BACKGROUND: As part of Yale-New Haven Hospital (Connecticut)'s Safe Patient Flow Initiative, the physician leadership developed the Red/Yellow/Green (RYG) Discharge Tool, an electronic medical record-based prompt to identify likelihood of patients' next-day discharge: green (very likely), yellow (possibly), and red (unlikely). The tool's purpose was to enhance communication with nursing/care coordination and trigger earlier discharge steps for patients identified as "green" or "yellow." METHODS: Data on discharge assignments, discharge dates/ times, and team designation were collected for all adult medicine patients discharged in October-December 2009 (Study Period 1) and October-December 2011 (Study Period 2), between which the tool's placement changed from the sign-out note to the daily progress note. RESULTS: In Study Period 1, 75.9% of the patients had discharge assignments, compared with 90.8% in Period 2 (p < .001). The overall 11 A.M. discharge rate improved from 10.4% to 21.2% from 2007 to 2011. "Green" patients were more likely to be discharged before 11 A.M. than "yellow" or "red" patients (p < .001). Patients with RYG assignments discharged by 11 A.M. had a lower length of stay than those without assignments and did not have an associated increased risk of readmission. Discharge prediction accuracy worsened after the change in placement, decreasing from 75.1% to 59.1% for "green" patients (p < .001), and from 34.5% to 29.2% (p < .001) for "yellow" patients. In both periods, hospitalists were more accurate than house staff in discharge predictions, suggesting that education and/or experience may contribute to discharge assignment. CONCLUSIONS: The RYG Discharge Tool helped facilitate earlier discharges, but accuracy depends on placement in daily work flow and experience.
BACKGROUND: As part of Yale-New Haven Hospital (Connecticut)'s Safe Patient Flow Initiative, the physician leadership developed the Red/Yellow/Green (RYG) Discharge Tool, an electronic medical record-based prompt to identify likelihood of patients' next-day discharge: green (very likely), yellow (possibly), and red (unlikely). The tool's purpose was to enhance communication with nursing/care coordination and trigger earlier discharge steps for patients identified as "green" or "yellow." METHODS: Data on discharge assignments, discharge dates/ times, and team designation were collected for all adult medicine patients discharged in October-December 2009 (Study Period 1) and October-December 2011 (Study Period 2), between which the tool's placement changed from the sign-out note to the daily progress note. RESULTS: In Study Period 1, 75.9% of the patients had discharge assignments, compared with 90.8% in Period 2 (p < .001). The overall 11 A.M. discharge rate improved from 10.4% to 21.2% from 2007 to 2011. "Green" patients were more likely to be discharged before 11 A.M. than "yellow" or "red" patients (p < .001). Patients with RYG assignments discharged by 11 A.M. had a lower length of stay than those without assignments and did not have an associated increased risk of readmission. Discharge prediction accuracy worsened after the change in placement, decreasing from 75.1% to 59.1% for "green" patients (p < .001), and from 34.5% to 29.2% (p < .001) for "yellow" patients. In both periods, hospitalists were more accurate than house staff in discharge predictions, suggesting that education and/or experience may contribute to discharge assignment. CONCLUSIONS: The RYG Discharge Tool helped facilitate earlier discharges, but accuracy depends on placement in daily work flow and experience.
Authors: Ayan Sen; Yan Xiao; Sun-Ah Lee; Peter Hu; Richard P Dutton; James Haan; James O'Connor; Andrew Pollak; Thomas Scalea Journal: J Trauma Date: 2009-03