BMJ Open Quality (Dec 2024)
Bundled de-implementation of recurring hospital orders with a novel electronic medical record order set
Abstract
Introduction High-frequency recurring orders placed through the electronic medical record (EMR) may contribute to unnecessary care in hospitalised patients. This quality initiative sought to develop and pilot test a hospital order set for bundled review and de-implementation of common recurring orders.Methods A voluntary-use EMR order set was developed to display low-frequency order alternatives for common hospital care components. The order set was introduced to hospitalists at a large academic hospital from February to June 2023. Orders for overnight vital signs, tubes/tethers (a composite of telemetry, continuous pulse oximetry and Foley catheter) and daily labs (a composite of complete blood counts and metabolic panels) were monitored twice weekly in hospitalised patients at low risk for clinical decompensation from December 2022 through June 2023. Paired t-test was used to assess for differences in order frequency before and after order set introduction.Results The order set was used in 48 unique encounters to place 80 de-implementation orders, most commonly for discontinuation of overnight vital signs (n=37). Two or more de-implementation orders were placed during 44% of order set encounters. Mean (SD) total high-frequency orders decreased by 0.22 per patient day (95% CI −0.39 to −0.06; p=0.010) after order set introduction, driven by a reduction in overnight vital sign orders of 0.17 per patient day (95% CI −0.23 to −0.12; p<0.001). There was no statistically significant difference in orders for tubes/tethers or daily labs before and after order set introduction.Discussion Introduction of a novel order set for bundled review and de-implementation of recurring orders was associated with reduced high-frequency recurring orders in hospitalised patients, driven by a reduction in overnight vital signs. Nearly half of order set use was for two or more de-implementation orders, suggesting that bundling de-implementation orders may be an efficient way to reduce unnecessary orders.