Research/Academia

Accessing the LLEAP De-identified Research Database:

The de-identified database provides data set to be used for research purposes. This project grew out of a need for access to raw data for research. The raw data can be explored using tools including Tableau, SPSS, R and Python. It is designed to be HIPAA compliant, meaning no identifiable patient attributes are included and all dates are shifted.  Thus, use of the dataset does not require IRB review or approval (IRB# 5190079).  It is built on top of Loma Linda's Enterprise Data Warehouse (EDW) and only contains Loma Linda data at this time.  This project provides the basic framework and core data set that includes:

References:

• Instructions on extracting and analyzing the dataset using SPSS
• De-Identified Database Overview – a short video demonstration of basic steps (mp4)
• Deriving a De-Identified Dataset from Epic for Clinical Research
• LLEAP Research Database - Journal of Translational Science 2020 publication
• Research Database IRB Determination

SlicerDicer

SlicerDicer offers self-service access to clinical data on customizable populations of patients as well as powerful data exploration abilities. SlicerDicer’s intuitive interface enables clinicians to investigate data, refine or reformulate their searches, and quickly test conjectures to better understand their patient populations without requiring much training.

SlicerDicer helps clinicians take reporting into their own hands and reduce the need to send requests to report writers before they can see interesting data. With SlicerDicer, physicians can analyze wide-ranging queries such as comparing the effectiveness of two courses of treatments for patients with a certain diagnosis or evaluating possible side effects associated with a medication. They can also identify patient cohorts for research purposes. Clinicians can review the results for their queries and then make changes to queries on-the-fly to refine those results and hone in on meaningful information.

In SlicerDicer, clinicians begin by selecting a patient population, and then create queries by:

• Filtering the patient population by using a variety of criteria such as demographics, diagnoses, lab results, procedures, and current medications.
• Comparing patient populations that match different sets of criteria to see, for example, trends and correlations in different populations and investigate potential contributing factors.
• Tracking trends in patient populations over time to compare, for example, patient outcomes.

Individual patient records gathered should only support direct patient care by those who have a treating relationship with the patient.  IRB approval MUST apply for any research use of the data.

Training for SlicerDicer is available through the Self-Tutorial in your LLEAP workspace environment.  In the search field, type SlicerDicer.  You may also have SlicerDicer on the main menu bar. Quick Start Guide.

If you are involved in a research project that requires broader access to patient information that is not normally granted in Epic, here is the current process to get the underlying data.

1. Researcher performs search in SlicerDicer and saves SlicerDicer session
2. Researcher submits a Data Intake Request with the name of the SlicerDicer session for underlying data and provides supporting Institutional Review Board (IRB) documentation
3. Data Governance solicits review/approval from IRB office
4. If approved, your request will be assigned to one of the analysts to export the data on behalf of the requestor / researcher
5. If the research project will need reoccurring data extracts, a Reporting Workbench and/or Clarity Report may be created to fulfill your data needs.