Baptist Health South Florida, a six-hospital system in South Florida, implemented a multidisciplinary low-dose CT lung cancer screening program in July of 2014, spearheaded by 72-radiologist practice Radiology Associates of South Florida, Miami, and Juan Batlle, MD, director of the program. Initiated prior to the era of reimbursement for the service, eligible patients were required to pay just $35 for the study.
As Batlle puts it, “We went from zero to sixty fairly quickly, and we had to adjust to that.” The program, which institutes a standardized approach across the system with one database, screened between 700 and 800 patients in year one and is on track to screen 1500 this year.
While there are many operational decisions that must be made in developing a lung cancer screening program, many of the key clinical decisions revolve around nodule management. “Images were being performed at outpatient facilities across Miami Dade and Broward, as far north as Cold Springs and as far south as Homestead Hospital,” he explains. “We really had to create a virtual system for a nodule clinic.” In a recent interview with Radiology Business Journal, Batlle describes how the program handles nodules.
How prevalent are nodules in the screening population?
Batlle: Lung screening is something that has been happening academically for more than 20 years. That screening has been happening in very heterogeneous populations, that is non-smokers, heavy smokers and everyone in between.
The numbers that you get for “false positive” rates range from 5%-10% all the way up to 30%.
With the National Lung Screening Trial, it was somewhere in the 20% to 25% range, but subsequent use of the ACR’s Lung-RADS™ structured reporting system dramatically lowered the false-positive nodule detection to the 10% to 15% range. We find that our “false positive” nodule-detection rate is somewhere in the 10% range—fairly low, and it’s usually lower than 10%. What ACR’s Lung-RADS did to push that number down was the creation of category 2, a benign category, with no need to follow up, just return to annual screening. That included all nodules up to 6 mm in size. If you find a nodule, and the average of two perpendicular measurements is smaller than 6 mm, then that nodule can be called a Lung-RADS 2 and it can be followed conservatively.
What that means is that these tiny nodules which were frequently found and over-called on previous exams no longer require any follow-up whatsoever apart from the annual CT exam. That eliminates 15% to 20% of patients who otherwise would have been called positive, and would have been called back for a follow-up scan and no longer are. We adhere strictly to that.
How are nodules assessed?
Batlle: I make sure I measure the nodules faithfully: a long nodule that measures 7 mm but is only 1-mm wide is not a 6–plus-mm nodule. The average of tdhose measurements is below 6 mm.
It’s important to make sure that you are measuring accurately, and that you faithfully go through the entire record to see if you can find a previous CT scan that includes that area to try and clear that nodule as benign and stable. I attempt to look for other signs of benign nature, like central calcification or complete calcification of the nodule, fat within the nodule, or characteristic features that may help
you attribute it to an infection, like other small clustered nodules in the area. These are all things I go through to make sure that when I call a nodule back, one that needs a six-month follow-up scan, that is done very rarely, less than 10% of the time.
The other thing that is useful from the ACR Lung-RADS perspective is that in its Lung-RADS 3 category, what they recommend is one additional 6-month scan. If a person is getting their scan January 1, 2016, and they are due to have the next scan in January 2017, the only extra scan they are going to get is July 2016. Then they go back to a yearly schedule: July 2017, July 2018. That “false positive” creates just one additional low dose CT scan.
I think it’s a very straightforward, streamlined, economical way that the ACR has helped develop to help us make decisions for patients that don’t overly worry them. The vast majority of our patients, 90% or more, are told come back in a year.
Which nodules are biopsied?
Batlle: Again, we follow the ACR’s Lung-RADS. Once the nodule gets to 8-mm or more, we look to characterize it with a PET scan and see if there is metabolic activity. If there is metabolic activity then that is a patient who would be referred, again, with our multidisciplinary panel, to some kind of intervention, and these days, there’s more than one way to intervene with a nodule.
There is interventional bronchoscopy: A pulmonologist navigates to and biopsies that nodule. There’s the percutaneous interventional-radiologist approach through the skin, and a needle into the nodule. There’s a path straight to surgery.
There may be a solitary nodule that is suspicious for lung cancer, say a 12-mm spiculated nodule and the multidisciplinary panel agrees that it looks suspicious. That’s a patient that might go straight to surgery, remove that nodule and not only is it diagnosed, but if it’s a cancer it is cured. A stage 1 lung cancer may be treated with surgery and absolutely nothing else.
Because of the variety of options, we felt it was important to have that multidisciplinary panel so that there is agreement and all of the people are at the table to help decide for that particular patient: “I would agree that this is a bronchoscopically approachable lesion, what do you think,” and allow that kind of discussion between the specialties.
Our general pathway follows the American College of Chest Physicians and the NCCN guidelines. For 8-mm nodules that are metabolically active, we do seek some some sort of histological sampling of some kind. For those that are metabolically negative, we open the door to continued follow up with CT scan. If a PET scan shows that the nodule is not active, that allows us a little more leeway in going with a conservative approach and not going for histology right away unless a CT scan shows growth.
What are you using to measure nodules? How do you manage inter-observer variability and inter-vendor discrepancies in nodule assessment?
Batlle: That’s a valid concern because its been shown that nodules can be measured at different sizes just based on how the patient held his or her breath. It may actually be impossible to have perfect inter-observer measurement.
We have one PACS to measure nodules. We have a very small subset of radiological readers. That’s very important. If your group had a heterogeneous population of readers, and mammography was newly developed, you might not choose to dump mammographic screening on all of them and say, “Now let’s all start reading mammograms.”
A similar process has occurred with CT lung screening: precision, faithfulness to the database, faithfulness to the process, knowledge of the process, and because of that, we have restricted the number of readers to a small handful who will read the lung screens. I think its important for the person reading the scan to really know the background of how lung cancer screening is done and how it should be done properly, so that the best outcome for the patient is achieved.
We use calipers, we measure in two dimensions, we average those dimensions, and then we faithfully compare to the previous exam—and, we zoom when needed. For small nodules, zooming onto the nodule to get the precise borders is very important for measuring the same thing the same way.
That’s the overview of how we do our nodule measurement. We don’t currently use any computer-assisted detection or any automated tools to give us volumes of the nodules, although that is certainly a promising thing in the future to reduce variability.
How successful are you at getting patients back for nodule follow-up? Any tips on making that happen?
Batlle: We are successful in getting patients back. We have very good compliance, and the keys to that are good relationships with referring physicians: making sure that they are well taken care of in terms of reporting important results, suspicious cases, and having a verbal communication between the radiologist and the physician.
I don’t contact them with every benign case, but with suspicious cases, I reach out to them and I speak to them and I thank them for referring the patient for lung screening. Hopefully that helps keep lung screening at the top of their minds, so when their eligible patients return, they are referred back.
We also contact the patients directly: Our nurse navigators call all of the patients before the initial scan, they also call with the results of that initial scan, assuming it’s on the benign side—we leave patient contact up to the referring physician if it is a suspicious Lung-RADS 4 nodule.
When they are due for the next exam, we have the navigators call that patient again, and we make sure that navigator has multiple ways of contacting that patient. We collect their email address, their cell phone numbers, their home address and their home number. We have multiple tools to get ahold of that person and bring them back in for a scan. We also have kept our price very low.
When are patients dismissed from the program?
Batlle: There are a variety of ways that a patient can be dismissed from the program. Number one, sometimes we don’t want them to be discharged, but they discharge themselves by failing to respond or declining a follow-up exam. After a certain period of trying to contact them, we discharge them from the program.
Then there are patients who “age out”. They hit 81 years old, because the USPSTF criteria are 55 to 80. Or, they quit smoking for more than 15 years. Once that patient hits 16 years, I call that graduating the program because their risk of lung cancer has now reached that of a non-smoker—it takes 15 or so years for a former smoker’s lung-cancer risk to return to normal.
That’s a controversial requirement. From a smoker’s point of view, and perhaps rightly so, if they smoked 30 pack-years, and they qualified at some point for lung screening, their anxiousness to continue lung screening annually is high; it’s sometimes hard to convey to patients that their risk has returned to baseline. It also is difficult to measure because it is purely self reported. I’d love to see that loosened in the future, as well as inclusion of younger patients and patients with CT-proven emphysema, but that is what we have now.
Editor’s Note: Visit www.thelungspot.com for more information about the lung cancer screening program Radiology Associates of South Florida helped to implement at Baptist Health South Florida.