The Rise of Specialist-Driven Alternative Payment Models in American Medicine

Narrator:

You're listening to ReachMD, and this is Inside Medicare's New Payment System, produced in partnership with the American Medical Association. This podcast was produced before final regulations for the quality payment program created by the Medicare Access and CHIP Reauthorization Act (MACRA) were released. Visit the AMA website for the latest news and more details on Medicare’s new quality payment program.  

Dr. Birnholz:

The advancement of the Medicare Access and CHIP Reauthorization Act, or MACRA, has catapulted alternative payment models into the spotlight for identifying new value-based approaches to care.  My guest today is leading one such care model which applies an innovative method analogous in many ways to sonar detection for tracking patients between clinic visits. 

This is ReachMD, and I'm Dr. Matt Birnholz.  I'm joined today by Dr. Lawrence Kosinski, practicing gastroenterologist and founder and Chief Medical Officer for SonarMD, a web-based platform that pings patients beyond practice settings to help get ahead of issues before they become emergencies.  Dr. Kosinski presented this model at the AMA House of Delegates meeting in June. 

Dr. Kosinski, welcome to the program. 

Dr. Kosinski:

Thank you very much for having me today. 

Dr. Birnholz:

So, to start, I've given a basic rundown of SonarMD from what I understand of it, which is limited, but help us understand what the SonarMD platform is and how it was inspired from the vantage point of your GI practice? 

Dr. Kosinski:

Thank you.  SonarMD is actually a platform that took some time to develop.  We initially began back in late 2011 looking for a way of bringing gastroenterologists into the value-based platforms.  Gastroenterologists, like so many specialists, have been highly dependent on fee-for-service driven colonoscopy reimbursements, and with today's changing environment, it was imperative that we looked elsewhere, that we looked to value-based initiatives to see if we could position our gastroenterologists for future practice. 

For the next large category of revenue that comes into a GI practice is for the treatment of patients with inflammatory bowel disease.  Twenty-five percent of our practice revenue in our practice -- and we are the largest gastroenterology practice in Illinois -- comes from the treatment of patients with IBD, so this seemed like the next best place to start focusing our efforts.  We realized, though, that we needed a lot more information and sought claims data from our major payer, Blue Cross Blue Shield Illinois.  After some time of discussions, they provided us with a very large claims database.  This was back at the beginning of 2012, so they provided us 2010 and 2011 claims database on their entire Crohn's population. 

Analysis of this led to Project Sonar, because it was very clear in looking at this database, which was quite large, 21,000 patients, that there was excess cost being spent for the treatment of complications.  How do we avoid these complications?  How do we intervene with these patients before these complications ensue?  One of the findings we had in our claims review was that less than a third of the patients who went into a hospital for a complication of this disease had any contact with their provider in the 30 days prior.  We couldn't find a CPT code on two-thirds of the patients, and yet, over half of the cost of care was for the treatment of complications, so we felt we needed to create a sonar system, a way of intervening with the patients in between their face-to-face visits so that we could intervene in their care long before they even realized they needed it.  That's what started Sonar. 

Dr. Birnholz:

Fascinating.  And how then did Sonar work?  What was the platform that it was based on for helping to reduce this complication rate? 

Dr. Kosinski:

Well, again, this was a learning process on our part.  We initially thought we needed some structured questions to send to the patients, so we pulled from a very accepted patient-reported outcomes questionnaire, the Crohn's Disease Activity Index, which has been around for over 30 years, and we pulled 5 questions out of it that the patients could answer, and we sent them those questions using our patient portal at the time.  And what we found was that the patients would report symptom data back, but their health-reported quality of life responses did not indicate that they sensed that there was any deterioration in their status.  So, our early findings confirmed the fact that using structured information, structured questions of symptom data, provided more information to us rather than just asking the patient how they were doing, because they didn't realize they were deteriorating. 

Now, we realized early on that the portal was very cumbersome.  Patient portals are wonderful, but they're really not designed at this point to be bringing back structured data and then calculating Sonar scores. So I partnered with a local company here in Chicago, and we developed a web-based platform that the patients could access via their Smartphones.  It's not an app.  There's nothing that resides on the device, and we are device agnostic.  All the device has to have is web-based access, which almost all of the Smartphones have today.  And so the patients receive what we call a ping, which is a set of 5 questions.  They receive them on the first Monday of every month.  It takes them no more than 30 to 60 seconds to do this.  We then get the response back from them in the form of structured answers that are validated and bring us back a score, which is a symptom score that we can monitor.  We have a team of nurse care managers inside our practice that then monitor these responses via a desktop dashboard, and the patient gets an immediate feedback after they fill out their survey.  The nurse care manager gets that same immediate feedback.  And before the patient even calls the office, 9 times out of 10 the nurse is calling the patient ahead.  So, we have been able to intervene, and we have been able to significantly save costs for the payers.  Most importantly, though, we're keeping the patients out of the hospital.  We're keeping them from having those complications.  They're remaining healthier.  The payer is happy.  We're happy. 

Dr. Birnholz:

Interesting.  And I understand that there were some challenges up front trying to get patients on board with participating in the program consistently, but you since turned that around into an almost unheard of compliance rate.  Tell me a little bit about that. 

Dr. Kosinski:

There are two parts to your question, and the first part is how do we engage the patients, get them to enroll?  I can't emphasize more how important it is for the patient's physician to be the one that conveys to the patient the need to enroll in the program.  Our nurse care managers do a wonderful job, but many times they faced reluctance from the patients, but one phone call from a physician and it usually turned around.  But with the patient portal, the second part of the question, maintaining that engagement.  The patient portal was very cumbersome for us to use.  We could never get more than the 27% sustained response rate, and that's why we went to the web-based platform.  But with this web-based platform, we have an 80% sustained response rate of our patients.  Now, it's not an automatic occurrence.  The patients all get, as I said, their ping on the first Monday of every month, and about 60, 65% of them respond to that first ping.  If they do not respond, they get a second one the second Monday of the month, and if they don't respond to that one, then the nurses have on their dashboard a tab for the patients who have not responded.  They then work that list.  So, it still requires some human input, but the web-based platform really has saved a lot of time and energy for our nurse care managers, and in today's cost environment, that's extremely important. 

Dr. Birnholz:

And I want to focus on that cost element in a minute, but the first thing I want to ask you about is:  Regarding patient care and outcomes, since that's the ultimate goal here for the value-based care, what changes have you witnessed since rolling out SonarMD? 

Dr. Kosinski:

The major thing is this.  I've been in practice for over 30 years, and like so many physicians, we will see our patients, we care about what's going on, and as they exit the office, we tell them, “I’ll see you in 3 months,” or 6 months or whatever it is, and, “call me if you have an issue.”  We’re relying on the fact that the patient will recognize that they need to call us.  Many times, if not most of the time, a patient with a chronic disease, especially a young one, is busy.  Their lives are as hectic as ours, and they learn to accept the symptoms they have and they don't always see when those symptoms are intensifying.  So, from a patient care point of view, what the Sonar platform has allowed us to do is to hover over them in a nonintrusive way that they don't mind participating with, and we get data back on a regular basis as to how they're doing.  And what typically has happened is, if we see a change in the scores, if we see it -- we plot the slope of their scores as well -- if we see a change in the slope of their scores, we then intervene.  The nurse will call the patient, or sometimes the physician will call the patient, and we have made medication changes frequently.  We have brought the patients in and intervened, changed therapies, and, as a result, we've markedly decreased their morbidity, the hospitalizations associated with the morbidity, and the ensuing cost. 

Dr. Birnholz:

Well, Dr. Kosinski, just to get on the granular level and understand the specifics of this type of questionnaire system that you’ve developed, what are some examples of questions that get pinged to the patient in your language? 

Dr. Kosinski:

We pulled questions from the Crohn's Disease Activity Index.  Now, the Crohn's Disease Activity Index, CDAI, has been a research tool for many years, and it is composed of multiple symptom-based questions combined with findings that a physician might elicit in a face-to-face visit.  We pulled from the questions only so that we could send them to patients without having to have them in the presence of their provider.  And they're very straightforward questions, first one of which is:  How many bowel movements do you have per day for the last 7 days?  And they have structured answers.  They have checkboxes that they can fill in.  Rate your abdominal pain.  There's a rating scale on that.  There is another question that basically is a (inaudible)* 11:58 answer.  Did you have joint pains?  Do you have any eye pain?  Do you have a skin rash?  Did you have fever?  Are you taking any drugs for diarrhea?  The final question is a health-reported quality of life question, basically a “how are you doing” question.  How do you feel?  What is your basic state of well-being?  And they have several potential answers there. 

The fascinating thing on this in analyzing how the patients respond is that they don't track together.  The health-reported quality of life question appears to track independent of the symptom questions in a very significant percentage of the patients so that patients will be noting that they're having more bowel movements, more abdominal pain, but their state of well-being hasn't changed.  And yet, there are other patients, and these are usually the patients who are deemed depressed at enrollment, will consistently have high abnormal scores on their health-reported quality of life question, but their symptom scores are low. 

So, it's interesting that the entire CDAI score over the course the last 30 years has come into contention by some researchers that it doesn't parallel completely what's happening to the patient, but when you dissect down to how the individual questions are being answered on a regular basis, you can start seeing what is driving the ultimate score. 

Dr. Birnholz:

Well, if you're just joining us, this is ReachMD, and I'm Dr. Matt Birnholz.  I'm speaking with Dr. Lawrence Kosinski.  He's founder and Chief Medical Officer for SonarMD, which is an Alternative Payment Model presented at the AMA's House of Delegates meeting in June. 

So, Dr. Kosinski, let me turn then to the cost side.  What has your experience been so far with insurers, such as Blue Cross Blue Shield of Illinois that you mentioned, to help make this model work financially? 

Dr. Kosinski:

They have been tremendous with us.  I have never thought in 30 years of practice that I would have a relationship with my payer like I have today, but we couldn't have succeeded without their information.  Physicians many times suffered group think amongst each other because we focus on our costs.  We don't see where the tremendous expenditures are being made in the care of our patients.  Looking at the Crohn's database, gastroenterologists only see 3.5 cents on the dollar that the payer is spending, so in our practices we might not think that inflammatory bowel disease is that expensive of a condition because we're only seeing the tail of the elephant and the elephants have very small tails, so we needed to see this data from the payer, and we have complete data.  We have every claim, every Pharma claim.  We can see where the costs are.  You give a group of doctors who are trying to do a good job access to data that helps them do a better job and we can make a difference.  We, early on, were able to move patients to lower cost sites of care for their infusions.  That was the low-hanging fruit.  But the most important component of our treatment is our appropriate use of biologic medications.  And we have increased our use of biologics in our patient population, but the overall savings from the morbidity from the disease have more than covered the increased cost. 

We did an analysis of our cost data in preparation of one of our national meetings because we had submitted an abstract that was accepted, and so we normalized our cost using Medicare payments instead of commercial payments so that we could eliminate site of service differentials, and after utilizing that normalization process, we were able to show a 9.8% decrease in the overall cost of care, which was mostly driven by over a 50% decline in inpatient expenditures.  We actually spent more for our biologics but, again, that was more than compensated for by the declines in treatment of complications. 

Dr. Birnholz:

And I'm sure a number of our listeners right now, both generalists and specialists, are listening to this and thinking this really sounds analogous to my patient base.  For whatever the chronic disease might be for that patient population, they're thinking my patients act in a very similar way and complications seem to be one of the largest cost drivers here, so there are definitely ramifications across the board.  But given that this particular model that you've helped develop came from a specialist-driven perspective, I imagine that as far as MACRA is concerned, you must have come across some challenges in trying to advance this, and a number of our listeners are probably thinking along this line for a specialist-driven APM.  What can you tell me a little bit about that in your experience with MACRA as a specialist? 

Dr. Kosinski:

Yes, you have your finger right on the pulse of one of our major challenges as specialists today.  The MACRA, at least our proposed role that we have so far been shown, really does not allow specialists to perform as APMs, as Alternative Payment Models.  We have to be part of a much larger entity such as an ACO.  But, there are very definite reasons to consider specialty-based APMs, and I'm going to pull from a term that stockbrokers might be very familiar with, something called high beta.  There are certain stocks that are called high beta stocks, and their variability in their stock price far exceeds the variability of the S&P 500.  There are illnesses that are also analogously can be called high beta illnesses.  Crohn's disease, ulcerative colitis, are good examples, but there are others -- chronic obstructive pulmonary disease, endstage liver disease.  We can go on.  But these illnesses are very high-cost illnesses, and their cost varies tremendously with mild changes in therapy, as opposed to treating hypertensives or diabetics whose illness may take years to show the evidence of management decisions.  

So, when we focused on Crohn's -- and I'm not going to claim I had any brainstorm on this, we learned it over time -- but when we focused on Crohn's, we realized that there was a very significant variation in the practices of the providers taking care of these patients, and that variability rapidly resulted in changes in outcome of the patients.  So, I would think it is in our best interest as a medical community to identify these high beta illnesses and to urge CMS to allow the option for specialists to participate in the care of these high beta illnesses on an APM platform.  We have had discussions at the highest levels of CMS, along with multiple medical societies, to see if we can allow these changes to occur before the final rule.  Time will tell. 

Dr. Birnholz:

And I understand that you actually had a chance to sit down with CMS, specifically, Andy Slavitt as the Acting Administrator and a leading voice in the MACRA initiative.  What did he make of the SonarMD approach? 

Dr. Kosinski:

I was part of a group of several specialty society-driven initiatives.  American College of Surgeons were there, the neurosurgeons were there, the orthopaedic surgeons were there and gastroenterology.  Dr. Joel Brill and I represented the AGA, and we did meet with Mr. Slavitt.  This was precipitated by Senator Cassidy, who is the only gastroenterologist in Congress.  And so this occurred in June, and we did meet with Mr. Slavitt, and he was very receptive to our meeting and patiently listened to all the presentations of the various Societies.  Ours was the last, and I opened our discussion similar to what we have said today, that I wanted to present a specialist-driven initiative that was partnered with a major payer that could form the foundation for a specialist APM.  I presented Sonar, although I only had several minutes to do it so I had to be very selective in how I said it, but he was very receptive and appeared to agree with our position.  And, as a result, we have future meetings now that are being scheduled next month and the month after that.  Hopefully, we will prevail that when the final rule comes out there will be more flexibility. 

Dr. Birnholz:

But it seems like that position that you described among a number of specialists seems to be one of opposition to MACRA just on the grounds of being or feeling excluded from helping drive APMs.  What's your perspective on the best foot forward for helping find solutions to this? 

Dr. Kosinski:

This goes to the heart of what we all should be doing in our practices.  The organizations that are responsible for covering the cost of care, CMS being the largest, but the insurance plans as well and self-funded employers, they want single-source solutions.  They really, I think, don't want to have to find the gastroenterologist solution, the surgical solution, the pulmonary solution.  They really, from a management point of view, would much rather deal with an entity that can take the risk, can manage the care and provide the product for the patients that are engaged in it.  But when we think more granular into those large risk-based organizations that have been developing over the last few years, the ACOs being the main example, they are going to need to be able to contract out to specialist-driven entities that will be able to provide slices of the care that they have to provide in total.  I always use the analogy when someone asks me this question of the automobile industry.  I'm old enough to know that back in the '50s and '60s and most of the '70s there were not that many foreign cars.  It was all GM, Ford and Chrysler.  And if you looked in the General Motors car at that time, the battery was made by Delco, a wholly-owned subsidiary of the General Motors Corporation.  The body was made by Fisher Body, another wholly-owned subsidiary.  But when competition arose from Europe and Japan, all of a sudden that model fell apart.  And if you look at an automobile today that might be manufactured by General Motors, the parts are coming from multiple companies that answered RFPs and could provide the best value-based solution for whatever they're creating, be it an alternator or generator, whatever. 

Healthcare is going to be the same way in my opinion.  Today we have ACOs that are trying to control everything, buying up the physician base and hoping that they will be able to control the cost by controlling the distribution channel and the entire value chain.  In all likelihood, as Medicare continues to cut costs, cut payments over the next decade and inject more value-based initiatives into the payment model, our ACOs, our large integrated delivery networks that are running those ACOs, are going to have to send out RFPs and find the most value-based provider for the various slices of care they will have to put together in order to remain competitive and provide an ultimate product to their patient populations.  Our group, the Illinois Gastroenterology Group, with our aggressive initiatives at this, hope to be one of those value-based providers in the future.  From a business point of view, I think this is the way we have to focus. 

Dr. Birnholz:

Before we wrap up then, let me just turn right back to your group again and ask you, if you look down the horizon, where does SonarMD go from here? 

Dr. Kosinski:

SonarMD has already grown outside of the Illinois Gastroenterology Group, and that's why SonarMD is a separate corporation.  We have gastroenterology practices from coast to coast now who are pinging patients and trying to build the same relationships with their payers that we have with Blue Cross Blue Shield Illinois.  We have an intensive medical home with Blue Cross Blue Shield, and we receive care management payments in an effort to work as a team with our payer.  Well, SonarMD has now about 20 large practices around the country that are all pinging patients, and the data is going into a common database.  This is allowing us the ability to create a national platform for the development of a specialty APM.

Dr. Birnholz:

Well, with that I very much want to thank my guest, Dr. Lawrence Kosinski, founder and Chief Medical Officer for SonarMD. 

Dr. Kosinski, it was great having you on the program today. 

Dr. Kosinski:

Thank you very much. 

Dr. Birnholz:

To access this interview and other related content, visit ReachMD.com or download the ReachMD app.  I'm Dr. Matt Birnholz, as always, inviting you to be part of the knowledge.