Identifying unnecessary duplicate genetic testing in a large medical center

Introduction Genetic testing has become ubiquitous in clinical medicine and plays an important role in making diagnoses and guiding treatment plans. Indiscriminate use of these tests can cause duplicate testing, which is typically not indicated because results from repeated constitutional molecular genetic testing should not change over time. Thus, duplicate genetic testing often represents inappropriate test utilization that can contribute an unnecessary burden on the laboratory and health care system. Objective The purpose of our study is to determine the incidence of duplicate testing of in-house genetic testing offered at a large medical center, which includes cystic fibrosis, factor V Leiden, prothrombin G20210A, methylenetetrahydrofolate reductase (MTHFR) C677T, and MTHFR A1298C, and to develop a tool to identify and block duplicate testing. Methods We retrospectively analyzed internal laboratory databases of all cystic fibrosis (n = 36164), factor V Leiden (n = 3264), prothrombin G20210A (n = 2890), MTHFR C667T (n = 1451), and MTHFR A1293C (n = 1290) testing performed at the molecular pathology laboratory at a large medical center from either January 2010 or January 2014 to July 2019. We analyzed and cleaned the databases with the R programming language, and we developed a prototype web-based app to proactively identify duplicate test requests with the Shiny R package. Results From January 2010 to July 2019, 3535 (9.8%) of the 36164 cystic fibrosis tests performed were duplicate tests for 3257 unique patients. Of these duplicate cystic fibrosis tests, 2997 were repeated once in the same patient, 244 were repeated twice in the same patient, 14 were repeated three times in the same patient, and 2 were repeated four times in the same patient. From January 2014 to July 2019, 99 (3.0%) of the 3264 factor V Leiden tests, 86 (3.0%) of the 2890 prothrombin G20210A tests, 49 (3.4%) of the 1451 MTHFR C667T, and 46 (3.6%) of the 1290 MTHFR A1298C tests performed were duplicate tests. We developed a proof-of-concept Shiny web-browser app that provides a user-friendly interface to determine if a patient has been previously tested in the molecular pathology lab. This app operates locally on a laboratory computer and uses spreadsheets automatically exported from the electronic medical record. These features allow for the app to be deployed quickly without needing to be integrated into the electronic medical record. Conclusions The results of this study indicate that unnecessary duplicate testing represents a small but significant proportion of genetic testing performed by the molecular pathology laboratory. Duplicate testing occurred more frequently with cystic fibrosis testing, which reflects its high volume at the medical center. Deployment of web-based apps using Shiny can provide straightforward and efficient tools for reducing duplicate tests.