I’m working on a very interesting project in a healthcare environment this week. It isn’t my usual type project where we are interested in improving the throughput of patients through a process like an Emergency Department, but it’s clear that improved throughput will happen as a result of this team’s efforts. This project involves the billing for immunizations at a healthcare organization’s 6 different satellite operations. It seems as though there are numerous billing errors that translate into lost revenue for the hospital, so they called me in to help them reduce these errors. The team make-up was important because it included Medical Assistants (MA’s), front desk personnel, and an expert in Clinical Informatics plus two Green Belts.
The first thing I had the team do was create a Pareto Chart of the six satellite operations to determine the distribution of billing errors between the operations. As you can see, there is a wide disparity between each of the offices in terms of number of billing errors. for data collected over a three month period (Oct-Dec, 2012). On the surface one might conclude from this analysis that we should review the process in Location # 2 and have all sites use their process. This would be a good approach if all things were equal, but digging deeper into the data indicated that there is a wide disparity between locations in terms of the number of immunizations delivered (i.e. number of patients immunized).
The next thing I had the team do was to develop a causal chain to determine potential root causes for billing errors across the different satellite organizations. This effort proved to be very valuable from a problem solving perspective, but even more so it was very enlightening for the Medical Analysts who comprised the majority of the team. It helped them crystallize their thoughts about what was happening in their own site location.
For those of you who may not have used a causal chain before, we started by developing a very succinct problem statement which in our project was Billing Revenue Lost. The object (Billing Revenue) is listed on top of the line and the state that it’s in listed directly beneath it (Lost). We then asked the question why? To determine potential reasons why revenue was being lost.The answer to this first “why?” was actually five potential reasons as follows:
1. Order (Object) – Not documented in TW (State). TW is the name of the database being used.
2. Order (Object) – Not Bubbled (State)
3. Billing Company (Object) – Commits Errors (State
4. Charges (Object) – Incorrect (State)
5. Standardized Work (Object) – Not followed (State)
We continued asking why until we arrived at an actionable item we could work on for each of the five potential root causes. The team’s most enlightening potential root cause listed on the chain had to do with the standardized work which had been developed for each of the sites. When the team asked, “why is the standardized work not being folled?” there were three reasons:
1. The standardized work was not communicated, meaning that several of the MA’s didn’t even know it existed.
2. The standardized work was not “owned” by the MA’s
3. The standardized work was not enforced by the leadership.
For reasons 1 and 2, the team asked “why?” again and concluded that the MA’s weren’t involved in the development of the standardized work so they didn’t own it and didn’t know it was even available to follow. As I’ve written about many times in this blog, if the subject matter experts (in this case the MA’s) aren’t involved in creating the process, it won’t be credible to them. And because it isn’t credible, it won’t be followed and used. It really matters not what the industry is, in every case what I typically see is leadership creating the process with very little, if any, input from the SME’s so it’s not owned, believed or used.
The team then reviewed each potential root cause to determine one of three different “ratings.” If they believed the entity was “solvable” it was color-coded as green; if they believed the entity could be influenced, it was color coded as yellow; if they did not believe they could solve the entity, it was color-coded as red. In all cases it was determined that all entities were either “solvable” or “influenceable.”
The team then decided to develop process maps for each different locations and then perform a value analysis on each one. What the team found was very interesting in that, not surprisingly, each of the locations had a completely different process. Although I’m not at liberty to visually display each of the different process maps here, suffice it to say that some of the differences were quite significant as were the number of billing errors.
In addition to the five current state maps, the team then developed two separate future state maps, one for scheduled patients and one for walk-in patients as demonstrated in the figures above. This was an important step because in order for the team to improve the process (i.e. build their future state), it had to first understand their current reality.
In the table below we see the results of this team’s actions in terms of the before and after for total number of steps, value-added, non-value-added, and non-value-added but necessary and as can be seen, the differences were quite dramatic. Of particular interest was the reduction in the number of yellow and red steps.
The team is very confident that if the two, new future states are implemented, the number of billing errors will be significantly reduced (i.e. >75% or more), but there is another benefit that is much larger. The team was able to identify a key policy constraint that is delaying the administration of immunizations. The team reasoned that if the physicians will simply enter the order into the data base, instead of waiting to verbally explain to the MA’s what the order is when they exit the examining room, the MA’s can prepare all of the necessary forms, supplies, medication, etc. needed to administer the vaccine in advance, thus significantly reducing the wait time for the patients. The team believes that breaking this policy constraint will have several very important effects:
1. By decreasing the patient wait times, patient satisfaction scores will increase which could have a positive impact on reimbursement rates.
2. Since the overall immunization cycle time will decrease, the number of patients that can be seen each day by each location should increase, thus increasing the overall revenue base.
3. Because the MA will no longer be getting verbal orders from the physicians, the number of mistakes due to interpretation of the verbal orders, errors will decrease significantly.
These effects are clearly very important, but the team believes that a significantly larger financial impact will be seen if their new process is expanded to other processes such as EKG’s, Point of Care Testing, etc. These items are much higher revenue generators than the revenue gained through immunizations, so the impact on the organization’s bottom line could be substantial.
This team learned several important things during this event, not the least of which was the value of bringing together the true subject matter experts, the people actually doing the work, to develop the “best” way to perform different functions. Comments like, “I finally have a voice!” were not uncommon.
One of the things I always do in an event like this is to measure the team’s pulse before and after the event. The purpose of this action is to get a feel of the team’s belief in two distinct areas:
1. Does the team believe that what they came together to do would be difficult or easy?
2. Does the team feel that it will be empowered or authorized to make their recommended changes.
The figure below is the tool I use to measure the pulse of the team by having them place red and green dots in one of four quadrants:
· Quadrant 1: Hard to do and not authorized to make changes
· Quadrant 2: Easy to do, but not authorized to make changes
· Quadrant 3: Hard to do, but empowered to make changes
· Quadrant 4: Easy to do and empowered to make changes
At the beginning of the event the team is instructed to place a red dot in one of the four quadrants and then a green dot at the conclusion of the event. At the beginning of the event, for the most part this team felt empowered, but that it would be difficult to make changes. At the end of the event, the team believed that they actually were both empowered and that the changes would be easy to make.