Lessons learned in migrating from one commercial genetics decision-support tool to another: Assessment of data integrity and utilization readiness.

Abstract

e23232 Background: Rapid advancements in health information technology have greatly influenced how clinicians engage with patient health data. In the context of clinical genetics, the electronic health record (EHR) often contains a wide array of data to identify patients at elevated risk for hereditary cancer syndromes. However, these data often exist in multiple forms and disparate locations, which in the presence of numerous commercial EHR solutions, results in significant variations in how institutions codify genetic data. Furthermore, within an institution, there is often internal pressure to migrate from one support tool to another, necessitating a focus on data integrity during such a transition. Methods: Fred Hutchinson Cancer Center (Fred Hutch) is a large academic cancer care center associated with the University of Washington. Between July and December 2023, the Fred Hutch Clinical Cancer Genetics and Genetic Counseling Service migrated their family history database from commercial pedigree software, Progeny, to another commercial pedigree and decision-support tool, CancerIQ. Following completion of the data migration on more than 13,000 patient records, study team planned to review a total of 500 randomly selected patient charts in both support tools. Each chart was reviewed to determine: (1) the presence of genetic intake information, (2) the presence and results of genetic testing, if available, and (3) whether the patient met National Comprehensive Cancer Network (NCCN) criteria for testing. A chart was noted to be concordant between the two support tools if there was agreement in all three of the above. Results: So far, a total of 240 patient charts were reviewed between the two support tools. Of these, 226 (94.2%) were noted to be fully concordant. Of the 14 charts that were noted to be discordant, the distribution of the discordant data was: 5 charts differed on whether genetic testing had been ordered, 5 charts differed on the results of the testing (often omitting one or more variants), and 4 differed on whether the patient met NCCN criteria for testing. Conclusions: At our institution, there was high data integrity following migration from one commercial and pedigree-generating software solution to another. While these results can ease clinician concerns following such support tools transitions, it also highlights areas for improvement in how family and patient genetic data is recorded.