King: Begin at the beginning. . and go on till you come to the end: then stop.

This advice given by Lewis Carroll in 1865 clearly was not meant for corporate planners. Planning has no start, middle or end and you cannot stop at any point – planning must be continuous.

From a practical perspective, Business Process Management should begin with a knowledgebase of pretty much everything you are likely to need to discover processes, map these processes, improve your processes and then, proceed to use your mapped, improved processes as best practices to guide the processing of instances. Start with a 30,000 foot view and then dive down to the details.

Examples of what you are likely find in an inventory of objects for process mapping include industry knowledge, evolving legislation, information on competitors, corporate strategy/policy and procedure, a portfolio of current products and products that may be on the drawing board, and, of course, an inventory of existing processes.

However tedious it may seem, there are important benefits to be gained from carving up objects into short, highly focused mini-objects.

A search for occurrences of a keyword in an industry standard in a free-form Kbase will highlight clauses in an organization’s policy and procedure, responding to questions such as “… where in our policy and procedure do we address a clause in a particular standard?”. And, with no added effort, the reverse search “. . . where in a standard does is say we have to follow a particular clause in our policy and procedure?”

Given a Kbase of processes and information needed to improve these and to develop new processes, it becomes relatively easy to prioritize initiatives at the Kbase. From here, projects can be planned, launched and managed, always with a return to the KBase to update status and re-prioritize initiatives.

Of course, there are Kbases, and then there are Kbases – if your objects are in a graphic environment you can use a “Russian Doll” approach to organizing these and you can go a long way toward accommodating thousands, even tens of thousands of objects on one screen. As anyone who has moved from a traditional computer monitor to a large monitor has discovered, the bad news is you will eventually run out of real estate but you can always go to a rotating/zoomable sphere and continue on.

I suppose the next step is to transition to a command center where you sit inside of a holograph and get at your data. They seem to be able to do this kind of thing on TV, so the wait may not be too long.

I suspect the travel business has stopped worrying about tele-transporter technology which dates back to the mid-1960’s, so who knows?

Development of processes is best done by functional units with the help of a facilitator if need be. This is easier said than done unless you have access to a mapping environment where processes can be mapped as quickly as stakeholders say “ . . . and then we do this”. The facts are some very large corporations lock themselves in time by using whiteboards and post-its and seem none the worse off.All this does is tell us that there is a separation between efficiency and effectiveness.

Beyond process mapping, the herd starts to thin out.

Few organizations where staff is mostly made up of knowledge workers reach a stage of maturity where best practices are encouraged, yet where ad hoc “processes of one step each” can be engaged at any time to the point where a series of ad hoc interventions can substitute for best practices.

In any case, the key point regarding process discovery, mapping, improvement and run-time use is that when you think you are at the end it’s really just a new beginning.

“So what software would people recommend for mapping out business processes where they are complex, interdependent and split across multiple disciplines”

And my recommendation was …

Absolutely no point setting a singular focus on mapping.

Once you have mapped a complex process, you have two options : stare at the paper process (to accomplish what exactly?) or subject the map to pattern recognition (e.g. compilation of the map) and forward feeding of process steps to an automated resource allocation, leveling and balancing environment.

Clearly, the compilation should be a one-click exercise, not some hay-wiring initiative, so this tells you that if the mapping environment does not accommodate tagging of steps with skill attributes, easy access to step-specific HELP, and instant access to any form(s) needed to collect data relating to performance of steps, you are not in the right environment.

Some BPM systems accommodate this, others seemingly do not.

A rough rule of thumb I use is that if a person performing tasks has to worry about WHAT the next step is along a process, WHO should perform it, WHERE, WHEN, HOW, and occasionally WHY, or what document is needed for data collection and as proof of completion of a step, the environment being used is not quite there and may or may not ever be there.

Then, we have the issue that for any work that involves knowledge workers, these people need to be able to skip steps, perform steps out of order, invent new steps that were not in the BPM template, revisit already committed steps, so the combo of Adaptive Case Management (ACM) and BPM works better for this than either of these two methodologies on their own.

See “Doing the wrong things the wrong way” at http://wp.me/pzzpB-f4 for a discussion on the relative merits of BPM and ACM and when one should be used versus the other.

Given a mostly linear process where steps can be performed by machines it makes good sense to automate. Of course, the decision to automate should be backed up by an ROI that turns positive within a respectable timeframe and once a commitment has been made to a particular initiative, stakeholders have to stay the course or the stated benefits may not be achieved.

If we change the scenario to one where the process is no longer linear, where a fair number of the steps require human input, where different skills are required for the performance of different steps, where steps are connected in complex ways and where there are numerous branching decision points along workflows, automation of the “work” becomes a less attractive strategy.

The management methodology called Business Process Management (BPM) is effective for automating work except for work that must performed by knowledge workers.  Here, Adaptive Case Management (ACM) does a better job of because it does not rely on pre-mapping steps that make up an initiative.

Whereas in BPM both the steps and their interconnections must be mapped out in templates, in advance,  in order to transform inputs into outputs, ACM accommodates ad hoc interventions and puts the focus on tracking progress toward the achievement of goals.  In ACM, the steps, their interconnections and their timing is left up to knowledge workers.

Since both BPM and ACM have as their fundamental objective the transformation of inputs to outputs to yield deliverables (goods or services, it does not matter), it’s not surprising to find proponents of each methodology taking up defensive positions on which approach works best for what.

Clearly, if we classify work according to the degree of automation (i.e. 95%/5% structured versus unstructured at one end of the scale, compared to 5%/95% structured versus unstructured at the other end), we are likely to see more BPM for 95%/5% work and more ACM for 5%/95% work. But, what about closer to the middle – which methodology is best for a mix of 50%/50% structured work versus unstructured work?

How about a methodology called ACM/BPM where we encourage everyone to focus on reaching goals using ad hoc interventions where appropriate and using “best practice” workflows where appropriate? ACM enthusiasts will have no problem with this, nor will BPM enthusiasts.

Wait a minute!  What’s with “encourage”? Will this not lead to some people following best practices whilst others get the notion that some best practices are not good enough and cannot be followed?

Yes, this does lead to work being done in different ways but the marriage of ACM/BPM brings together “Case Management” from ACM and auto-resource allocation, leveling and balancing from BPM and if you think about it EVERYONE routinely handles a varying mix of structured and unstructured work.

Each day we go to work, be it at a physical office or a virtual office, we attend to our fixed-time appointments (structured work) and between appointments we work on To-Do List items (unstructured work). Like it or not, we have all been doing ACM/BPM all of the time.

Given a proper ACM/BPM User Interface that allows work to be organized at a User Calendar (for structured tasks) and at a User InTray (for unstructured work), we have only to look at our Calendars for fixed-time appointments and look at our To-Do List. Events post to our calendars, tasks post to our To-Do List.

And here comes the Eureka!

With a setup as described, we get to the remarkable position where it’s all about tasks, not structured tasks versus unstructured tasks and if your User Interface has a Calendar on one side and a To-Do List on the other, you can handle 100% of your work at ONE screen with no need for navigation (ACM/BPM in the background will take care of any and all required navigation).

And the bottom line with a setup like this is management gets to where they understand that it’s not tasks that need managing but rather handoffs of tasks across people, thereby accommodating a shift away from doing the wrong things the wrong way to doing the right things the right way.

I encourage you to read this post over a few times, it took me about 10 years to figure all of this out.

Healthcare is complex. No two agencies manage patients the same way. One-size-fits-all “solutions” do not work.  At the other end of the spectrum you can hire a consultant to build the next-generation Ferarri in their garage. Bill Gates did it. Steve Jobs did it. Why not your consultant?

Your options at this point are down to one – configuring an off-the-shelf product to suit your needs.

But, how can you ensure that any software product will meet your current needs? And, who says, anyway, that your staff will be able to articulate the needs of your agency in language that a vendor will understand?

The simple formula for success in shopping for an EHR is to take one of your programs and have the prospective vendor build a process map for your program, assign skill categories for the performance of tasks, attach to each task images of any forms needed to record data at the task and to provide evidence of task completion and give you access to a run time environment where you can piano-play instances of your program. All at no charge.

If the vendor cannot do this within 2-3 days or, if the vendor cannot be bothered, you need to move on.

The watchwords listening to vendor-speak are “configuration” versus “customization”. Configuration requires little time and little labor, customization requires significant time and significant labor. Guess who ends up paying?

Here is a tip that will probably save you from being led down a garden path when shopping for an EHR.

It’s very easy to mock up a software system so that it appears to be doing this or that.  So, as and when your prospective vendor gets back to you with an implementation of your program, tell the vendor at the start of the web demo that you want to run the demo.  If you cannot process a patient through an instance of your own program without training, your staff will not be able to either in the eventual all-program EHR environment.

When was the last time you bought a car without test-driving it yourself?

Management can “encourage” the use of structured flows but individual staff members can do what they like so long as they are “boiling in their own puddings”.

If the nature of the work is such that it can be described in terms of a BPM flowgraph (linear for a straightforward process, a mix of process steps and decision points in a more complex process) then if a flowgraph includes in-line compliance checkpoints, rule sets can grind things to a halt as necessary until instance processing remedies are put in place.

If the nature of work is such that a flowgraph is not appropriate, things become more difficult.  If staff are free to do what they like, when they like, how they like, we have to rely on gap analysis between goals and progress toward goals to guide the processing.

How do we manage work in an organization that needs to manage a mix of structured and unstructured work?

Many clients I have worked with assume that if they provide an ACM/BPM case environment that accommodates any mix of structured/unstructured work, staff will use flowgraphs where these are available and will perform ad hoc interventions in the absence of flowgraphs.

Sadly, this does not work out.  Some staff will follow flowgraphs, taking the view that following flowgraphs is easier, others, in spite of great efforts by the organiztion to accommodate performing process steps out of sequence, skipping over steps, re-visiting steps that have been committed and inserting steps, will not.

Clearly, we want outcomes to be the same regardless of the work approach taken.

The key to consistent outcomes is to have a few background, system-level gatekeeping tasks that “sniff out” activity, consult rule sets and raise red flags when things appear to be starting to go off the rails.

Here are several examples of gatekeeping activity.

-          You cannot get a case charge number  unless/until you open a case and input a few demographics

-          No payments of any consequence will be made by accounting in the absence of an invoice.

-          If you don’t ship via the shipping department, the shipping fee is on your account and QA will start grinding their axes as soon as they hear about this.

Otherwise, do what you like and so will I and we will both arrive at the same outcomes.

Change managers play a decisive role in getting staff on board with new software systems.  

The greater the change, the greater the resistance, it seems. 

I was fortunate to discover this business rule early on and figured that the easy way out with new software was to pay a lot of attention to User Interfaces. 

In the area of EMRs/EHRs where it is self -defeating to try to capture information faster than scribbling something on a small piece of paper (never mind the fact that no one will be able to read the note), it seemed to me the next best thing was to provide staff with a UI that mimicked a traditional day calendar and “to-do” list.

If you think about it we all attend to our fixed appointments each day and work on our to-do lists between appointments. So, a single UI window with a calendar on one side and a listbox on the other side is familiar and covers 100% of what we all do, every day, all day long.

Still, once in a while you will come up against a group that is particularly resistant to a new system. This is where change managers come in.

I recall one situation where my client warned me in advance that things would be difficult. Several stakeholders were specifically singled out as ‘dangerous’.

The client recommended multiple training sessions over an extended period of time. 

I countered that the best way to handle this would be to provide no training and to set up one computer terminal within a perimeter with a small sign that read “Keep Out”. 

Phase II of the plan was to pick a product champion and have this person work at the terminal with explicit instructions not to divulge any information regarding what he was doing.

After about 2 days, we started to get complaints – “. . . why was I not informed of this initiative?”.  We dodged all such questions. 

Two days later the problem had escalated to the point where one individual came in and pounded the table demanding to be given access.  The staff member had no clue what it was she was being denied access to, but she wanted access.  Again, we said we were sorry but powerless to do anything.

After a few more days, we announced that due to popular demand, all staff would have access to the new system and if anyone wanted training, we would be happy to provide it.

So far as I am able to recall, the implementation went forward with minimum fanfare.

Reading this over, a reasonable expectation might be that the best environment in which to carry out “case management” would be a “case management system”.

In the world of EMRs/EHRs, there are only a few software suites that have their foundation in Adaptive Case Management.

According to the Workflow Management Coalition (WMC)  adaptive case management is “. . information technology that exposes structured and unstructured business information (business data and content) and allows structured (business) and unstructured (social) organizations to execute work (routine and emergent processes) in a secure but transparent manner.”

It’s not clear to a casual reader that ‘structured’ applies to “best practices” in healthcare nor that ‘unstructured’ applies to “ad hoc interventions” in healthcare.

So, healthcare agency staff roaming exhibition halls to pick up information on the “latest” EMR/EHR software is not likely to ask whether a particular product is a “case management  software” system. Asking the question probably would be pointless in any event.

As Alice would say “. . . curiouser and curiouser”.

CA is about to introduce budget cuts of close to $1 billion.  

Whereas it’s  easy to understand that budget cuts are necessary at this time, the obvious concern is whether the bureaucracy will be proportionately reduced along with program cuts or will we end up with a larger than ever bureaucracy to “manage” the downsizing.  

What’s going to happen to the 70 natural parks  in CA that are being closed?

And, to put a focus on this post, what’s going to happen to healthcare services?

CA, of course, is not unique, cutbacks have already taken place elsewhere across the country and are taking place, and the big question for healthcare agency administrators is “What are your plans for 2012 and beyond?” 

One option is to scale down to where wait times become ridiculous, where services have deteriorated to the point your agency can no longer provide even essential services and where morale for the few remaining employees is in the basement.

The great EHR Meaningful Use (MU) incentive program was well intentioned in its planning but the math was not right. You cannot expect providers to invest in technology that has a focus on long term outcomes data analysis but does next to nothing in the area of improving outcomes for individual patients.  The EU did a major study on the ROI of EMRs/EHRs and found that the payback period was much longer than most investors’ financial attention spans.  Could have/should have.

The jury is still out on MU but one thing clear is that “dressed up” twenty-year old billing systems are not going to increase staff efficiency, increase patient throughout, decrease admin/clinical errors or improve outcomes for individual patients.  What agencies need is software that allows them to “do more with less”.

One sign things make little sense in all of this is why have “case workers” using billing software for patient management instead of using Case Management software?  

If you would like to learn more about Adaptive Case Management for healthcare you are at the right place – check out some of the more than 70 posts on ACM and Fixing Healthcare at this blog.

I suppose the result of this post is going to be that 500 EHR systems will overnight become ‘Case Management Systems’. 

This is where buyer research comes in – when you need to go into the market, do your homework by first documenting your strategy. What is it you wish to gain in the way of benefits from EMR/EHR software?

Ask the vendor to show you an end-to-end demo that features one of your programs.  To make sure the demo is not rigged, tell the vendor at the start of the web demo that you want to run the demo yourself. If they say no, you might as well move on, if they say yes, the “test drive” will give you a level of comfort with the proposed software solution.

Above all, don’t lose sight of what seems obvious but may not be – you need a case management environment to do case management. 

I use a mix of three mainline work management approaches with my clients (CPM, BPM and ACM). Reading these backwards, you get A, B, C which looks auspicious.

Goal attainment on projects planned and monitored using Critical Path Method (CPM) requires little attention because of the inherent focus on predicting arrival times at project milestones. Analysts start tracking “float” from day one and management has no problem relating to summary level reports that show plan, actual and projected-to-completion dates for project activities.  Management knows that if float starts to erode during the 1^st 20 percent of the timeline it becomes increasingly more difficult to avoid being late at project milestones.

In discrete manufacturing where Business Process Management (BPM) process maps are characterized by multiple decision points and where branching can only be determined at the time these steps become current along a process instance, it becomes next to impossible to predict arrival times at process end points.  But, goals usually are clear (e.g. build a prototype, test it, ship it).

In the world of office/services where a high percentage of the work is unstructured, Adaptive Case Management (ACM) is the preferred “methodology” but, in the absence of pre-defined logic connections between pre-defined steps, tasks or interventions, the only methodology, really, involves intuitive interventions such as opening a case, adding material to the case, performing whatever work is required to reach objectives and then closing the case.

The big question with ACM becomes “When do we close the Case”?

In healthcare, deciding what interventions a particular patient needs is difficult, but when to close is easy – we close (i.e. discharge) when the problem(s) have been sorted.

The appropriate time to close a case for a new product release initiative that involves research and development could be when the product team run out of funding, when a particular level of functionality has been reached, when it becomes evident that the product “will not fly”, or when the product team has been leapfrogged.

One way to “predict” case closure in an ACM environment is to assign relative values to objectives/sub-objectives and track ‘Earned Value”.   The EV model does not have to be complex.  If there are five interlinked ‘deliverables’, a simple matrix with five weighted rows and five state columns (not started/started/in progress/essentially complete/complete) with calculated summary data points 0,10,25,50,75,90, 100  may be sufficient. If the case involves final assembly of five otherwise unrelated deliverables, you can use the same model, set the weighting to 1.0, and report 0,20,40,60,80,100.

Bottom line, whatever non-subjective/subjective approach to predicting project/case closure you use, the information should post each time anyone opens the project/case.

Adapting to 45 CFR Part 170 Subpart C of the Final Rule for Health Information Technology.

The consensus is sooner or later healthcare agencies are going to be herded into adopting “certified” EMR/EHR systems.  The party line is “a certified EMR/EHR system is a good thing” and there is truth in this, but the question is “good for whom”?

Right now we the industry is at MU Stage I (don’t bother to read the 2000 pages in the final rule).  Basically, it says that for eligible professionals attending to Medicare patient populations, you can, over a period of five years, get a $44,000 incentive payment per eligible professional.

Nowhere do we see what the required outlay is relative to the $44,000 incentive.

Remember the old saying “if you torture statistics sufficiently, they will tell you want you want to hear”?

ROIs that exclude hard-to-identify costs look good but most agencies don’t prepare ROIs and an even smaller number have the ability to cost out the indirect costs, so is payback from the investment positive or negative?

The EU did a big study on ROI of EMRs/EHRs – the payback period was far longer than most agencies in North America have patience for, besides, things are different over here, so, again, who knows?

Chances are it does not matter if the prediction of mandated EMR/EHRs comes true so the key issue becomes can a chosen EMR/EHR increase staff efficiency, increase patient throughput, decrease admin/clinical errors, and improve compliance with internal and external rules and regulations?.

The question I put to my clients when they ask whether they are doing the right things is “Are your workflows working for you or are you working for someone else’s workflows?”

On the presumption that we will see mandated EMRs/EHRs, Civerex has put together a blueprint for 2012 and beyond that will see a consolidation of our two healthcare products (CIVER-PSYCH® and CiVER-MED) to a single product called CIVER EHR.

The rationale for this is quite simple – at present there are no BH MU criteria – right now, the only category open for outpatient  is the Ambulatory category.  It includes various criteria such as Vital Signs, Submissions to Immunization Registries, Public Health Surveillance reporting and Automatic Measure Calculations/Reporting.

It may turn out that BH agencies can exempt themselves  from some of these criteria but we don’t know.  A consolidated MU product does not have to deal with the uncertainty – we will certify with ONC as CIVER EHR and content will then set the focus on BH, Medical or both.