by Lee Fried, on 18 Nov 2007 02:29 pm
The Journey
The Question of “Spread”
Just and FYI to the readers: I have decided to sunset for a while the Quote of the Week posting. I have exhausted my notebook of favorite quotes and will be rebuilding over the next couple of month.
Today I want to challenge you with a case study and am very interested in hearing your thinking. I have a feeling that many of you are working or have worked in the past through a similar challenge and am wondering what steps you have taken or took. I will provide my current thinking as well for feedback.
Description of the current state:
- Twenty-six service centers (medical) across a two state region
- High variation in customer needs depending on geographic
- No two work environments the same: size of centers ranges from 20 to 100s of employees
- All centers provide basic services, but number of special services provided depends on center
- All centers utilize same information system
The question: How to build standard work and “Spread” improvements across a system with high variation?
This was the question that I discussed last week with several highly committed leaders that are looking to move from a manage by results to a manage by process model. How can we possibly implement standard work across so many centers that are so different? Possibly even more challenging is trying to figure out how once we have standard work in place across all of our center will we spread an improvement once one center has figured out how to do something better?
I have done a lot of reading about how other companies approach this challenge, but have not been satisfied with what I have read. Most cases and articles talk about the need for best practice sharing session, promotions for latteral learning processes or “meet it or beat” policies. While I believe all of these strategies have merit they are not sufficient. None of them go deep enough. They all seem to be a single component of a much bigger system.
So here is my latest thinking of how to manage this challenge. To try and identify and spread best practices across the system in the form of standard work is unsustainable. We would be solving the symptom and not the cause. I believe that variation is the result of non-standard management work. Thus the best approach that we can take is to focus first on creating a consistent and standard management system that all centers, regardless of size, location or service mix would adopt. In this system the role and the processes of management will be the same in every center. All managers would follow the same improvement process, learn how to manage by data and facts, teach staff, follow a standard checking cadence, and focus on bringing processes under-control by implementing standard work within their centers. At the same time the organization would need to work to redefine our planning and measurement systems. Only then could we focus each team on the most important improvements and identify with relability best practices.
So am I on the right track?
on 19 Nov 2007 at 4:11 am 1.Rob said …
Is it not worthwhile to identify common, core business processes across all areas? Then focus on the creation of a model value stream for one core activity, with the idea that when it is successful it could be horizontally expanded to others? You can also identify the CTQ factors in this process and minimize variation, thereby stablising the process. This will then set-up a foundation for standard working practices … ideas surrounding this may also come out of your future state VSM. Don’t forget about communication too!
on 19 Nov 2007 at 12:30 pm 2.Anand said …
I believe that standard work can be implemented successfully across multiple divisions only when the same metrics are used to measure outcomes. The metrics should measure the departments based on the market and the consumer rather than products and services.
I will try to explain my point using an example.
The current problem for Group Health (GH) is high operational cost. The underlying reason to reduce variation is to reduce cost. If we draw a value stream for the whole of GH relative to the consumers’ premium, then we can identifying the cost of operation in each division (each box in the VSM) that is involved in the value stream. This analogy is similar to the work that is done at the lower levels at GH the difference being that the planning and coordination needs to happen at the top. We need to use the same metrics for each division. The metrics that can be used for the individual divisions are, for example, ‘operational cost’ and ‘revenue’. This is not Management By Objectives (MBO) if the senior leaders are involved in this process. There is bound to be one box that is the bottleneck adding to the cost. To reduce operational cost, we need to have cost transparency within GH. The employees need to know the budget allocation for their group and how it is managed every quarter. We need to have ‘Cost’ report outs just like the ‘Lean’ report outs. Then we will know what and how individual division is doing to reduce variation and hence reduce cost.
In short, in my opinion, if the work done by different groups can be measured using the same metrics then it is easier to spread standard work.
on 20 Nov 2007 at 10:37 am 3.Kent Schnaith said …
I think the key is to put the 26 service centers in touch with each other. So they are aware of what the other service centers are doing. As suggested, this will likely involve making the management more transparent and measuring the service centers the same way.
Spreading improvements sounds like telling people what to do. I think what you want is more of an attractive model where an improvement by one group is voluntarily adopted and adapted by other groups.
on 21 Nov 2007 at 3:34 pm 4.Shawn Kelly said …
First I think your approach is sound but incomplete. It would be my sense that system wide you should be able to design processes, “what to do”, and supporting metrics to represent all of the work that the system performs regardless of location or patient. Since everyone in the system is using the same basic process for each type of patient care, you would be able to perform analysis across to system to determine trends either at the system level or the center level.
After defining the “what to do” at the system level, then use Standard Work, “how to do it”, at the center level.
Defining process at the system level should reduce variability across the system.
Defining Standard Work at the center level should reduce variability within the center.
For example, if one of the specialties offered is Orthopedic Physical Therapy (PT), then:
The process: validate prescription, validate insurance, perform evaluation, develop treatment plan, schedule first treatment within 3 days of evaluation, perform treatment, perform 30 day re-evaluation, etc. should be the same process used across the system.
If we take the perform evaluation and perform treatment components, from a standard work perspective, there may be different locations for the equipment or different equipment used to perform those processes because of the difference in center layout. The basic process is not changed but the how to perform it would be.
The second item that I consider incomplete is a lack of focus on patient care / customer focus. Whether it’s the process or the supporting metrics the patients will be the ones affected along with the employees.
There is a different dynamic compared to that of designing standard work in a manufacturing environment. In this case, the patients are the objects work that have their own variability (injury, feelings, disposition, stress, etc.) as opposed to inanimate parts and assemblies. Assemblies won’t gripe when you don’t certify their insurance correctly and you hand them a $300 bill!
on 25 Nov 2007 at 3:44 pm 5.Lee Fried said …
Hi Guys,
This is great advice from all of you. I appreciate your willingness to share your thinking.
Best,
Lee
on 28 Nov 2007 at 2:39 pm 6.Barbara Fetty-Solders said …
I would like to add to Shawn Kelly’s point. Part of the difficulty with designing “standard work” whether management processes or direct care processes, is that the needs of each of our patients can be “Non-Standard”. Each is unique, requires assessment, and a creative and individualized plan of care which statisfies the customer, meets their needs and doesn’t use up all of GH’s resources.
on 29 Nov 2007 at 6:17 pm 7.Ted Eytan said …
In practice, I celebrate the fact that every person is unique - no story can be explained by a chief complaint alone. This is what makes primary care satisfying.
At the same time, I enjoy the ability to focus on the unique aspects of a person because other aspects of the care experience are delivered through standard work. This includes the way they are introduced to the experience, the way we reduce uncertainty through testing, and the way we provide compassionate treatment and reassurance when needed.
If I (or a patient) can’t predict how a care experience is going to happen and need to adjust these parts - one way when working with one person, one way when working with another -, then we don’t get to spend the time we want to focusing on the special parts.
on 03 Dec 2007 at 9:45 am 8.Curious Cat Management Improvement Blog » Management Improvement Carnival #24 said …
[...] The Question of “Spread” by Lee Fried - “All managers would follow the same improvement process, learn how to manage by data and facts, teach staff, follow a standard checking cadence, and focus on bringing processes under-control by implementing standard work within their centers.” [...]