The case of Inova Mount Vernon hospital, a small, 237-bed hospital in Mount Vernon, VA, presented in the March 2011 edition of Quality Progress provides some useful insights for hospital quality management. The hospital used several quality methods to reduce its average length of stay for discharge home patients within the emergency department (ED) from 266 minutes to 135 minutes. The hospital used abbreviated kaizen events, value-stream mapping, spaghetti diagrams, metric definitions with regular reporting, brainstorming and control charts to facilitate the performance improvement.
Some concrete steps that were taken after gathering relevant data and identifying the problem are as follows:
1) The project members realized that metrics related to key time elements within the ED process were not being tracked. The hospital implemented an ED tracking system (using Picis software) in November 2008 to track key information about patients. This include information regarding chief complaints, acuity level, diagnoses and patient dispositions. The software was also used to track process steps such as registration time, triage time, time for patient to arrive in the ED room, time for medical provider to first visit the ED room, time for medical provider to make disposition on patient, time from disposition to patient leaving ED (to be released to go home, admitted to another hospital room or transferred to another hospital).
2) This time tracking was done at the level of individual doctor, registered nurse or shift management coordinator. The required trust was built with the concerned hospital staff so that they realize that the objective is not to punish but to understand the process. The helped in gaining confidence and obtaining cooperation from all members of the hospital.
3) The data were downloaded daily, and the results were posted and distributed to key managers and management coordinators. With the data, the hospital began to understand the parts of the process that need fixing and root cause analysis was done to devise corrective actions.
4) With baseline metrics in place and an understanding of the vision and mission, the ED, floor units, and other departments met to identify the critical needs in the ED. To achieve consensus, all parties maintained focus on the needs of ED patient.
The following six steps highlight the specific action taken by the hospital to reduce the length of stay:
i) Streamline the triage and bedding process:
The first change activity focused on the front-end processes of getting a patient triaged and into a bed, as shown in the highlighted boxes in the emergency department (ED) process flow in the figure below. Spaghetti diagrams drawn using a floor layout were used in this step to better understand patient, registration and ED staff movement. This helped the team visualize excess movement between areas, and facilitated moving some process steps closer to each other to remove wasted staff and patient movement.
Because the ED nurses and technicians were involved in creating the new process, buy-in was readily obtained. The team decided to modify the triage process and only have a quick triage prior to moving a patient to a room—after the patient was registered and given a hospital identification wristband. This quick triage involves obtaining the patient’s vital readings, chief complaints and initial acuity level. The initial acuity level is the emergency severity index that stratifies patients into five groups—from one (most urgent) to five (least urgent) on the basis of seriousness and clinical staff resource needs. The patient then is walked back to a room assigned by the charge nurse, and the full triage is completed later by the assigned nurse. The doctor is given access to the patient as soon as the patient arrives in the room. The time-to-doctor metric ultimately dropped by 10 minutes as a result of this change. It also helps reduce instances in which patients are sent to the waiting room and left there for extended periods.
After the triage changes were made, the team discovered that the ED doctors often were seeing patients before the assigned nurse completed the full triage. This was a cultural change for some nurses, but ultimately they came to appreciate some of the benefits. One obvious benefit was that the nurse could obtain the doctor’s orders more quickly. The bedside nurse could combine the nurse assessment and triage (as shown in the shaded boxes in the figure presented above) with execution of the orders—for example, blood draws—and accelerate the care of the patient. Some rules had to be established based on the ED activity, and these were shared and agreed to by ED management after the kaizen event. The partial up-front triage will be in effect whenever the ED has more than two of the 18 ED rooms available for more acute patients, regardless of staffing. When the ED is full (two or fewer rooms available), the triage nurse returns to performing complete triages and getting patients to rooms when they become available. Advance triage protocols (ATP) also were established to allow the triage nurse to get some initial orders started so results are obtained before the doctor sees the patient or shortly after the patient arrives in the room.
The overall performance gain from the changes initiated in step one was more than 20 minutes.
(ii) Streamline discharge-to-home process:
The portion of the ED process from doctor disposition to patient discharge also was addressed early in the improvement endeavor. All nurses and charge nurses were trained to review the ED status board regularly. The ED doctors also were charged with verbally calling out whenever they discharged a patient to go home. This would start the ball rolling more quickly to get the patient their discharge instructions. Individual nurse performance on the disposition-to-discharge times is tracked regularly, and the comparative times are posted openly for review. Coaching is instituted when necessary. The goal was to complete the patient discharge instructions and get them on their way within 20 minutes of disposition. Total time savings from this change has been about 10 minutes. An additional six-minute improvement will be necessary to hit the 20-minute goal for this step of the process.
(iii) Measure Length of Stay for doctors:
The ED data-capture system allowed the hospital to easily gather information on all the key metrics in the ED, including the time from when the doctor first sees a patient to time of patient disposition. Each doctor’s disposition time was monitored and graphed monthly. The disposition time by doctor also was shared in a daily report so we could investigate when someone had either a slow or fast day. The team looked at patient acuity, but the one area where the team noticed longer length of stay (LOS) times were with patients with stomach or chest pains and patients being treated for alcohol and substance abuse. Over time, patient loads by doctors by acuity seemed to even out, so the monthly charts are good indicators of comparative performance. The monthly charts are used by the ED medical director to counsel ED doctors who are slow on their dispositions. The key is charting performance by doctors and understanding which doctors need training and coaching. For example, slower-performing doctors would spend additional time double charting patient notes. When counseled, the doctors—being competitive, as expected—tend to change their patterns and improve their performance. Faster doctors provide small, informal seminars for the group and share techniques they use to increase speed. The total time saved from step three was about 35 minutes as presented in the following figure:
(iv) Implement an ED fast track for less-acute patients
Early in the improvement process, ED leadership recognized that lower acuity and less complex patients were spending more time in the ED than their conditions warranted. If a fast-track process (manned by a nurse practitioner) was established for this type of patient, patients could be treated in the ED more quickly and the ED could focus more resources on more severely ill patients. This process would take a little longer due to the need to hire nurse practitioners and adjust nurse and tech staffing. The hospital also needed to renovate the ED’s registration and triage areas to accommodate the fast-track staff and patients.
The hospital carved out space for this area and added a nurse practitioner desk and one ED bed for exams. Space also created in the cast room for ED staff to treat less-acute patients and apply casts and splints. This room is not included in the 18 dedicated to acute patients, but is one of the 20 in the ED. It is across the hall from the fast track and can be used as a second fast-track bed because it generally has low use. In addition, a family waiting room was commandeered and four recliners were added to accommodate fast-track patients. The entire fast-track space was developed with little capital, but is considered temporary. The hospital has plans to complete a more extensive renovation and expansion of this function using space from the larger patient waiting room.
The fast-track approach achieved its goals of enabling less-acute patients to be seen more quickly and dispositioned. Use of the patient waiting area dropped significantly, and LOS times for the fast-track patients have stabilized at about 90 minutes. Another advantage is that the number of patients leaving without being seen was reduced by 75% (as shown in the figure below), which actually added to the total revenue gathered by the ED. The impact to overall LOS from step four was a reduction of more than 15 minutes.
v) Implement bedside triage
The team organized a kaizen event with value stream mapping to see what further improvements could be made in the ED process. The goal was to get the patients to the doctors as quickly as possible. With that in mind, the team decided to look at revising the triage process to bring patients directly to the ED rooms and complete the quick and full triage at the bedside. The team thought this would get patients to a room quickly and that the doctors could see them more quickly. This value stream change resulted in a decrease of 10 minutes to inpatient LOS. The following figure shows the overall LOS metric plotted daily on a control chart to keep focus and allow awareness of when to take actions.
The several hurdles to overcome in this step mostly related to equipment. The hospital had to ensure all patient rooms had the necessary working equipment to perform triages and make the necessary inputs into a room computer. Triage nurses also had to be retrained to make their first call to the charge nurse, who would then direct them to the proper room where they would begin their quick triage.
Teamwork also played a big part in this success. If the triage nurse is bringing a patient to a room and another patient is ready to go to another room, the charge nurse or other free nurse handles the second patient. When the ED becomes busy and all rooms are occupied, the process must revert back to full triage at the front triage station. This is OK because there will be no rooms to place the patients. After step five reduction of the discharge LOS dropped another nine minutes as a result of the bedside triage process.
Source: Watson, R. Q., Leeson, K. 2011. On the Clock: Hospital uses quality methods to eliminate delays and move patients along. Quality Progress March 2011, 44-50 (and on-line supplement)