In my last posting we discussed two examples of why it’s so important to look at the entire process rather than isolated parts. If we don’t do this, we could end up with a localized improvement that could have a negative effect on other parts of the process. Today I want to discuss this same situation and how it could negatively impact the next step in this particular process. Once again this is a healthcare example that I was personally involved with.
This scenario took place in a Radiology Department of a hospital in the Chicago area. When the project charter was written, I questioned its scope because what they had chosen to improve was a middle part of the process in question. Specifically the Champion had elected to improve the part of the process from registration being completed to the start of the radiologic exam. In addition, the Champion had limited the process to CT Scans, only one of many test that were routinely done in this part of the process. My concern was that if only this part of the process was improved, and I had no doubt that it would be, then patients would end up being “pushed” to the exam area creating an excessive queue of patients waiting to be tested. My concern was heard and considered, but in the end the leadership decided to continue on with their original charter scope.
The team leader’s first order of business was to review the project charter with the embers of the team, making sure they understood the boundaries of the process being studied and the targets this team was supposed to achieve. The team reviewed existing cycle time data and found that, on average, the time required from registration complete to exam start was 26 minutes with the goal set by the project champion being .15 minutes. The team received our customary training in the basics of Lean, Six Sigma and Constraints Management and began the hard work to reduce the cycle time by nearly 50% (i.e. 26 to 15 minutes).
The team chose to walk the process to better understand its flow and to gather any potentially meaningful information. This information would be used later to create a current state process map of the process from registration being completed to the start of the patient’s CT exam. The team noted that even though the process was relatively simple in nature, there were still two separate waiting rooms. This observation would become an interesting and useful bit of information. The team next created a SIPOC Diagram to better understand the process flow from a high level. Based upon the estimated times collected during the Gemba Walk, the team determined that the patients actually spent a total of 19 minutes between the two waiting rooms. Because patient satisfaction was another important performance metric, it was clear to the team that the extended waiting room times was an important factor to consider as they designed a future state. Figure 1 is their SIPOC which has both the original estimated times and what they believed their future state could be. The team used the SIPOC to identify their system constraint and worked to remove time from it.
Using the current state process map, the team then completed a value stream analysis and color-code each step as either green (value-added), red (non-value-added) or yellow (non-value-added but necessary). The team then created their ideal state map by eliminating all steps that didn’t add value which theoretically reduced the average process time from twenty-six minutes to about seven minutes. They then created a future state map which came in with a cycle time of just under 12 minutes. Even though they had met their project target of less than 15 minutes, the team understood that they had made a localized improvement that could have a negative impact on patient satisfaction. Their concern was that the constraint had probably moved to the actual exam and that by reducing the processing time, the number of patients entering the exam areas would probably increase.
Figure 2 is the team’s current state Process Map after completing their value stream analysis and as you can see there were 18 activities coded red as non-value-added (NVA), 6 non-value-added but necessary, coded as yellow and only 6 value-added (VA) activities coded as green.
Figure 3 is the Ideal State developed by the team and Figure 4 is their future state map.
In Figure 4 you’ll see that there were 0 reds, 5 yellows and 7 greens. The anticipated cycle time improvement moves from 26 minutes to only 12 minutes. One of the most noticeable difference is the complete elimination of one of the two waiting rooms which should have a positive impact on patient satisfaction.
As the team implements their future state recommendations, I’ll follow their project and see if their concern was valid or not. No matter what, this team now understands the potential danger of making localized improvements without considering the possible negative impact on the system.