Improving Patient Flow with Real-Time Capacity and Demand Matching

Carondelet St. Mary’s Hospital
Tucson, Arizona, USA


The team from Carondelet St. Mary’s Hospital is a participant in the IHI Learning and Innovation Community on Improving Flow Through Acute Care Settings.

Diana McBroom, MSN, MBA, Chief Nursing Officer, Project Sponsor
Patricia Hess, RN, BA, Director of Patient Care Services, Project Leader
Richard Radecki, RN, BSN, Senior Clinical Supervisor, Assistant Project Lead
Bill Loy RN, BSN, Emergency Center Supervisor, Flow Improvement Team Member
Frank Lopez, RN, BSN, Clinical Supervisor, Flow Improvement Team Member
Christine Miene, RN, BSN, Clinical Nurse Leader




By focusing on patient throughput challenges related to patient placement, St. Mary's Hospital will achieve the following goals for Fiscal Year 2009:

  • Emergency Center (EC) left without treatment (LWT) rate will be reduced by 34 percent to a LWT rate of 2.89 percent by June 30, 2009
  • Emergency Center length of stay (LOS) for admitted patients will be reduced to 73 percent of the current baseline of 531 minutes to 300 minutes by June 30, 2009
Left Without Treatment (LWT): Patients who sign into the Emergency Center are triaged by the nurse, and leave the Emergency Center before Medical Screening Exam is provided by the emergency physician, physician’s assistant, or nurse practitioner.
Length of Stay (LOS): Emergency Center patient arrival (sign-in) time to time of departure from the Emergency Center for all patients admitted from the Emergency Center.




  • Percentage of patients moved from the EC to inpatient bed within two hours [Goal: Greater than 90 percent]
  • Percentage of patients left without treatment [Goal: Less than 2.89 percent]
  • Emergency Center weekly median length of stay for inpatients [Goal: Less than five hours]
  • Average length of stay for inpatients (does not include behavioral health) [Goal: Less than 4.6 days]
  • Boarding burden comparison January through April 2008 vs. 2009


Carondelet St. Mary’s Hospital joined the IHI Flow Community in November 2008. We used the real-time demand/capacity process and tools, aligned our infrastructure as a “whole” system approach, and began small tests of change. The following tests of change were spread throughout the organization:

  • Predicting capacity, focusing on 8:00 AM–2:00 PM: Prior to leaving each day, Case Managers discuss with unit Clinical Nurse Leads (CNL) any patients with high potential for discharge the next day. This list is updated overnight by CNL in the electronic bed tracker. By 8:00 AM, potential discharges are verified in the electronic bed tracker by the CNL. The focus is on achieving a success rate greater than or equal to 80 percent. The current average house-wide success rate is 83 percent. This information is proactively used to match capacity to demand.
  • Predicting demand, focusing on 8:00 AM–2:00 PM: The demand is calculated using historical data and real-time data (using whichever number is larger). Again, our focus is on achieving a success rate greater than or equal to 80 percent.
  • Improve feedback communication, “closing the loop”: Feedback communication was expanded from next-day Leadership Meeting review to include electronic daily reports, weekly reports, and monthly FY08/09 comparison reports. This effort supports the shifting culture from compliance to commitment.
  • Differentiated role responsibilities of Emergency Center nurses: The Triage Nurse focuses on incoming patient assessments. The Patient Flow Nurse focuses on rapidly placing incoming patients in EC beds (Demand). The Zone Nurse focuses on 20 minutes to discharge after orders and 30 minutes to transport admissions when a bed is assigned. The Clinical Nurse Leader (CNL) focuses on identifying patients that can be discharged or transferred from the Emergency Center (Capacity).
  • Physician Assistant (PA)/Nurse Practitioner (NP) assigned to triage during peak patient flow times (12:00 PM–8:00 PM):  Replaces nurse in triage during peak hours. Eliminates patients being placed in EC beds if PA/NP determines they do not need to be seen in EC.
  • Streamlined the admissions bed request process in EC using EDTracker and Teletracker interface: This reduced disposition determination to bed request time by over 10 minutes by eliminating “middle person” (EC Clinical Nurse Leader).



Graphy_StMary'sHospital_AverageLengthStayforimpatients.jpgSummary of Results / Lessons Learned / Next Steps

Summary of Results:
As seen by the graphs of our results above, we have been able to impact LWT and EC LOS with sustainability. Our tests of change were done with innovation and creativity, not capital. We have begun to shift our culture to support an infrastructure of sustainability, integration, and communication. All of these elements will help us learn together and meet our community’s ever-changing needs.
  • Graphs 1, 2 and 3: The spike in January 2008 is considered to be seasonal since this is the busiest time of year for the hospital and we do not see the same spike in January 2009. This suggests that the real-time demand/capacity process has had a positive impact. 
  • Graph 4: A drop and sustained reduction in EC length of stay for inpatients occurred in April 2009 when process changes were implemented that led to a pull system being hardwired into the patient care processes on the inpatient units.
  • Graph 5: The average LOS for inpatients is not showing an impact from flow improvement initiatives yet. This is expected, for we are just beginning to drill down and address barriers. This measure, having the most variables impacting it, will take the longest to show improvement.
Lessons Learned:
To date our key lessons learned include the following:
  • A system-wide infrastructure must be established to support and sustain changes
  • Senior leadership support is necessary to ensure integrity of the work
  • Physician engagement is an integral piece in systems operations
  • Culture change is needed at front line to ensure a “pull” environment – inpatient unit ownership of patients is assigned, which helps create “commitment instead of compliance”
Next Steps:
Going forward we will continue to implement small tests of change to improve patient flow. We will continue to use Plan-Do-Study-Act cycles to monitor test of change success before implementing and hardwiring them into processes. With capacity and demand, we will be working to expand our test of change to 24 hours a day, five days a week. We will be refining our capacity accuracy down o specific patient successes and we will also work on refining our trending of measures to include “plan success.We understand this is not a process change but a journey we have begun.

Contact Information

Patricia Hess, RN, BA
Director of Patient Care Services
Carondelet St. Mary’s Hospital

Average Content Rating
(0 user)
Please login to rate or comment on this content.
User Comments