Through the use of value stream analysis and process design methodologies, investigators were able to reduce costs and significantly improve outcomes of patients with lung cancer who underwent video-assisted thoracic surgery.
Nabil P. Rizk, MD, MS, MPH
Through the use of value stream analysis and process design methodologies, investigators were able to reduce costs and significantly improve outcomes of patients with lung cancer who underwent video-assisted thoracic surgery (VATS), without sacrificing quality of care, according to Nabil P. Rizk, MD, MS, MPH.
Investigators conducted a retrospective review of perioperative practices between January 2015 and March 2016 for patients undergoing VATS lobectomy. They used an outcomes database to identify cost drivers, practice variability, and rent seeking. Utilizing these data, investigators implemented a process redesign with constant review and a formal value stream re-analysis every 6 months over the course of 2 years.
Results showed that these efforts led to an overall 187% reduction of time in the operating room (OR), going from 297 minutes to 159 minutes. The redesign also led to substantially fewer chest x-rays per patient (mean, 6.7 vs 2), laboratory draws (100% vs 5.7%), and consultations (100% vs 5.7%); this translated to a 234% reduction in the mean length of stay, from 4.4 days to 1.88 days, and an overall cost reduction of 40%.
"Ultimately, we found that we could cut 40% of the costs to providing a lobectomy for a patient over the course of 2 years; this is pretty significant and could potentially serve as a template for other service lines to start to look at these methodologies," said study author Rizk. "The way we did that was by performing value stream mapping and systematically eliminating different steps. We were able to do this without any sort of downstream consequences with regard to quality."
Notably, these changes led to a decrease in pulmonary (16.9% vs 8.6%) and cardiac (13.2% vs 8.6%) complications, as well as readmission rates (13.6% vs 2.9%).
In an interview with OncLive, Rizk, chief of the Division of Thoracic Surgery at Hackensack University Medical Center and co-director of Thoracic Oncology at John Theurer Cancer Center, discussed the process redesign further, explained how these procedure redesigns could be implemented at other health systems, and shed light on how this approach could help hospitals that are struggling with the increased number of patients during the COVID-19 pandemic.
OncLive: Could you start by providing an overview on the utility of VATS lung lobectomy in lung cancer and how it compares with other approaches?
Rizk: There are 3 basic approaches to performing lung cancer surgery. One is the traditional approach, a thoracotomy, which involves a 4 to 5-inch incision where you spread the ribs, usually between the fifth and sixth ribs. Close to the majority [of these procedures] are done that way in the United States. The second approach is VATS, which has been around for at least 20 years; it has been more popularized but has plateaued over the past few years. Most recently, there have been robotic-assisted lobectomies, also called RATS, which uses the robot as another means to do minimally invasive approaches. Usually, VATS has 3 incisions and RATS usually has 4 smaller incisions and 1 slightly larger incision. VATS and RATS are basically the same. The benefits of these minimally invasive approaches typically include shorter recovery times and less pain [for patients]. They have shown through several studies to be equivalent cancer operations so there is no real compromise on the cancer side of things. As such, in general, most people would consider [these approaches] to be beneficial, without many, if any, disadvantages.
Could you speak to the previous review of outcomes of patients who received VATs lung lobectomy that was done in 2015?
We have 2 basic groups of patients. We have the control group, or those who were done in 2015 to early 2016, in which we retrospectively looked at our results. We did a certain number of lobectomies, about 44, during that period of time. We wanted to see what our outcomes were, length of stays, complications, OR time, and most importantly, cost data. We were not particularly satisfied with what we saw.
Our length of stay was not much different than the expected length of stay, but our OR times were longer than we thought. Surgeons tend to underestimate how long they are in the OR and, in particular, our costs were significantly higher than had been expected. All those issues made us take pause and decide to try and improve this. What we did over the next couple of years was then the experimental group.
What were some of the key takeaways from that review and how did that lead to the process redesign?
There are 2 aspects to it: one was the cost aspect and the other was the outcome and resource utilization aspect. In regard to the cost, about 55% of costs for video-assisted lobectomies were in the OR and 45% were after surgery. Within the OR, the primary cost is actually time in the OR. The estimate, at least at our hospital, is about $25 per minute; as such, it is very expensive time in the OR and anything you do that adds to that time, significantly adds to the cost of the operation.
Also, some of the instruments that we use in the OR can be particularly expensive. Staplers are expensive, but you cannot do without them. [Some of the] energy devices you could, in theory, do without; this could potentially be advantageous, but not necessarily so. We weighed this in our analysis. Other small procedures are done while in the OR which can also contribute to both time and cost.
[For example], arterial lines, which have routinely been done for lobectomies to measure blood pressure, were being routinely placed in our patients. The other thing we noticed was that in approximately one-quarter of those patients, they would draw blood from the arterial line and send it for a blood gas. This was never acted on; it was more of a knee-jerk thing by the anesthesiologist, which added to cost and time. Then we eventually showed that it did not really benefit patients. These are some of the systemic things we looked at in the OR. Post-operatively, the most expensive aspect of hospitalization is the length of stay; every day in the hospital is very expensive, for the bed that the patient is using and then for the bed that can be used for another patient. That is the primary source of cost, but other things like laboratories, consultations, and x-rays that all add to the cost.
Beyond cost, we also looked at the quality and outcomes side of things. In regard to quality, you are looking at complications, length of stay, patient satisfaction, pain scores and other variables. We looked at all those variables in that initial control group and although the length of stay was equal to what the national benchmarks were, they were certainly longer than we thought we were keeping our patients. Again, a lot of what surgeons perceive as their outcomes are not necessarily what reality would show when you read it closely and looked at it in a granular way.
We did that and then we did what is called a value stream mapping; essentially, you map out everything that you do for patients. You take your routine lobectomies, you see what you do in the OR, what you do post-operatively, and then you start systematically looking at every aspect of that value stream map, and you ask yourself, "Does it add value?" Meaning by doing this, are you benefiting the patient? Are you unnecessarily adding cost without benefit to the patient? We looked at every aspect of what we did and systematically eliminated different aspects to see whether we can maintain similar, if not, better outcomes while at the same time perhaps lowering costs.
Could you expand on what some of those process redesign efforts looked like?
Many process redesign methodologies are out there; they are mostly used in industry. They have been used in healthcare, usually more from a top-down approach. For example, someone in the administration decides that they want to lower costs and length of stay, so they dictate certain things and physicians tend to listen [and maybe make adjustments]. However, because the types of things they usually recommend are not completely clinically relevant, they are not very much adhered to.
What we did was a bottom-up approach. We decided what steps we would address and which methodologies we would use and then using those, we were able to accomplish what we did. Something called Lean and Six Sigma are 2 generic things that we used. Lean, essentially, ensures that there is absolutely no waste in what we do; as such, everything that you do adds value to the care of the patient. The primary goal is value and you want to ensure that everything that gets done adds value and what does not add value, you get rid of.
In order to do that kind of thing in healthcare, you have to be somewhat agnostic to what people have been doing forever. For example, with doing lobectomies, people have pretty strongly held beliefs about what adds value to the care of the patient; however, in reality, there are no real evidence to show that. We essentially generated evidence through databases by looking at outcomes and then systematically generated new evidence to question whether it works. Step by step, it eliminated many things; it became very lean. Six Sigma looks at it in a bit of a different way, it looks more at complications and outcomes and seeks to minimize errors. While trying to improve value, which is quality divided by cost, we always focus on quality, and in doing so, minimize any potential for complications. Our outcomes reflected that.
What were the results of the analysis?
We found that we could cut 40% of the costs to providing a lobectomy for a patient and we were able to do this without negatively impacting quality. Our cardiac, respiratory, and renal complications all actually decreased; they were all relatively low, and certainly lower than what the Society of Thoracic Surgeons standard benchmarks are. As importantly, our length of stay went down significantly, as well; it started above 4 days per lobectomy to under 2, and the median was 1 day. That equates to significant savings with regard to length of stay in the hospital.
We also, for example, eliminated patient-controlled analgesia because that required an anesthesia consult. However, we also monitored our patient's pain scores and we did a lot of preemptive pain management and our patient's pain scores actually improved. Overall, quality improved significantly, although it was not statistically significant because the numbers are not large enough to achieve that. It was done at the same time as lowering costs significantly, so it showed that you can achieve both at the same time.
Do you feel that these process redesign methodologies are easily replicable in other institutions and health systems?
Absolutely. That was the principal goal in doing this in addition to benefiting our patients. [We wanted] to show others that these are the methods we used, the specifics of each method. The value stream mapping is obviously going to differ from service line to service line, but this is how you can use evidence-based methods to generate really rapid change in a way that can lower costs and improve quality. As such, there is no reason that this cannot be done in any specialty. I am confident that anyone can use this as a template.
Are there any other emerging efforts that you're looking into to further improve outcomes?
We have now moved on to robotics. We started this about 1 year ago, doing robotic lobectomies, and we are achieving similar results to what our final group in video-assisted achieved. However, our cancer operation is maybe a little bit better; that’s principally due to some of the technical capabilities of the robot. Ultimately, we will probably achieve even better lengths of stays and maybe better quality, as far as outcomes for patients. We are starting to do a lot of segmentectomies, for instance, which are sublobar resections for lung cancer. Again, there is not a lot of evidence saying what is best. Especially in the era of COVID-19, when you are trying to spare lung capacity for patients who have lung cancer, if you can do less yet achieve similar cancer outcomes, then it is definitely worth trying. Robots allow us to do that much more easily than video-assisted does.
As such, about half of our lung cancer operations have undergone segmentectomies. The additional benefit of segmentectomies is almost every one of those patients is going home in 1 day. Probably within a year or 2, I will be comfortable saying we could send the majority of those patients home same day. It is definitely pushing the envelope, but we are doing it in a way that we are monitoring everything. As long as it is being done safely, and patients do better, we are ultimately going to continue to push the envelope.
What is your take-home message for your colleagues regarding this research?
The biggest misconception that many people have is that costs and how much you do to patients directly correlates with outcomes. However, there is actually an indirect relationship many times. The more we do to patients, in general, the more the potential for causing harm. For example, putting a Foley catheter in at the time of surgery because we have always done it; this should only be done because of the length of the procedure, not because of the type of procedure.
For VATS, everyone still utilizes a Foley catheter; there is a consequence to that. You can potentially cause a urinary tract infection or urinary tension in some patients and so that is something that you can eliminate, or do less, and still have better outcomes. Again, many of the things we do when we provide care to our patients, do not add to the value of the care that we provide. Once you start really looking critically at what you do, you can start to realize that this is the case. There is a lot of that in the system, to be honest.
If we can cut 40% of costs in 1 service line, then I estimate that we can probably do the same thing across many service lines. Imagine what that would look like in healthcare economics in the United States, which is already extremely expensive. The argument I am making is that we probably overspend to achieve no better quality than other parts in the world because we are essentially overspending, and we do not need to spend that much.
Is there anything else I didn't ask you about that you would like to highlight? Or any other research you're working on that you wanted to share?
The timing of this publication in the midst of the COVID-19 pandemic is actually fortuitous, but one of the problems I foresee that hospitals are going to have to deal with going forward, with the whole issue of social distancing and keeping patients safe in the hospital, and make patients feel comfortable about going back to the hospital. At our hospital, many of the beds are 2 patients per room. I cannot imagine that going forward that that is going to be the norm. I suspect we are going to go to private rooms to keep patients a bit safer. As such, imagine the issue with many hospitals who most of their rooms are currently 2 patients per room. How are you going to do in a patient census if you ever get back to the volumes you were at previously?
The second problem is, New York is mandating that 30% of their beds remain empty in the hospital because of the need to be prepared for the next surge of patients with COVID-19. If you are telling hospitals they need to be at 70% capacity, and most patients are going to demand that they have private rooms, where are we going to house patients? One way to fix that problem is to create bed days by shortening the length of stay. Our service line, for instance, last year saved about 600 bed days if you take lobectomies and other procedures that we did, relative to what it would have been perhaps at other hospitals.
If every service line does that, then maybe you do not need to create new space; you can more efficiently use the space that you have. That, to me, is something the whole healthcare system can look at over the course of the next year, as we are being forced to do so. We have shown in the past months that we can be very innovative and create many solutions to major problems. This will be the next major problem and there is a way to fix that problem very rapidly and safely and improve quality at the same time. The timing is somewhat fortuitous, and we can say that you can do this, and this is the way you can do it.
Yeo JH, Shariati NM, Pelz GB, et al. Lean lobectomy: streamlining video-assisted lobectomy to increase the value of lung cancer care [published online April 30, 2020]. JCO Oncol Pract. 2020. doi:10.1200/JOP.19.00590