Executive Summary – Patient Flow Improvement Cuts Length of Stay (LOS)

As a management consultant, I worked with a client hospital emergency department to reduce Average Length of Stay by 10%. Specific improvements implemented include patient arrival time and volume modeling, aligned clinical team scheduling, standardized nurse / doctor communication practices, and innovative charge-nurse-esque patient outflow coordinator.

As a management consultant I completed a wide range business improvement projects with world class companies in tons of industries. I have profiled a few of my projects here on RickMaher.info, but out of respect for my clients confidentiality, I have excluded their names and have changed some material facts and figures.

Situation

My client, was a health system that operated three hospitals each of which with one sizable main facility and numerous satellite facilities. This client was focusing on a growing trend in hospital operations – patient flow – or the optimization of patient movement through the various stages of care in a hospital in order to optimize the quality of care in tandem with hospital bed utilization. Under this focus, projects were conducted at each location, and scoped to include significant projects within many discrete departments at each location. This profile will highlight the work that I did in the adult emergency department (aka emergency room) of an 800 bed (large), university medical school, teaching hospital.

The emergency department was faced with growing demand due to community population growth and aging, and was struggling to maintain their high service level standards. In short, patients were experiencing longer wait times, and reporting reduced satisfaction in the quality of care and attention they received. As the only level I trauma center within reasonable commute, the department was also facing pressure from the state department of health to ensure constant available capacity in case of a sudden surge in trauma patients.

A capacity expansion project was being planned, but would not be ready for ~3years due to the need for a significant construction project to expand the physical space.

Task

In a management consulting role, I was brought in to identify and implement process and management improvements to increase the capacity of the emergency department within its current physical space and set-up. Further, the client insisted that patient care not only must be maintained, but indeed must be improved along with the throughput improvements. Finally, while the client and the university medical school generally worked well together, the doctors were not employees of the client and therefore were identified as a gray area in terms of in/out of scope. The core group of project stakeholders included hospital management and clinical support staff (e.g. nurses, techs, transporters, etc.).

Action

As with all of my projects, my first step was to get to know the key people, and develop a feel for the culture. The culture itself however, was absolutely unlike any other. If you’ve ever thought the chaos shown in emergency rooms TV must be dramatized, you’d be wrong. The level of activity and movement was every bit as fast and varied as I’d seen on TV, and I saw things that I will never forget – ranging from funny to scary and sad to joyous. Fairly quickly though, I realized that the level of activity was not at all chaos, most of the time it was just fast and varied. This was a key learning in getting to understand the culture and identifying opportunities to improve patient care.

Patient In-flow Patterns

In the process of getting to know the people in the department, I heard them say over and over, “you never know what is going to come through the door.” And while I came to understand their sentiment, and agree that the clinical details and demographics of patient was extremely varied, I looked to the data to uncover the real truth. In fact, as with most businesses, customer trends (in this case patient trends) are quite predictable. I was able to identify that patient arrival volumes and times were fairly predictable based on the time of day, the day of the week, and the week of the year. Further, while the clinical specifics of the patients were not very predictable, the acuity (the health severity and the amount of care resources required) of patients, by definition, was fairly predictable.

This was the first half of the story – by understanding when patients typically arrived, I was able to develop an improved staffing schedule to ensure that there was appropriate resources to address patients and prevent a significant backlog. We implemented this quickly, but saw almost no improvement.

Patient Outflow – The Real Root Cause

It became very clear that the inflow was only half of the equation, and that the outflow of patients from the emergency department was the real root cause of the long wait times. A common performance metric of hospitals and emergency departments is Length of Stay (LOS) or Average Length of Stay (ALOS). This metric is calculated by figuring the amount of time between when a patient first arrives, to when the patient departs. On aggregate, if 100 patients come in to the ED in a day, the ED Average Length of Stay for that day is the sum of each of the 100 individual Lengths of Stay divided by 100.

While the Length of Stay includes time that the patient spends sitting in the lobby waiting for care, the bulk of most patients’ LOS is accrued “in care,” the time from when they are brought to a bed from the waiting area until when they are moved out of the emergency department. Patients generally are moved out of the emergency department through two unique channels:

1. Discharged – patient care is complete and the patient leaves the hospital
2. Admitted – patient requires admission to hospital for more care, further diagnoses, further observation, etc.

In fact the mix between these two outflow channels was about 50%/50%.  Recognizing that patients admitted to the hospital often faced delays due to a similar capacity constraint in appropriate hospital beds, I focused on identifying solutions that would expedite the progression of patients to be discharged. The first solution was create a standard procedure where the physician and nurse team assigned to each patient would discuss the likely outcome as early in care as possible: discharge or admit. By identifying patients in this manner, understanding that the disposition of the patient could change at any time, the team could track progress and eliminate idle time in any likely discharge patient’s process. For example, if a patient was identified as a likely discharge, and the doctor determined that the patient should be given a dose of medicine then observed for improvement, the administrative tasks related to that patient could be completed during the observation time such that as soon as the observation was complete, the patient could be quickly and safely discharged.

Care Team Communication – The Patient Flow Jackpot

This small change of forcing the doctor and nurse to talk directly to each other turned out to hit right on the real constraint in the process. While there was rarely a situation where the doctor and nurse were not communicating with each other, a surprising amount of the communication was indirect (e.g. notes in the patient chart) or left un-said, each side assuming the other was “on the same page.”

Next, communication delays were identified within the physician team. The physician teams, depending on time of day, were made up of 4-15 residents reporting in to 1-5 attendings. A phenomenon became apparent that residents were sometimes reactive or inactive in communicating up to the attending until the attending or a group of attendings called an update meeting. These update meetings would commonly go 3, 4, even 5 hours in between each, but within the 15 minute update meeting, there would be a batch of often 10+ patients that were suddenly cleared for admission or discharge.

To cut down on the physician batching I worked with the ED management to create an innovative nursing role. The position of charge nurse is common in hospitals and emergency departments, and in this busy ED, was fully occupied in managing the inflow of patients. Typically the charge nurse would work with the triage nurses to identify the patients waiting and their priority, then when a bed became available, the charge nurse would bring the patient to their bed, perform any minor immediate care tasks, and would assign a nurse to care for the patient and communicate the patient’s vital information to the nurse. We created a role that nearly mirrored the charge nurse, but was responsible for patient outflow. For example, if a care nurse was busy attending to patient A, but patient B appeared ready for discharge or admission, the outflow nurse, or Patient Flow Coordinator as we titled the role, would prod the physician to give next steps, or would assist the patient in preparing for discharge and exiting.  The Patient Flow Coordinator also became an advocate for resident – attending conversations. Using tactics from gentle hinting to blunt commanding, the PFC recognized the importance of keeping a patient’s care progressing, and if a patient’s care was complete, expediting that patient’s discharge so that patients in the waiting room could receive care.

The Patient Flow Coordinator also filled a missing role in being the collaborator of all patient cases in the emergency department that were to be admitted to the hospital. Previously when two or more patients were admitted from the ED at about the same time, the first admission into the system received the first available appropriate bed. However the PFC was able to identify opportunities to increase quality of care by designating which admitted patients would be best served at the moment a bed became available.

Finally, I coached the Nurse Managers to create a regular routine of getting out of the office and administrative activities that tended to tie them down, and to work with the PFC and Charge Nurse to remove barriers to safe patient flow. Usually this meant that the Nurse Manager would identify any patient who had been in the ED for >6hrs and to convene a quick briefing on the patient with that patient’s care nurse, the resident, and the attending. In this meeting the Nurse Manager would identify the root cause of the delay and work with the care team to create a specific action plan to ensure that the patient received the appropriate care… Generally a patient in the ED >6hrs would receive better care if they were admitted to the hospital.

This active management by the Nurse Managers also kept them close in tune with any delays caused by hospital support services. For example, if the imaging department was on a backlog, the Nurse Manager could communicate that to the entire ED allowing the care teams to make clinical decisions as necessary.

Results

The Emergency Department Safe Patient Flow Project was a success. Operational improvements were measured as follows:

• Adult ED Average Length of Stay decreased from 5:05 to 4:36
• Average Pre-Care Wait Time was cut from 40 minutes to 25 minutes
• Reduced Average Daily Elope Rate from 7% to 2%
• Improved % Patients with LOS >6:00 from 45% to 27%

By reducing the strain in ED, the ED was able to accommodate patients to be admitted for a longer LOS when the hospital was at maximum capacity and unable to take in further patients from the ED.

The creation of the empowered Patient Flow Coordinator broke down the culture of doctors dictating patient care progression.

Nurse Managers performing regular and targeted Active Management rounds, prevented isolated variances from becoming crippling roadblocks for the entire ED.

More Information & References Available

If you are interested in learning more about this project or the work that I have done, please call me by clicking below:

I can also be reached via email at rickmaher@gmail.com, and would be happy to share client references upon request.

To track my current projects and work, you should follow me on Twitter @Rick_Maher

Executive Summary – Lean Implementation Saves Food Manufacturing Plant $800k/yr

As a management consultant, I worked with a client food manufacturing plant to kick-start a stalling lean implementation. By focusing the effort on throughput at the constraint, lean methodologies such as 5S, Visual Management System, and Total Productive Maintenance (TPM), plant throughput skyrocketed nearly 60%. This improvement was achieved with zero capital expenditure and allowed for competitive advantage as low cost producer in growing demand.

As a management consultant I completed a wide range business improvement projects with world class companies in tons of industries. I have profiled a few of my projects here on RickMaher.info, but out of respect for my clients confidentiality, I have excluded their names and have changed some material facts and figures.

Situation

My client, the high volume manufacturing plant of a line of household name brand food products, was facing growing demand, but serious margin pressure from their largest retailer. Current plant performance was failing to match the demand, but a capacity expansion project would have been prohibitively expensive.

The facility operated 24 hours a day and 5-7 days each week. The facility was staffed to run 3 shifts per day and 5 days per week, thus incurring higher overtime wage costs for 6th and 7th running days each week. Further, the facility was set up on a continuous process manufacturing facility – meaning that the entire plant could only produce one product type (flavor) at a time. Thus to accommodate demand for multiple flavors the factory had to shut down and change-over the equipment, ingredients, and packaging.

Task

In a management consulting role, I was brought in to identify and implement business improvements that would increase capacity to match demand, and simultaneously cut costs by at least $500k/yr to protect profit margin.

Action

First I developed a working relationship with plant management as well as line staff. By building these relationships I gained a deeper understanding of what had been tried in the past, what the perceived current opportunities were, and what solutions seemed the most logical. I relied on these relationships as the project progressed to quickly develop buy-in and momentum on my recommendations.

Next I analyzed past performance and the current state process. I dispelled the perception that the packaging department was the constraint by showing that when all packaging equipment was running smoothly, its capacity was far greater than the extruders – the actual constraint. This was critical in changing the culture from focused on packaging line output to focusing on extruder throughput.

After the constraint was confirmed to be the extruders, I tracked any and all downtime there. I tracked it from two perspectives:

1. Qualitatively – what was the root cause of the downtime?
2. Quantitatively – what was the net throughput impact of the downtime?

Using this data I created a root cause pareto that illustrated the most impactful reasons for lost throughput. It turned out that line change-overs were the largest cause occurring on average four times per week and causing almost two hours of down time each occurrence. Indeed almost an entire shift of production each week was lost to line changeovers. Other big hitters of lost time included mechanical failures upstream starving the constraint or ingredients or downstream choking the constraint to reduced or stopped throughput.

Once the pareto was sufficiently consistent and accurate, I developed solutions to attack each unique root cause.

To reduce changeover downtime I looked to reduce the frequency of changeovers as well as the duration of each changeover. To reduce the frequency of changeovers I worked with the master scheduler to understand the demand and storage situation. Together we optimized the schedule to extend product type run lengths by as much as possible without ever sacrificing a customer shipment deadline or filling the warehouse to its capacity. Next I tackled changeover duration by performing a 5S Kaizen event on the changeover areas and implementing Single Minute Exchange of Dies (SMED) principles.

In effort to cut down mechanical failures along the line leading to downtime at the constraint, I implemented numerous changes with the plant mechanics. First, we identified the machinery that was failing most often and stationed mechanics nearby to reduce the amount of time between its failure and a mechanic on site beginning the repair process. Next I created a schedule for machinery check-ups and preventative maintenance. Finally with plant management I developed a long term machinery replacement plan allowing for proper budgeting and equipment lifecycle management.

Another throughput loss root cause was an uncountable number of points along the manufacturing line where perfectly good product would fall off the fast moving conveyor belts and onto the ground – of course rendering it unusable. While this root cause accounted for very little total production loss, it was very visible and generally a pain for the line staff to clean up and dispose of. I found this to be a perfect opportunity to generate serious buy-in from the line staff. I created a rotating “swat team” of line staff tasked with identifying one common “spill point” per day and a practical solution. These fixes ranged from having a mechanic spend 15 minutes fastening a shield rail to a conveyor to placing a sanitary bucket under the spill point and creating a communication process to ensure that it was replaced before overflowing.

Results

Ultimately the project was successful. Average hourly production increased from 60,000 lbs per hour at the beginning of the project to 95,000 lbs per hour by the end of the project 5 months later. This production increase accomodated the increased demand, and allowed for the elimination of overtime shifts – saving $800k/yr in labor costs. Specific operational improvements included:

• Reduced average change-overs per week from 4.0 to 2.5
• Cut average change-over constraint downtime from 115 minutes to 40 minutes
• Improved equipment down mechanic response time by 45% leading to constraint choke / starve downtime by 77%
• Eliminated 15,000 lbs of average daily food waste disposal – a 33% improvement

More Information & References Available

If you are interested in learning more about this project or the work that I have done, please call me by clicking below:

I can also be reached via email at rickmaher@gmail.com, and would be happy to share client references upon request.

To track my current projects and work, you should follow me on Twitter @Rick_Maher

I’m coming to the end of my project here in Hong Kong at a financial services company and have been loving all the time I’ve been spending coaching my team.

Whereas most projects are set up such that there is a team of consultants and one full time client resource dedicated to the project, this client provided 7 “analysts” with skills on par with junior consultants.

As it relates to this post, there are two key impacts of this set-up:

1. The analysts have not been hired under the requirements of, and through the training generally given to our firm’s consultants
2. The analysts have vastly more knowledge of the client’s (their employer’s) business than any of our consultants would ever have

What I’ve really enjoyed in this set-up is coaching this team to harness their knowledge of the business, but at the same time really think beyond the sacred old ways that are holding it back.

Today I worked with one of them to set up a strategy to gather some specific information:

• What mix of activities does an average employee encounter on any given day?
• How much time does it take to complete each variety of activity?
• What barriers does an employee encounter that prevents him from focusing solely on their primary responsibilities
• What expectations are communicated to an employee by their manager throughout the day
• How explicit is the manager in controlling the priority / sequence / quality / details of an employee’s work

You get the point.

To gather this information, the analyst sat smack in the center of 3 of the employees in question and took an inventory of everything that happened.

I checked in on him throughout the day and asked for any highlights or interesting information that he had gathered. He provided a concise summary each time. Then at the end of the day we met in the project office to outline how the summary would look. At first it caught me off guard when he was at a loss for words when asked to list the top three highlights of the day.

Quickly I realized that he had a great story – the ones we discussed when I checked in on him throughout the day – he was just struggling to format them into highlights.

We spent the next 30 minutes or so talking about how to mold the nitty-gritty details of what he observed into a few punchy points and one recommendation. We reverted to the question that was asked first thing in the morning that led to the observation in the first place, then talked about how his observations related.

I was content leaving for the evening after he had developed his own summary and recommendation with only some pointed questioning.

This project, more than others, has made me aware of the what I really enjoy about this job – coaching others.

I think tonight I’ll go back to a few classic Manager Tools podcasts on Coaching to see how I can improve.

I started this post to talk about another topic completely, but found it would be helpful first to post a quick run-down and commentary of my day today:
730-900
Build an Excel workbook to convert operational indicators (e.g. cases per hour) to financial results (e.g. saved $20k or lost $5k)
900-930
Set expectations of my team for the day – they all need to be wrapping up their portion of the project before we leave next week
930-1000
Pre-present the ops / finance excel workbook to a client project manager for first impressions
1000-1030
Sit in on a daily department planning meeting and offer advice on how to make a more exact plan for the day – this is a fire-fighting effort as the consultant in this department failed to instill the necessary changes
1030-1200
Chair meeting with a consultant on my team and a client director to understand their needs in performance metric tracking – challenge here is to convince the manager that they are not the victim of, but rather the maker of all that goes on – good or bad – in their department. This is very common rampant.
1200-1230
Enjoy lunch – eat at my computer and read emails – Today: Salmon Crayfish Salad from Pret a Manger (delicious)
1230-1400
Observe and coach client manager (same department as meeting at 10 this morning) in tracking daily progress and setting clear expectations with his employees and making the most effective use of their time. It was obvious that personal interaction between this manager and their direct reports is not a common occurrence. That changed today.
1400-1500
Check on progress of my team, verify that any deliverables will be completed when required. 5/5 say we’re on track but the pessimist in me thinks probably more like 3/5 are on track and I’ll have to be the bad guy tomorrow and make a long Friday for the other two tomorrow – too bad.
1500-1515
Client president showed up unexpectedly looking for my director who is not on site today, discuss progress of project with her and tell her how she can push a few key buttons to keep the project moving. It’s amazing how much can get done in 15 minutes with the right people.
1515-1600
Take a video from a presentation from earlier in my project up to the office of the other project going on here and get pulled into a meeting to convince the client managers that their counterparts in my project suffered through the same process and came out better for it. (Sunny – you owe me)
1600-1700
Review today’s work with junior consultant and give some tips on how to summarize and present. This coaching was my favorite part of the day.
1700-1800
Read / respond to a few emails. Reschedule a key project meeting from next Friday to the following Monday… Client President was unavailable at the scheduled time – whoops.
1800-1840
Make to-do list and plan of attack for tomorrow… why does it always seem like there is 10x more work to do then time to actually do it?
1840-1855
Write this post and make some quick notes for the post I actually planned to write – about the time I spent today coaching especially the period from 1600-1700.
1855-??
Out for noodles with the team… Christopher gets to go home tomorrow… lucky…

Head on over to http://asasu.net/usg/ to view the nearly completed website of the Undergraduate Student Government. The USG is the undergraduate arm of the Associated Students of Arizona State University (ASASU).

I expect the USG site to be the most comprehensive and largest of the three branches (GPSA and PAB being the other two).

It is being powered by a number of MovableType “blogs” working more as a content managment system than an actual blog. I pulled inspiration for this design/layout from a number of sources including http://simplebits.com, http://mezzoblue.com, and even the ASASU page created by Omar in the SDMU over the summer.

Check it out and let me know what you think.