Electronic Capture of Written Handoff Information: What Are the Next Steps?

Communication lapses during patient care transitions are reported to be frequent and may result in patient harm. The primary objective of our study was to assess the completeness, accuracy, and usefulness of our electronic handoff system to guide future software changes and educational interventions. We randomly selected and reviewed 707 of 2840 available handoff records generated on the medicine service of an academic medical center between August 1, 2012 and December 31, 2012. We used both quantitative and qualitative analytical techniques to characterize sign-outs in the following dimensions: completeness, usefulness and accuracy of information content, handoff task category, logic, internal consistency and appropriateness of assigned tasks, and composition and complexity of assigned tasks. The degree of completeness of information varied considerably across domains. Completeness was highest for entry of assigned tasks (99.9%), nearly as high for hospital course/presenting illness (95%), and relatively high (87%-98%) for entry of provider name and contact information, principal diagnosis, allergies, current clinical condition, mental status, and code status. Eighty-eight percent written handoffs described clinical condition and hospital course and whether there were tasks to complete. In 58% of suitable records, all problems listed in the electronic health record (EHR) were also present in the history of present illness. The accuracy of entered information also displayed wide variation. Only 80% of cardiovascular medications matched the contemporaneous EHR pharmacy record. Birth dates and allergies were identical in the handoff system and EHR in 95% and 86% of respective records. Of assigned tasks, 8% contained at least 1 unnecessary component or illogical/internally inconsistent element. Use of a handoff system, which organizes information entry through a standard template, promotes completeness of written handoff information. Inaccuracies in handoff data are associated with manual entry and should be discouraged. Programs should be encouraged to develop robust interfaces between the EHR and handoff platforms to promote entry of complete and accurate data and to enhance provider workflow.