University of Pittsburgh

PULM-TREE: A New Tool for Lung Transplant Research

University of Pittsburgh investigators are developing novel research software, PULM-TREE, designed to store and integrate the clinical and genetic data of lung transplant recipients (LTRs) from multiple sources into one easy-to-use platform. 

Description

Understanding and assessing clinical and genetic data can allow for an understanding of the risk for transplant rejection promoting early intervention and improved patient outcomes. Currently, much of the data are stored in multiple databases and requires manual extraction for each patient. This novel agile software will extract and combine data in its native form in one database improving research results and patient outcomes.

Applications

• Lung transplant research
• Other research

Advantages

Current data from lung transplant recipients are stored in several registries, and collating data sets for research purposes is currently a laborious process. Demographic and biospecimen data are stored in the Lung Transplant Biospecimen Registry (BREATHE-LT). Electronic Health Records (EHR) contain health data including medication use, pathology reports and mortality, and separate databases contain information on whole genomic sequencing and biospecimen inventories.

This novel software aims to consolidate all data from LTRs into one repository, PULM-TREE. Data will be extracted and stored in its native format and will have sharing capabilities, allowing for use in multiple research projects. Using an agile software development process (SCRUM) functionality can be delivered rapidly in small pieces, building to a large repository over time, as opposed to waiting for full development before roll-out. Data can be displayed on a timeline and include rapid access to additional information including microbiology data to enhance monitoring of pulmonary function, identify patients at risk of organ rejection, and be used as a large database for research purposes.

Invention Readiness

PULM-TREE has been developed and written in Python 3.9.7 using a variety of supporting modules including NumPy, Openpyxl and docx to combine data from multiple sources. It has four modules: data extraction, user interface, business logic that supports data harmonization, and a patient management tool. A system containing 720 patients, 20,818 lab results, 9,219 pathology reports, and 21,441 microbiology findings was tested which identified 270 patients with potential Chronic Lung Allograft Dysfunction (CLAD). Development is ongoing to include supporting research projects and standardized terminology.

IP Status

Software