As one of three Six Sigma Master Black Belts at Commonwealth Health Corporation (CHC), Nick Nauman mentors 35 Green Belts. At the time of this interview, CHC’s team had a total of approximately 180 completed or ongoing Six Sigma projects throughout the institution, focused on a range of issues involving both operational and clinical improvement.
Q: Since CHC was one of the early adopters of Six Sigma and now has a team of highly skilled and self-sufficient practitioners in this methodology, I thought it would be helpful to gain some perspectives on the approach and how it is being applied in your organization.
Clearly, some of the early wins for healthcare have been operational in nature – improving the billing process, reducing cycle time, streamlining workflow and so on. Now the emphasis seems to be increasingly moving toward the clinical side. Can you share some of the things that CHC is doing in this regard?
A: It’s still somewhat early in this evolutionary process, but I would say that we have perhaps 10 projects that could be classified as clinical improvement efforts. There are some projects, such as within the radiology department, that may impact patient wait time and it could be argued that they also affect quality, but we also have projects that are more obviously within this realm and focus on areas such as surgical procedures and infection rates.
Q: Can you give me an example of one of these clinical projects?
A: In March of 2000, we began data collection for a project to study order patterns between urinalysis and urine screens. The Green Belt for the project was the assistant lab director. This particular project was part of an overall focus on maternal care patients (labor and delivery), and improving the processes that surround this category.
Generally, we have a group of Green Belts working on a single, large focus area – so if there are 15 Green Belts in a training session they would all focus on one area. As we looked at benchmarking data for maternal care, we found that for DRG 373 (vaginal delivery), the costs were higher in some cases and there was an opportunity to apply some of the Six Sigma tools.
Through our hospital information system we reviewed the standing orders option that allows the physician to issue a set of standing orders when the expectant mother presents to the hospital. One such order was to automatically request a urinalysis test, and base the subsequent plan on the urinalysis results.
Looking at the data, we found there were 10 of 14 physicians routinely ordering this test. But there is a less resource-intensive, less costly test available involving a urine screen. When a urine screen is performed, if certain criteria are present, then a urinalysis test will be run. Our objective in this project was to get the physicians to order at minimum the initial urine screen, which would conserve resources and benefit both physicians and patients since they’d get the results in a shorter timeframe. To help achieve this objective, we had a physician work directly with the OB/GYN doctors to change the standing orders from urinalysis to urine screens.
If the OB/GYN physicians encounter a particular need they can still order the urinalysis and other tests as indicated. Early on, we met with the lab director’s team to figure out what they could do and how the labor and delivery staff could be re-educated to understand and adopt the new standing orders.
Using the Six Sigma DMAIC (Define, Measure, Analyze, Improve, Control) process, we went from a “z” score of 0 to 3.2. During the control phase, we used the chi-square test as a mechanism to basically lock the new operating procedures in place. Our UCL (upper control limit) was set for 13% defects. CHC’s hospital information system prints a report that shows how many urinalysis tests were ordered for a particular physician, and we can also run a report for the urine screen test.
Moving forward, the standing orders for the 10 OB/GYN docs would be for initial urine screens, and subsequent urinalysis only if warranted by the screen results. We completed this project in November 2001 and are now monitoring our progress to be sure we remain within the control limits.
One of the deliverables is to understand what this is costing the system and what are the CTQs or the elements critical to quality and customer satisfaction. We haven’t yet been able to quantify customer satisfaction as accurately as we’d like, but we still address it as an issue and identify needs.
Q: How has this approach been received by the physicians, given that you are basically changing their practice?
A: We’ve left the options somewhat open to give the physician some flexibility and to protect clinical quality and safety, and the physicians have ensured that there has been no adverse impact on patients. In general, doctors don’t want to be told how to practice medicine, so you walk a fine line in suggesting that they should order this or that test. That’s why it’s important they we had a physician directly involved to help us drive this process. He was interested in Six Sigma and worked closely with the assistant director.
There have actually been positive implications beyond the boundaries of the original project scope. This project has made physicians stop and look at standing orders for other areas in the hospital. The other two hospitals in our system don’t deliver babies, but they do perform lab tests. We’ve looked to see if we can transfer the learnings to other areas. For instance, we had a lab director at Franklin hospital – she was a GB in the class – who looked at this project, took the overall idea and went back to her department to see where changes could be implemented.
Q: Are there any other clinical projects at Commonwealth Health you could share with us?
A: There’s an interesting project underway that is being led by an infection control nurse to look at infections in surgical site cases for patients having bowel procedures. The nurse collected quite a bit of data on different patients and how they respond. Physicians may require different preparatory steps for patients to go through prior to surgery. There has been a certain amount of variability in how each doctor preps the patient, and we wanted to look at every opportunity to decrease the incidence of post-op surgical infections.
For the purpose of the project, we looked at any patient who develops an infection within 30 days of surgery. We started with an initial “z” score of 1.9, and collected 61 data points based on the day patients developed the infection. The nurse found that within the data, there were 40 zeros indicating patients who hadn’t developed an infection, and in some cases infection developed up to 10 days after surgery.
The team used a process map showing the day of the week, surgical prep, room, antiobiotics, hibiclin bath, type of drape used on patient, how they were irrigated, type of oxygen used…it was really a very thorough study. They came up with a significant causal factor in the specific type of dressing applied post-op – whether it was dry, betadyne, bacitracin, bactotroban and so on. The difference in those types of dressings was found to be one of the critical “x’s” in the process.
Another factor was O2 usage. The nurse tested the type and concentration, whether it was above or below 40 percent. Using a hibiclens bath as pre-op to surgery was also found to be an important factor.
Working with one of the physicians, the nurse sought to verify whether or not these were in fact the most the significant factors to focus on within the project. To achieve buy-in and create change, you certainly want to present to docs information that is statistically significant. Some of the things found have been backed up by research in medical journals, so the medical literature seems to support the project findings.
We were going to send a letter to surgeons summarizing the recommendations and indicating that if you are treating certain patients, please use the hibiclens bath, don’t use betadyne. We decided not to mandate this, but recommend the change as part of the new SOPs. Usually, if they hear it from another well-respected physician, they’ll more than likely heed their advice, although there may be exceptions.
This particular project has just completed the R2 phase and we’re using it as a pilot. We’ll collect more data points to test the validity of our assumptions and determine if physicians are following the recommendations and whether infections have decreased as a result.
Q: That sounds like an interesting and medically significant project. It will be important to follow up with you once this project has entered the control phase. I understand you also launched a pulmonary project at Commonwealth?
A: Yes, we actually had about 15 projects focused on pulmonary DRGs, since the benchmark data we were using showed that CHC had slightly higher costs for some of the common DRGs involving pneumonia, CRPD and so on.
In one project a pharmacist is the Green Belt and works in a small pharmacy outside the surgical suites. This project targeted the potential for switching antibiotic administration from IV to PO (intravenous to mouth) in pneumonia patients. The goal is to decrease the amount of time the patient is on IV antibiotics, since the longer one is on an IV, the skin begins to breaks down and it’s uncomfortable for patients. There is also a greater risk of infection.
The pharmacist measured patients to see how long they were on IV therapy, and began with a “z” score of .7. The upper spec limit was set at 72 hours and anything greater than this was considered a defect. We determined that if a patient was able to eat food and could take medications by mouth, but were still on IV antibiotic, we’d need to move them away from this and toward PO administration.
We’re piloting the process on a couple of units now and we’ve found that one was falling out as to being on antibiotic longer than desired. We review charts daily to see if the patient is able to be on PO and show signs of improvement. Where appropriate, the charts would be updated to indicate that the patient is taking medications by mouth and the attending physician/nurse should consider removing the IV and placing the patient on PO. As part of the project, we would monitor how long it took to remove patients from the IV.
Data was collected from August 1, 2000 to November 30, 2000, and analysis demonstrated that they had made a measurable difference. They were able to decrease the length of antibiotic therapy by .75 days. On average patients were on IV antibiotic for roughly 4.5 days or 6700 minutes. We were able to reduce that by 3/4 day by monitoring charts and working with docs and asking them to consider removing the patient from IV earlier when they’re ready. This change made a difference both in terms of dollars saved as well as less discomfort and risk to the patient, while maintaining quality and outcomes.
Q: With the experience you’ve gained in Six Sigma, what are your thoughts about the evolution of the approach in medicine and the expansion into new areas?
A: From my perspective at CHC, I think we’re really in a transitional period with Six Sigma, now that we’re four years into the process. Early on, we focused primarily on culture, training and communication. We’ve had Green Belts working on a variety of projects part time, but we’re now adding full-time people into the initiative. We’re also trying to make sure we focus on the right areas and align projects with metrics that feeds into our overall objectives for the organization.
Moving forward and looking at projects on the horizon, we may do a project on blood transfusion documentation. And another project led by the lab director will be to look at the success rate for obtaining blood from patients. Basically, the tech comes up to the patient’s room and the policy is to try twice before they go back downstairs and another tech comes up to try the procedure.
Our medical director is a Green Belt and he wants to launch several clinical projects. He’s talking about one project involving the wound center (hyperbaric chamber, oxygen on wound etc.) so there are many opportunities to apply the methodology and expand our results throughout the organization.
Thanks, Nick, for sharing this valuable information regarding clinical improvement with us and bringing us up to date on the Six Sigma initiative at Commonwealth Health Corporation.