Sustainable Mechanism to Reduce Emergency Department (ED) Length of Stay: The Use of ED Holding (ED Transition) Orders to Reduce ED Length of Stay.

OBJECTIVE: The objective was to evaluate the effect of an emergency clinician-initiated "ED admission holding order set" on emergency department (ED) treatment times and length of stay (LOS). We further describe the impact of a performance improvement strategy with sequential plan-do-study-act (PDSA) cycles used to influence the primary outcome measures, ED LOS, and disposition decision to patient gone (DDTPG) time, for admitted patients.

METHODS: We developed and implemented an expedited, emergency physician-facilitated admission protocol that bypassed typical inpatient workflows requiring inpatient evaluations prior to the placement of admission orders. During the 48-month study period, ED flow metrics generated during the care of 27,580 admissions from the 24-month period prior to the intervention were compared to the 29,978 admissions that occurred during the 24-month period following the intervention. The intervention was the result of an in-depth, five-phase PDSA cycle quality improvement intervention evaluating ED flow, which identified the requirement of bedside inpatient evaluations prior admission order placement as being a "non-value-added" activity. ED output flow metrics evaluating the admission process were tracked for 24 months following the intervention and were compared to the 24 months prior.

RESULTS: The use of an emergency physician-initiated admission holding order protocol resulted in sustainable reductions in ED LOS when comparing the 2 years prior to the intervention, with median LOS of 410 (interquartile range [IQR] = 295 to 543) and 395 (IQR = 283 to 527) minutes, to the 2 calendar years following the intervention, with the median LOS of 313 (IQR = 21 to 431) and 316 (IQR = 224 to 438) minutes, respectively. This overall reduction in ED LOS of nearly 90 minutes was found to be primarily the result of a decrease in the time from the emergency physician's admitting DDTPG times with median times of 219 (IQR = 150 to 306) and 200 (IQR = 136 to 286) minutes for the 2 years prior to the intervention compared to 89 (IQR = 58 to 138) and 92 (IQR = 60 to 147) minutes for the 2 years following the intervention. It is notable that there was a modest increase in the door to disposition decision of admission times during this same study period with annual medians of 176 (IQR = 112 to 261) and 178 (IQR = 129 to 316) minutes, respectively, for the 2 years prior to 207 (IQR = 129 to 316) and 202 (IQR = 127 to 305) minutes following the intervention.

CONCLUSIONS: We conclude that the use of emergency physician-initiated holding orders can lead to marked reductions in ED LOS for admitted patients. Continued improvement can be demonstrated with an effective performance improvement initiative designed to continuously optimize the process change.