Month: June 2020

This is the intro to a series. See Part 1 here, the context to Part 2 here, Part 2 here, and Part 3 here.

Every medical specialty has its own organization (or sometimes a few competing organizations), and the main organization for internal medicine physicians is called the American College of Physicians (ACP). I think it has such a general name because it wants to be representative of internists AND also all the people who did internal medicine on their way to sub-specializing (e.g., cardiology, infectious disease, rheumatology, gastroenterology, pulmonology, . . .). Yes I know most of them consider themselves internists as well, as they should!

After the AMA (which seeks to include all physicians of every specialty), the ACP is the second-largest physician group in the U.S., which makes sense because internal medicine is the most common residency program for medical school graduates to do. I just checked the data on residency positions, and there were 8,697 internal medicine positions for 2020. The second most common one was family medicine, with 4,662 positions.

Random fact: What’s the smallest specialty in terms of residency spots? Apparently it’s preventive medicine. It had one spot available in 2020. I didn’t even know you could do a preventive medicine residency.

As a large medical specialty organization, the ACP has some significant power to wield, and that’s important because one of its main purposes for existing is to promote the interests of internal medicine physicians in the political arena. I’ve seen this in action twice in D.C., when I joined ACP members from all 50 states to meet with quite a few members of congress to advocate for specific issues.

And now the ACP has entered the realm of my primary interest by releasing a new set of policy proposals specifically talking about how to fix our healthcare system.

I haven’t read all their proposals yet–I’ll be doing that over the coming weeks–but I can guess the general approach they will take because I’ve met with enough physicians who are active in the ACP to know the general type of person who gets involved in this flavor of organized medicine and would be the ones influencing these proposals. At the risk of being accused of overgeneralizing, I will describe them as passionate and caring individuals whose morals drive them to push for social justice and to advocate for system changes especially to improve care for the underserved. Of course there are many involved in the ACP who have totally different priorities, but I would venture to say that that description applies to the most prevalent motivation to be involved.

Another reason I can guess how these proposals turned out is because I’ve heard ACP’s Senior VP over Government Affairs and Public Policy, Bob Doherty, speak a number of times, so I have a sense for what he is all about. Bob is not a physician–he’s a policy guy–and he’s articulate and impressively comprehensible when speaking on health policy issues, which tells me he knows his stuff super thoroughly. He’s been with the ACP for a long time, and I am willing to bet that, intentional or not, he had a significant influence on the shape these proposals took, if even only indirectly by being the primary person teaching ACP leadership about the issues.

My point in talking about Bob Doherty and about what kind of person gets involved in ACP leadership is to say that these proposals are going to reflect the minds of those individuals, and that they would probably have ended up looking very different if I and a randomly sampled group of regular ACP members came up with them.

It’s just interesting that now we have the second-largest physician organization essentially saying, “This is what physicians believe about how to fix healthcare.” And the rest of the country will not have any reason to question that.

So, I guess it’s time to learn what I, an internal medicine physician, believe about how to fix healthcare! There are three main papers that lay out the ACP’s proposals, so over the next few weeks I’ll take each one in turn and see what they have to offer.

I have mild persistent asthma, which means I take a daily suppressive inhaler and occasionally also need to use a rescue inhaler. I moved a year ago to a new job, but I had enough medication stockpiled that only a few months ago did I finally need to get more inhalers. You would think that, as an internal medicine physician and also a health policy researcher, I have the rare perfect skill set that would allow me to get the best deal on these medications. But you would be wrong.

I have a high-deductible health plan and haven’t made any claims this year, so I was confident that I would be paying the full price for my medicine since that’s what it says in my benefits description. And since inhalers can be expensive, I decided I should start by logging into my health insurance profile to see if they have a formulary that would list which inhalers are the cheapest.

I didn’t find a formulary. But I did find a very interesting tool that I guess is supposed to be better. It’s called “Rx Savings Solutions.” It lets me look up a medication and then shows me prices. Simple, right?

So I looked up fluticasone proprionate–not to be confused with the nasal version fluticasone proprionate, or the other fluticasone proprionates that are combined with other medicines, and good thing I know that HFA means it’s an inhaler. It said the estimated cost was $192 for a 1-month supply, but that was for a retail pharmacy. I could instead use a mail pharmacy and pay $562 for a 3-month supply (I initially didn’t notice that price was for 3 months instead of 1).

But it went further than that. It offered alternative formulations of the same medication (inhaler vs diskus, although the diskus was listed as the same price as the inhaler), and it even offered alternative medications (fluticasone furoate, beclomethasone, budesonide, . . .). I guess if you are using that tool real-time while at the doctor, that could be helpful in steering them to an equivalent but cheaper medication. But it would probably be a frustrating time-waster for the doctor!

Luckily, I was writing the prescription myself. And since I knew that any of those other medications would work fine, I felt confident. I was even savvy enough to look at the recommended dosing frequency and the number of doses per inhaler to make sure they would each last the same number of days. Yes, I was confident at this point.

Beclomethasone and mometasone were both within $1 per month of the price of fluticasone, so I just decided to go with fluticasone because that’s what I’ve been using. I selected it, which then took me to a page listing the actual prices at a variety of pharmacies. In this case, every pharmacy listed had the same price, still that $192 for one inhaler. Other medications I have searched have showed different prices depending on the pharmacy. I’ve heard of various pharmacy-specific discounts, so I wasn’t sure if I should trust the prices listed, but I wasn’t willing to call each pharmacy to ask.

But $192 seemed like a lot of money per month. Remember, I’m a young doctor paying off large debts, and also remember that I grew up in Canada where medications are cheap. So I decided to broaden my search.

I checked GoodRx. They claim they have prices that will beat my insurance plan’s prices! I’m sure they do, but in this case, the price for one fluticasone inhaler was $256 at all the pharmacies listed, so no luck there. I didn’t go out of my way to check the GoodRx prices for beclomethasone and the others, although I should have if I wanted to truly be confident I was getting the best price.

I’ve heard of Canada mail pharmacies, so then I looked into those. There are a few options, so I started checking prices on all of them for a few of my medication options and their various formulations. This took some time, and eventually I was able to find one that offered the exact same fluticasone inhaler for $35. This was a killer deal! Except there are uncertainties about if ordering meds from my motherland will get me in trouble.

Ultimately, I had run out of fluticasone and my asthma was starting to get bad, so I gave up my search for the best price and just went with what I knew I could get the next day. I called in a prescription for a fluticasone inhaler to our local Walmart. When my wife picked it up for me, she said they only charged her a $36 copay! But they had to switch it to the diskus version.

I don’t know why they didn’t charge us the full price. But I think I know why the switched it to the diskus version. Looking again at the Rx Savings Solutions website, it turns out the price listed there is $182 for the diskus (as opposed to $192 for the inhaler), but I only see that price when I search the diskus directly. If I search for the inhaler and then look at the price of the diskus in the alternative formulations section, it lists the price of the diskus as $192. Go figure.

In a way, it’s good that the cheapest price is the one any normal consumer of healthcare would have gotten. Unfortunately, that’s probably not normally the case. But any other person on my same insurance plan who needed a refill for fluticasone would have gone to their doctor and asked for a refill, then the doctor would have electronically sent a new fluticasone prescription to their usual pharmacy (having no idea whether it is the cheapest option for that patient’s insurance), and then the patient would have picked it up and paid whatever price the pharmacy requested.

Do you want to know how doctors figure out how much a medication will cost a patient? Well, the real answer is that we don’t figure it out. But for those times when a patient is requesting a price due to severe financial constraints, we call the patient’s pharmacy and have them run a test prescription. The computer then spits out how much the patient’s out-of-pocket cost will be. That’s the only way to know the actual out-of-pocket cost. Not even a pharmacist can tell you that information without running a test prescription. Can you imagine a physician really trying hard to find the best deal for a patient? They would have to call multiple pharmacies and have them run a variety of test prescriptions. That is completely impractical.

This is yet another manifestation of the despicable complexity in our healthcare system, and it wastes so much time and money of patients, doctors, and pharmacists.

Did you notice the impossible number of variables someone has to sift through to find the cheapest price for a single treatment? The insurance plan, the number of months’ worth of medication you’re buying, the outside rebate programs, the mail options (U.S. and international), the alternative equivalent treatment options, the alternative formulations for the same meds, the frequency of dosing, the search term used to find the price of a medication, and the pharmacy-specific price impacts.

There are clearly other variables I still don’t know about, which is why, even after all of that, I still didn’t know what my actual out-of-pocket cost would be.

The last two weeks, I have written about the money-and-time-wasting complexities of hospital billing and inpatient provider billing. Are there solutions?

Remember, first, that to identify a solution to a problem, you have to accurately diagnose the cause of the problem. So we should back up and talk about that first.

I should give a disclaimer and say that, while I’ve read a variety of book chapters and articles talking about how our billing system evolved, I am no expert on this subject. Nor do I want to try to go into the deep billing history here. But maybe a big-picture summary will suffice.

Why would people design such a complicated system? On the provider side, the two reasons that stick out are as follows:

1. To make sure providers are actually delivering enough care to warrant getting paid their fee. It’s trying to prevent that small minority of “innovative” providers who would cut corners in a way that isn’t necessary illegal but that would enable them to generate way more RVUs per hour than good, thoughtful, conscientious care would allow.
2. To try to quantify differences in complexity, especially so that providers are rewarded for doing more difficult things.

On the hospital side, those same two reasons also apply, although the second one is much more relevant. There are, though, issues with the first reason as well, which is why 30-day readmission rules have cropped up.

So there you have at least a start to an understanding of why these complicated systems have evolved to what they are today. And your mind is probably brimming with solutions already based on that short explanation. For example, if most providers end up billing about the same average complexity, why not just simplify things by only having a single complexity level and trust that things will average out? Some providers will take a pay cut and some will get a raise, but most will still make the same amount of money and will save a lot of time and expense (less overhead dedicated to paying for coders to review charts, etc.). Or why not have fewer modifiers that change what DRG a patient gets assigned to? Or, heck, just fewer DRGs in general? Better yet, just have a closed integrated provider-insurer organization that can make whatever billing rules make the most sense to everyone in that organization?

We could easily come up with lots of other solutions. But remember I’m interested in fixing the whole healthcare system? Would any of these help us make progress toward actual and significant big-picture value improvements in our healthcare system? Or might they risk creating more barriers to getting to the real solutions?

Remember that my explanation of how to fix healthcare (as detailed in my Healthcare Incentives Framework) is to enable more people to identify and then choose higher-value providers and insurers. This means people need to know price and quality beforehand. (This isn’t always possible, but it is possible in more circumstances than people realize.) And when people start to know provider quality beforehand and then choose them based on that information, the issue of corner cutting by providers is greatly reduced. This would obviate the need for complicated billing requirements that are designed to make sure providers are really doing what they are supposed to do to provide good care.

But how would all those complicated documentation rules actually go away? Well, if providers start to shift to making their own prices (and insurers allow it because patients are paying most or all of the differential between provider prices), this would mean that those rules no longer apply because the providers are the ones in charge of determining how much they get paid rather than having to adhere to arcane insurer rules that dictate their fee for them.

This is probably a topic that requires further exploration another day. But my point is that all aspects of the healthcare system would change if we get more patients to make value-sensitive decisions when choosing providers and insurers.

(As a sidenote, I am sometimes asked by coworkers, “Would your solution to healthcare fix [insert specific issue here]?” And often my answer is, “Not directly.” And then I usually leave it at that because I know they don’t want 15 minutes of explanation about financial incentives and value-sensitive decisions and different types of pricing arrangements. Billing complexity is one of those issues that wouldn’t directly get “fixed,” but it would sure get better!)

In summary, the targeted easy-to-think-of solutions to all these expense-inducing coding complexities we’ve been talking about would make a difference in the problem they’re designed to solve, but remember to keep the big-picture solutions in mind to make sure any new policy doesn’t create additional barriers or detours to getting to that final destination.

Last Tuesday, I wrote about the complexities in hospital billing. This week, I’ll look at provider billing.

The same preface is required, so let me copy that from last week:

Any provider’s goal with billing is to bill honestly to the highest amount they can for the necessary work they’ve done. Not billing for something that we are legally allowed to bill for is leaving money on the table and unnecessarily lowers our income. So keep in mind that, when I talk about billing the maximum possible, this is an honest and appropriate goal that is very different than taking advantage of the system or committing some kind of fraud.

When I was in residency and would work in our outpatient clinic, we had coders reviewing every single note we wrote, checking to make sure our documentation supported the bill we’d submitted for each patient. Us residents would get messages from them fairly frequently (some of us multiple messages per week), asking to go back and clarify things like the name of a problem or to change the billing code we’d used. The worst was after doing a half-day of dermatology clinic, where there were also tons of little procedures that required procedure codes and different specific details in our notes to justify. One time, I remember taking more time to fix all the notes to the coder’s specifications than it originally took me to write all of those notes.

We hated these messages from coders because it was just one more administrative annoyance required of us. (The coders themselves were lovely and patient with us, but that can only make it so much better.) We’d have fairly frequent coding training sessions, which helped us do the right thing the first time a little more often, but short of dedicating a full week or more to coding training, there was no way to teach us all the things we needed to know to do everything right the first time. And every minute spent training us on billing meant one less minute training us how to be good doctors.

When I started working as a hospitalist, I had a coding training session as well. And then I, like every new hire, was on a probation period, during which time a coder reviewed every single patient note I wrote to ensure I was doing everything properly.

My first day off after starting my new job, feeling overwhelmed at all the medicine I was supposed to know, do you know what I did? I went to a public library and spent a whole day studying the pages and pages of coding requirements the coder had given me. Over the next week or so, I took copious notes, and eventually I was able to clarify things in my mind enough to feel confident. This required multiple email and phone discussions with another very patient coder. Sometimes communication was a challenge because we had to try to translate her coder’s brain and terminology to a lowly physician’s clinical brain and terminology.

I wanted to be studying medicine, but without getting the bills right, the hospital doesn’t get paid!

Do you want to hear a little of what I learned? Let me tell you. I’ll insert the five-digit billing codes next to each one, just for fun.

When a patient is seen the first time by a member of my hospitalist group (not to be confused by the first day a patient is in the hospital), we need to code for an initial encounter. There are only three levels of complexity we can use, so let me lay out the requirements to bill at the highest level (99221).

First, their level of “risk” has to be high. There’s a long list of things that qualify, such as a change in neuro status, giving them a drug that requires intensive monitoring, or a “severe” exacerbation or progression of a medical problem.

Second, they also need to have at least four problems we are treating (a max of two of them can be self-limited), or at least one new problem that requires a workup, or a new problem without workup plus at least one other problems, or just a worsening established problem plus at least two other problems.

If they don’t meet the high-complexity level based on their problems, you can instead use a points system to meet the second criterion. If you get at least four points, it counts. There is, again, a nice list of things that get you points, such as ordering or reviewing labs, ordering or reviewing imaging, ordering or reviewing an EKG, personally viewing an image/EKG, discussing results with a provider, deciding to obtain collateral information about the patient from another person, deciding to get a consultation from another provider, deciding to request records from another location, and actually getting collateral information/consultation/records. Each one counts as a point. But if you order two imaging studies or EKGs or labs, you can’t double-count. This points system is part of what explains why there are so many seemingly unusual comments in a patient’s chart, such as, “I personally reviewed the EKG.”

So, let’s say the patient qualifies for high complexity based on those first two criteria, now I need to make sure I put all the other information in the patient’s note that will allow me to bill to that level. This includes at least four details of history from the patient (there is a specific list of what qualifies, but it includes the location of a symptom, its quality, severity, duration, timing, context, . . .), at least one symptom question about at least 10 different organ systems, and at least one thing about the patient’s medical history, family history, and social history.

Then, I need to document an exam with at least one thing in eight different systems. Or, if I’m a specialist focusing on one specific organ, I need to document at least eight things about that one organ. And some things don’t count. I thought I could use “Head/eyes/ears/nose/throat: atraumatic” as one of my systems, but it turns out I can’t. But I can instead say “Head/eyes/ears/nose/throat: Vision grossly intact.” Physicians, even internists like me, are rarely concerned about eight different body systems at the same time, so we have learned to say things to check off the eight systems without having to do too much extra work. For example, my standard physical exam includes the very useful “Dermatologic: No rash noted” because, when I’m looking at the patient’s face and belly and legs, I didn’t note a rash. My standard exam also includes the very useful “Psychiatric: Normal affect” and “Genitourinary: No Foley.”

Now you know the requirements to bill for a high level of complexity for an initial care episode. If I don’t meet every single one of those requirements, I can’t bill at that level, and I need to see what lower level my documentation would qualify for.

There’s another set of requirements for moderate level (99222) and low level of complexity (99221), each with their own list of what constitutes the right level of risk, the right number of problems or points, and the right number of history and exam details. And there’s yet another set of requirements for all three levels for subsequent encounters (99231, 99232, 99233). And another set of codes if the patient doesn’t qualify as sick enough to be considered an “inpatient” but is only here “under observation” (99218, 99219, 99220).

But you could bill for time instead of complexity! In this case, the documentation requirement is much simpler. You just have to have a statement in your note that tells how much time you spent on the patient (but time spent doing procedures counts separately), and at least 50% of it has to be related to direct patient care and also coordination of care. You also have to give some details about the content of what took longer than would have been otherwise justified based on the patient’s level of complexity. And you need to remember that, to bill for a high level of complexity, you need to have spent at least 70 minutes on an initial encounter or 35 minutes on a subsequent encounter. And for a moderate level of complexity, you need to have spent at least 50 minutes on an initial encounter or 25 minutes on a subsequent encounter. And for a low level of complexity, you need to have spent at least 30 minutes on an initial encounter or 15 minutes on a subsequent encounter.

But what if the patient was admitted after midnight? That means they have already been billed for that calendar day. But if you spend at least 30 minutes on the patient (or, if you are doing an initial admit and spend the expected amount of time plus at least 30 minutes more than would be expected), then you can bill for a “prolonged service” (99356). But you have to make sure that the total time the initial provider took plus your time is at least equal to 30 minutes plus the expected amount of time the initial provider should have taken based on the level of complexity and encounter type they billed for. This is why we have to put our time in every note, even if we aren’t going to bill for time.

And if the total extra time on the patient was at least 60 minutes, then you can add two prolonged service codes (99356, 99357), one for each 30-minute chunk. But make sure you specify how much actual face-to-face time if the patient is on Medicare, because that’s the only kind of time that counts for prolonged service codes for Medicare.

Maybe I should stop. I know everyone also wants to know the details of how to bill and properly document if a patient is admitted and discharged the same day (99234, 99235, 99236), which code to use on discharge day (99238, 99239, or 99217), when you can bill for critical care time (and what you have to document) (99291, 99292), but I’ll stop.

When I think about the cost to code this way, it’s significant. There are costs for the armies of coders reviewing providers’ notes and sending messages, costs in doctors’ time (learning this stuff, responding to coding queries, getting refresher trainings, documenting so many things that are irrelevant to medical communication), costs in software to make sure we are all billing to the maximum allowable, costs in consultants and administrative personnel analyzing our coding data and figuring out how to optimize it, and (most important of all) costs in quality of care because doctors have to spend so much time on this rather than learning more medicine.

I talk sometimes about despising the complexity of our healthcare system and that it is a major component of its crazy high cost, and billing complexity is one of the things I am referring to. I hope now you can see why.

And since I like thinking and writing about solutions, so I’ll give some thoughts on that next week.