I've said it before: The ED is the front door to the hospital. And the Agency for Healthcare Research and Quality (AHRQ) agrees. The agency noted that the ED is the entry point for more than half of all patients admitted to the hospital in the United States. (“Hospital Admissions that Began in the Emergency Department.” AHRQ: Healthcare Cost and Utilization Project. Statistical Brief 1, February 2006; www.hcup-us.ahrq.gov/reports/statbriefs/sb1.pdf).
That statistic means 40 visits are made to the ED for every 100 citizens each year. An astonishing 43 percent of patients present to the ED with moderate to severe pain, and almost 40 percent of patients wait more than an hour to see a physician. A quarter of the patients seen in the ED spend more than four hours there, and another 40 percent of patients present with complaints of chest or abdominal pain with potentially serious diagnoses to be uncovered. (“National Hospital Ambulatory Medical Care Survey: 2006 Emergency Department Summary.” National Health Statistics Reports; No. 7. Hyattsville, MD: National Center for Health Statistics. 2008; www.cdc.gov/nchs/data/nhsr/nhsr007.pdf.) Yet the ED continues to be a place of inefficient, unsatisfactory, and often unsafe health care encounters.
Banner Health, an organization with hospitals around Phoenix, recently began a new process to improve intake and patient flow at eight of its hospitals. The project was sophisticated from the start, involving many ideas championed in this column for the past four years. Queuing theory, demand capacity management, patient segmentation, keeping patients vertical, and medical teams are all part of this new intake model implemented at Banner Health. Banner dubbed its model “D2D SPF,” connoting door-to-doctor split-patient-flow. The providers agree to split the ED at input. Sicker patients based on ESI scoring are ushered to beds in the usual fashion, but less acute patients are kept vertical, and moved through the system in a clinic-like fashion, not occupying a room.
Banner uses a queuing model that requires input of census and acuity to determine how many providers are required in each area to match capacity to demand. The beauty of their model is that they have adapted it to low- and high-volume departments. The nurse does a “quick look,” and assigns the patient to a patient stream based on ESI. ESI 1 and 2 patients are placed immediately in a bed and assessed by a physician, nurse, and technician, and the workup is articulated by the physician for the team. ESI 3, 4, and 5 patients are taken to an intake area smaller than a traditional ED room for an assessment by a physician and a nurse, and then testing or treatment occurs from the results waiting area; the patient never occupies a bed. There can be movement from one stream to the other depending on the patient's condition and results. The details of this process and all the tools necessary for implementation are posted on the Banner Health web site:www.bannerhealthinnovations.org/DoortoDoc/About+D2D.htm.
The Banner model is different from most traditional EDs because patients who are not that sick do not own a bed; instead they move among treatment areas as they would in a clinic. They are not undressed, remain vertical as much as possible, and wait for their lab and other results outside the flow. The medical assessment by the nurse and physician occurs simultaneously. This change in the intake process and patient flow has the advantage of keeping these patients, the vast majority, flowing rather than waiting in the lobby for an initial assessment during busy times, and it removes the bed as the capacity limiting factor.
Preliminary results using this model are very promising. Implementation of D2D SPF has decreased the time from patient arrival in the ED until seen by a physician, which in turn reduced the rate of patients leaving without treatment by 30 percent to 60 percent across eight busy EDs. Tools developed for implementing this process include queuing network analysis, which helps capacitate different facilities for this new model based on patient volume and acuity.
Innovation like this at the front lines with assistance from those familiar with queuing models and the principles of demand capacity management is exactly what we need for solving our flow problems today.