Presentation
Work System Factors Influencing Medication-Related Discharge Processes in an Orthopedic Unit
SessionHC6: Clinician Workloads
DescriptionHospital discharge is a high-stakes transition where medication errors are common, yet little is known about how work system factors affect nurses' ability to coordinate and communicate medication instructions to patients. Using the SEIPS human factors framework, this study observed 16 nurse-patient discharge interactions in an orthopedic unit at a large academic medical center in Southern California. Researchers identified 468 contributing work system factors across four primary categories: communication (31%), coordination (28%), tools/technology (20%), and operational processes (15%). Patients were discharged on an average of three new medications, with one patient managing 24 total medications at discharge.
As part of a pilot analysis, AI-assisted nano-coding was applied to the tools/technology category. Seven refined nanocodes were developed, with discharge instruction assembly, medication payment visibility, and instruction clarity emerging as the most frequent challenges. AI classifications required human expert review and refinement to achieve sufficient granularity, though AI and human coders ultimately agreed in 88% of cases. These findings underscore that discharge is a complex, technology-mediated workflow requiring systems-level redesign to reduce medication safety risks.
As part of a pilot analysis, AI-assisted nano-coding was applied to the tools/technology category. Seven refined nanocodes were developed, with discharge instruction assembly, medication payment visibility, and instruction clarity emerging as the most frequent challenges. AI classifications required human expert review and refinement to achieve sufficient granularity, though AI and human coders ultimately agreed in 88% of cases. These findings underscore that discharge is a complex, technology-mediated workflow requiring systems-level redesign to reduce medication safety risks.
Alternate Presenter
Event Type
Lecture
TimeThursday, October 22nd3pm - 3:20pm PDT
Location
Similar Presentations