On the intrinsic value of CPT codes

I would never review technologies on a code basis. Coding is strictly nomenclature.

Now, coding is a placeholder in a big fat book published by the American Medical Association or by CMS. It does not guarantee reimbursement. I couldn't care less if there's a new code for such and such procedure or technology.

Coding is nomenclature: There's a million phone numbers in the phone book. That doesn't mean I'm going to call each one of them. In other words, that doesn't mean I'm going to patronize each one of those phone numbers.

There are certain phone numbers in that phone book that you mean something to me: One of them's my parents, one of them's my friends, you know, so on and so forth. Those numbers are worth me dialing.

Just because there's a phone number in there doesn't mean I'm going to call it – doesn't mean I'm going to use it – in other words, doesn't mean I'm going to reimburse for it.

On getting paid

Since the middle of December, maybe – I bet I've had five phone calls with companies around the United States asking me my opinion about these new codes.

All of these in this email that you just forwarded me or similar – but primarily these, saying, "Hey, Nic. Hip hip, hooray! There's a new code!" and I say, "Yeah, doesn't mean it's going to get paid."

Every technology I denied coverage for? Had a code.

So, having a code does not mean you'll get reimbursed.

Not having a code can mean we'll either use an unlisted code to reimburse for you, if we like the technology, or it'll mean, "Ah, we like the technology, but we'll wait ‘til there's a specific code, then we'll reimburse for it."

Think about it as: Just because there's a code that does not mean I'll get reimbursed.

I think this is my own gut reaction, knowing CMS very well, and certainly knowing commercial payers even better than I know CMS.

These three codes that are coming out seem a little bit like a red herring to me. I think they're setting themselves up to just deny the heck out of digital health for the next little while.

And all these digital health companies are coming out right now, going, "Oh, my gosh, this is perfect timing! Here we are in the market with our new stethoscope, our new preterm birth urine test or whatever the thing is, which is all digital health, now we can get reimbursed for it because there's new codes and CMS is going to go, "Yeah, let's see what you got?"

"Oh, you don't have any studies? Denied!"

”Oh, you don't have any peer-reviewed papers in the National Association of Managed Care Physician's guideline clinical dossier? No, thank you. We ain't reimbursing for it."

There's going to be a ton of digital health companies out there that did not structure themselves is a consumer item – that you would buy on aisle nine at Walmart, and somebody who just pays $10 at the checkout stand – and goes home it; with no reimbursement needed.

They didn't structure their companies that way. They built their company as a conventional medical technology that requires CPT, HCPCS, so on and so forth, billing and coding mechanisms.

And they said, "Great! There's a code. I'll get paid!"

No you won't.

There was a code for Urolift when I denied coverage for Urolift.

There's a code for robotic surgery for pancreaticoduodenectomies when I denied coverage for pancreaticoduodenectomies.

So just because there's a code doesn't mean you're going to get paid.

On what my client needs to do now

What [client] needs to do is have a very long, hard discussion with their board of directors and their investors. "We need to go raise more money. Now."

If the company is, you know, "shoestring-ing" this, or maybe they've raised a couple million, but they certainly don't have enough to do a full-fledged study, they're going to have to, if they want to get reimbursed.

Now, maybe my gut reaction's wrong.

Maybe CMS is just going to say, "Look, digital helps here. There's no fighting against that.

For example, we assume stethoscopes improve patient outcomes. We assume when a doctor looks at a portal and sees all 60 of his patients on there, you know, half of whom had an arrhythmia yesterday, and he can now prescribe a new drug, that that's a good thing. Ipso facto, it improves outcomes."

But between me and you, that has never happened. That would be a major outlier in the trend of reimbursement for the last 20 years.

Joe Hage: Be specific when you say 'that' has never happened. What has never happened?

Nic: 'That' being, they've never carte blanche reimbursed for a code. They always demand real data.

And CPT is a little different than HCPCS.

CPT implies there is data and you've already done Phase III studies – you've already shown that current coding in CPT is not sufficient – so you've got to have a new code.

But for HCPCS...? For this HCPCS G2012, G2010 and so on – those are brand new codes.

And CMS is going to say, "Yes, we will issue a code because, currently, there is no nomenclature to capture that particular technology that needs to be reimbursed. But unless [client] brings to us an AMCP clinical dossier with – I'm making up a number – three peer-reviewed prospective studies that show clinical utility..."

Not just clinical validity; clinical validity is not sufficient for reimbursement.

"… that they have done three studies, or five or 10 studies or whatever – they're not studies but papers; you can get multiple papers out of one study – that show demonstrable improvement in patient outcomes commensurate with the price they're charging, we're not reimbursing."

So, I couldn't care less if I issue the new code. That doesn't mean you're going to get paid.

So what these guys need to do, Joe, unless it's already very, very, very well established, they need to go do a clinical study.

If CMS is not going to let them piggyback off of well-established data that using remote patient monitoring works, they have to do a study – a very well defined, acute study that matches what the code is for G2012 or G2010.

And, then, they approach CMS saying, "Hey, you have these two codes. We've done a randomized prospective study with 67 patients over 24 months. They used [client]. Look how well they turned out. Now, would you mind reimbursing G2012, G2010 whatever amount of money for those codes?"

And CMS will review that data, as will every commercial payer, and then they'll say, "Yes, you know what, we really like this data. Yes, we'll reimburse it. This is great."

Joe: They will retroactively reimburse that which happened over the previous 24 months?

Nic: Not necessarily, no.

On not covered, possibly covered, and covered

You can get reimbursed for clinical trials. You can use specific codes for that. But no, no. In fact, they may help plan to do a few things.

They could set up that code as not covered, possibly covered, covered.

Not covered is just a blatant "if that comes through in the system, we aren't paying for it. Whether they use [client] or something else, we don't care. We're just not paying." It's automatically denied.

Possibly covered would say, "You know what, we will pay for this but only if it's in patients that are over 50 years old, that have diabetes and no family history of…. Then, we'll pay."

And then covered is just blatant, "If it ever comes through, we pay for it."

On reimbursement for equipment

Well, now, if that doctor has to go sit down at a console every five days or whatever and look at all those patients and he has some new technology that is going to have in his briefcase that he's using or that he's sending home with a patient, that's a separate issue.

That's why CMS has issued these codes to say, "You know what, maybe physicians are going to buy a box of [client]'s for $20,000 and he's handing it out to all of those patients. He needs to get reimbursed for that just as though he is handing out pacemakers to his patients. And he needs to get reimbursed for interpretation of those graphs and charts every five days when he looks at them on his portal."

Joe: Tell me about that scenario and likelihood of "Yes, we need to pay him back because he handed out pacemakers."

Nic: Well, in the hospital, pacemakers are put – I'm not sure – in the outpatient setting, but the hospital would bill HCPCS code...

Joe: For the benefit of those on the call and, frankly, me, is HCPCS all about paying for the product?

Nic: For a piece of equipment. Bingo. For the pacemaker, for the stent, for the crutch, for the brace...

CPT is for the physician. For their time and skill.

Then if you said, "Well, look, we're no different than a stethoscope." Then, CMS would be like, "Great. Don't bill us for the HCPCS code." You go, "Ding it. That backfired."

Joe: Why not? I wasn't handing out stethoscopes before, but I am now.

Nic: Then, CMS would say, "Well, then don't, if you're no better."

Joe: What do you mean "If you're no better?”

Nic: If it's no better, if it doesn't change anything, why hand out a stethoscope to everybody? If [client] approaches CMS and says –

Joe: "Look, I have a stethoscope I can hand out." They'll be like, "Why would you do that?" Unless I said, "Well, look at these three clinical studies that say that's why I do it." Then they'll be like, "Oh, snap, man. We have to get you some payment."

Nic: Bingo.

But if they show up and they say, "Look, we're no different than just a stethoscope." Then they go, "Then why are you billing us the extra 20 grand?" You go, "Oh crap. If we don't improve anything, why am I billing that?"

On the rush to market

To be completely honest, over these next six to eight to 12 months,

… as people start billing for this code, I think there's going be some broken hearts.

I'm scared digital health companies are going to poison their own well, and they're going to come to market and they're going to think they're unique and they don't have to do what UroLift did, and they don't have to do what Cologuard did...

Joe: By "did," you mean all those clinical studies?

Nic: Yes.

Joe: Okay.

Nic: Yes, I think every digital health company I've ever spoken with wakes up and goes, "You know what, we're computer programmers. Wouldn't it be cool if I could do this? And wouldn't it be amazing if doctors could do this and this and this?"

And also, we're making an assumption that doctors want to sit at a portal and read at-home EKG readings and things. That, I'm sure, the company's already figured out – whether doctors are willing to do that or not.

Now, if they are willing to do it, they'll get reimbursed the CPT code for it.

Joe: I think you subtly just said, "I'm going to give them the benefit of the doubt. But I really don't know if they had any data to support that there is a firm customer demand for this concept."

Nic: I would want to see, as a Payer, [client] in 50 patients versus not [client] in a commensurate group of 50 patients.

And I would want to see 12-18-month follow up. Whatever it is, the follow-up period determines what endpoint you're trying to capture - the increase in stroke, decrease in heart attack, ...I want to see this stuff.

On successful clinical trials

Joe: It seems so extraordinary to me.

You're talking 18, 24, 36 months. There are so many things we can't control against that – could I even feel good about saying, "This had a 5% fewer incidence of heart attacks. Therefore, we're a go."

I don't know this business. But if I were the decision maker, I'd be, like, "Okay, but I don't know that I could statistically claim this is repeatable."

Nic: Yes. I often say in these lectures,

Logic and intuition are poor substitute for science.

Logically, it makes a whole lot of sense to me that if I handed out 50 [client]'s to people, I'm going to get fewer heart attacks and strokes and so on.

Well, have you proven that?

Because what I may end up seeing is that I'm inducing supplier-induced demand, speaking like a health economist.

I'm going to get a whole bunch of people measuring their vitals all day long and giving themselves white coat hypertension, and then they end up having a spike in this and that, and now we got a whole bunch of people demanded more anti-hypertensives.

That's how a payer thinks.

On driving down hospital costs

Joe: One of the main concepts that drives the interest and pursuit of the market is the money-losing proposition that a typical office visit for a chronically ill person is.

Nic: Yes.

Joe: Every time the diabetic shows up, every time the COPD patient shows up and just needs a reading or a checkup, the hospital loses money. Won't hospitals flock to this concept so they can obviate that cost and save lots of money?

And logic dictates that if you check your pulse every day, that's better than checking it once a month. And we can catch things earlier and have them come in if there's an unusual reading.

Nic: Yes, but let me ask you a health economic question. If I have to use this in 10 patients and I have to pay 100, I have to pay $1,000 for all of them.

Let's say my number needed to treat is tenths, meaning I have to give this to 10 patients to help one patient avoid a thing that I wouldn't have avoided had I not used it.

Joe: Yes.

Nic: So, it only benefits one of those 10 patients.

Joe: Yes.

Nic: Well, what if the benefit is $900? Well, why would I pay 1000 to save 900?

Joe: Well, that's like an insurance policy. And aren't you, as a Payer, pretty much medical insurance? That's the game we play, isn't it?

Nic: Sure, but that's why they hire people like me – health economists – to sit down and do exactly this. And, just minutes ago, you said, "Nic, logically doesn't it make sense for those people that are measuring their stuff that they're getting better outcomes?" I go, "You bet. That's completely logical, but there's also finance and economics.

On the value of an ‘improved patient outcomes’

Joe: Well, this goes at the heart of the conversation I had you, Gunter, and Beth talk about two, three weeks ago. What is the value of an improved patient outcome?

So, let's use your example. There are 10, and one guy may have averted a heart attack. How much is that worth?

Can I justify paying $10 million to save that guy a heart attack or no? Of course not.

Can I say it's worth at least a nickel? Oh, yes, of course.

What's that number in between a nickel and 10 million? And who decides that?

Nic: I decide that.

Joe: And what did you use as a metric at Intermountain?

What is that number?

Nic: Well, my number is, I would go to [coworker], and I would say "Pull for me the last 300 heart attack patients we had." She goes, "All right. Give me a minute."

She types it all in. Look, I'm not kidding. literally pulls up 300 heart attacks. And, then, I say, "Okay. Give us the billed, allowed, and paid charges for each of those."

And she pulls all those up for me. And she says, "Well, billed was $49 million, allowed was 42 million, and paid was 41 million."

And I go, "Okay, divide that by 300." That tells me what, on average, billed, allowed, and paid charges are."

Joe: Right.

Nic: Now, I go back to [client]'s papers. And I say, "All right, [client], what was your number needed to treat to avoid a heart attack? Because these patients were so diligent and adherent to your protocol when they went home with your device."

Joe: Right. So, we saved one in ten.

Nic: And I had to pay $1,000 for all 10 even though it only benefited one of them. I wasted money in nine to benefit one.

Joe: Yes.

Nic: So, if that is less than the other number – whatever that billed, allowed, paid charge was that I got from Julene – then, I'll pay for it. I'll reimburse.

Joe: So, it's really divided by the paid number. The other two are nice to know, but what’s my out-of-pocket?

Nic: Yes.

Joe: So, let's say it was at least 30,000.

Nic: Sure.

Joe: So, if you saved one guy's heart attack, that's worth 30,000 to me. But you are saying, "I paid 1000 to save 30, then I'm in?"

Nic: Yes. Heck yes.