With MediQuant, clients receive, over time, a consolidated data set(s) with an independent schema the client owns and understands. This removes the need to maintain legacy system knowledge in order to interact directly with archived data. MediQuant’s approach gives clients the ability to access datasets directly from DataArk, where information is stored in one schema for each data type, allowing advanced data searches from one location.
2. DataArk is not a static archive that’s read-only. It’s a healthcare active archive that is dynamic, changing as updated information is received. For example, revenue cycle personnel can continue to work down accounts receivable among archived files. When a collection is made, the data in DataArk updates.
3. Since ALL data of a particular type is stored centrally, users can access the information they need via a single sign on through normal workflows.
4. MediQuant provides one simple UI sitting on top of one well-documented, well-tested and mature data schema, a simple and elegant approach that brings together years of health information systems technology in one platform.
5. MediQuant continues to innovate, experimenting with AI strategies to enable significant improvements to our solutions. However, AI can’t manage the wide variation between legacy data schemas that exist in ELT databases.