The Problem
The collection and review of medical records for clinical transplantation evaluation and submission of required UNOS forms requires immense resources to extract information from internal and external health care facilities and providers. This work to create a transplant referral summary or to fill and upload regulatory forms is largely a manual process often taking hours per patient and is redundant in that it must be performed by multiple members of the transplant mutlidisciplinary team.
Our Solution
The proposed solution is to implement natural language processing/large language models to extract annotated defined data elements to streamline the processes and reduce duplication of effort to perform the initial transplant evaluation. The same process will be developed and utilized to fill specific UNOS forms for candidates placed on the waiting list and patients undergoing transplantation.
Anticipated Impact
Use of technology solutions will decrease time spent by each multidisciplinary team member for patient evaluation without diminishing the quality or completeness and improve the clinician/staff satisfaction. The application of technology and utilization of available APIs for UNOS form submission will further improve efficiency in achieving regulatory requirements. Time saving translates to increased capacity for new evaluations/ transplants, increasing throughput without increasing personnel.
