ICD-10: Get Ready for Another Efficiency Scam

Seth Bilazarian, MD


September 24, 2015

This feature requires the newest version of Flash. You can download it here.

ICD-10 Implementation: Ready or Not

Seth Bilazarian, MD: Hi. Seth Bilazarian on on Medscape, for Practitioner's Corner. This month, I wanted to prepare for International Classification of Diseases, 10th edition (ICD-10) and comment on the process. I'm sure practitioners around the country are dealing with this. Everyone knows by now that we're not far away. October 1 is the deadline for implementation—one of those funny governmental things that says the change is required for everyone covered by HIPAA, which is equivalent to saying that it pertains to humans.

The Centers for Medicare & Medicaid Services (CMS) had made a concession that during the transition fromICD-9 to ICD-10, they will continue to reimburse for wrongly coded claims as long as that erroneous code is in the same broad family as the correct code. There is some parsing of words there, but it seems like some leeway will be granted as we make this transition. Unfortunately, this transition is coming at a difficult time for many practices, including ours, during which we will be simultaneously implementing ICD-10 and self-auditing for meaningful use. It will be a significant challenge.

I wanted to call this blog "ICD-10 Scam." I'm not using that as a pejorative term; I want to be very clear about the word "scam." A scam is a confidence game or some other fraudulent scheme in which the scammer promotes or alleges a benefit that is patently false. The benefit accrues to the scammer, but the person who is scammed is someone who enrolls and derives really no benefit. That's why I think the ICD-10 is a scam, and I'll go into that here in a little bit greater detail.

I'm not entirely cynical. If anything, I'm generally optimistic and trusting. However, I feel that I previously missed scams that were foisted on us, and of course one of them is the electronic medical record (EMR) scam. This is now decades old. We were promised a really valuable tool, but, of course, EMRs continue to be hard to use; they degrade the patient experience, they're not reliable, they cost too much, and the templates don't match our workflow for patient care. There's still no interoperability, no customizability, and it's a woeful intrusion on the physician/patient interaction. That's where we are with the EMR. Yes, there are some good things about it, but really that was a scam foisted on us as something that was going to be better, safer, and easier to use.

Of course, another big scam that we benefited from was those physician leaders who taught us that the maintenance of certification was completely a scam. In this case, I've gone through all of the required education for ICD-10, I've attended classes, I've done webinars. I've estimated that my training thus far for both group-wide as well as department-specific ICD-10 is about 10 hours. I've been told that ICD-10 is going to enhance our ability to measure quality, safety, and efficacy of healthcare. That's probably the one thing that's true about this, but the rest of the promises are not.

Increased Efficiency: Not Likely

The claim that there'll be a reduction of the need for extra documentation to describe a patient's condition to satisfy reimbursement —not likely. It will streamline patient system design—not likely. It'll be more efficient and less cumbersome for healthcare delivery systems—absolutely not. The promise of major benefits that relate to the EMR and collection and storage of data—no. Prevention of fraud and abuse? I don't really understand how a more complex coding system can help with that. There may, of course, be some improved tracking of public health; I certainly think that that's a possibility.

I was instructed that payers cannot currently pay claims fairly because the healthcare industry cannot accurately measure quality of care under ICD-9, and I think this is the scammer's approach to foisting this on us. Two things I heard time and again in my teaching from consultants that our practice brought in and others is that this will save time and money and improve quality. I have several comments on that.

One problem is that there is specificity. I was repeatedly taught that I need to convert ICD-9 to the highest level of specificity. That's a quote: "Convert to highest level of specificity." I understand how specificity might be valuable from a public health standpoint but not from the standpoint of clinicians and patients. Clinicians are already burdened by digital input requirements with the EMR and meaningful use. The best people in healthcare expected to code these diagnoses are not physicians, and I think that this is going to further jeopardize our patient interactions.

I'm sure that many people have seen that there are all kinds of specific issues about coding. A lot of them are very funny; for example, distinguishing between a macaw bite vs a parrot bite. This will, of course, produce databases that are more easily interrogated for research and quality outcome measures, which I think is a positive thing for medical care. But this specificity comes without accuracy. The consultants I was able to meet with weren't able to provide any useful insights in this regard. For instance, there's an ICD-10 code, "coronary disease due to lipid plaque." I asked our consultants what this meant; they had no insights. I asked my physician colleagues, and I got four different answers from the five cardiologists in my group.

One physician said that coronary disease due to lipid plaque means insignificant or intermediate lesions. Another physician trained in CT angiography said that it's a stenosis identified on CT angiography with a lipid plaque. Another physician said that it's coronary heart disease in patients with severe hyperlipidemia. As an interventional cardiologist, I thought that it was plaque identified with intraluminalimaging on optical coherence tomography or intravascular ultrasound. So, which one of these is it? How will this be used in a specific manner without a definition?

Another area that doesn't seem like it will relate directly to cardiologists or even internists—but I think it will—is fractures. We can no longer code for an ankle fracture or a wrist fracture. Picture this: A patient comes to my office who needs a preoperative assessment for open reduction and internal fixation of a joint. The patient is in a cast, they have a broken ankle, and they may be seeing me for syncope or preop evaluation. How do I code for that? I can't code ankle or wrist if they're in a cast and I don't know specifically what bone it is. Do I guess that it's a talus or a distal radial or distal ulnar fracture? If I guess wrong, is that fraud?

Complex Coding

These are the complexities of the specificity that will be especially difficult when this may not be core to our specific specialty. Some of it is extremely specific. There's a code, "atherosclerosis of a bypass graft of a coronary artery of a transplanted heart without angina." How would you get to that without knowing that that exists? If you're a transplant cardiologist, then probably you'd be familiar with this code. But for the rest of us—I might code as atherosclerosis of a bypass graft or atherosclerosis of a transplant heart, but to be able to find the code for all of those, I think, will be very difficult because there are very poor search functions.

There is now an ICD-10 code for takotsubo cardiomyopathy (I51.81) but not for Lyme carditis or other kinds of cardiomyopathies. In our EMR (by GE Centricity), they've assigned the same code for rheumatic and rheumatoid myocarditis. I think that's an error, but there's no code for rheumatoid myocarditis, and they've assigned it to the rheumatic myocarditis code. There's no code for amyloid heart disease or amyloid cardiomyopathy, so our EMR vendor has assigned a nonspecific cardiomyopathy (I43) for amyloid heart disease. If you search for amyloid heart disease in our EMR, you will think that you've found the appropriate code, but it's really a nonspecific code.

These are really the problems that lead me to think that the nonspecific nature and the nondefinitions are really an issue. As for the issue of specificity, while the accuracy has many other hazards, one of the issues that also comes up is that ICD-10 makes it clear that when we treat a patient with an acute myocardial infarction (MI), we should treat the MI as a 30-day procedure. So, when the patient comes to see us at 1 or 2 weeks follow-up, patients have a ST-segment elevation MI, and they're compliant with their dual antiplatelet therapy, that's still within the acute MI presentation. Beyond that, that's not the case. We don't have that kind of insight about acute congestive heart failure patients. If a patient has acute decompensated heart failure and sees me at 1 week follow-up, are they still part of that acute decompensated heart failure, or is this something different? We have no insights, and our consultant couldn't provide us any information. Some other consultants may be confident that they have an idea about it, but they're just making it up as well because there's no real guidance from CMS in this regard.

Smoking is ill-defined. Some coding consultants have recommended that practices design their own internal recommendations for how to use the smoking guidelines to define a former smoker. Do you continue to categorize them as a former smoker if it's 1 week, 1 month, or multiple years? My intuition as a clinician would be that they're a former smoker if they ever smoked, but there is different guidance being given by different groups.

Bizarre Codes

Everyone I'm sure has heard about the CMS funny codes. They can just Google ICD-10 funny or outrageous or hilarious or bizarre or hysterical. Many of these codes deal with trauma. These include struck by a turtle, bitten by a pig, struck by a duck, pecked by a chicken, and knitting or crocheting accidents; they all have trauma codes that we're encouraged to use. Spacecraft collision has a code, hit or struck by a falling object from a canoe, a swimming pool in a prison, burned due to water skis on fire. There are locations for trauma including art gallery, an opera house, or even hurt in the library. I have spent a lot of time in libraries, and I've never been hurt in a library, but we do have an ICD-10 code (Y92.241) for being hurt in a library.

There are a variety of other things that I think that we're going to run afoul of. For HPI (history of present illness), if I see a patient who is having a stressful issue with a family member or perhaps an in-law, there is actually a code for that (problems in relationship with in-laws). The way our practice works, I'm sure our coders would categorize me as deficient in not coding Z63.1, problems with relationships with in-laws, if I had noted that in my history and physical. There are other codes that I'm not sure how to use. How do you use "bizarre personal appearance" (R46.1)?

Most physicians are dependent on an enabling machine and device (ie, a cell phone or a smart phone), and there a code for that. If we put in things about a patient in our prose notes and then don't code for it, theoretically we are at risk for inappropriate or unspecific coding.

No Going Back Now?

What are we going to do about this? This is federally mandated. We essentially live in a single-payer system. I know that there are multiple payers, but pretty much everyone follows CMS guidelines, and we have to use these codes for all the payers, so we essentially have a single-payer situation. It's reasonable to tell patients that we're struggling with the implementation of a really rigorous bureaucratic system that will affect our interactions as patients and physicians. I already do this with discharge of patients. I tell them that the reason it takes so long to discharge you is federally mandated rules about how long it takes to discharge patients.

I don't think we can avoid this. I'm not suggesting any sort of civil disobedience, but the issue of better diagnosis coding leading to better compensation is what's driving this for most practices. It would be really good if we had reasonable search functions. None of the EMRs that I know of have good search functions like the search functions we're used to on the Internet. Misspellings and synonyms are not categorized as part of the search, so when you search something, you have to constantly guess about whether your guess is correct. If you search for apical ballooning, you have to remember it's takotsubo.

Definitions that are agreed on and disease entities that are frequently updated would certainly improve the consistency and the accuracy. If this whole process had been more transparent from the federal government, I would have welcomed that rather than foisting on us this alleged scam.

What could have been? If there had been transparency with the public and the physician community, that would have been much better. Having tools adequate to the task such as a really good search function, really good definitions, and a full cardiovascular data set (eg, including resistant hypertension and amyloid) would be valuable.

Reimbursing physicians for the added time and effort that this will require would also be reasonable. There should have been a better certification process for coding experts, but in fairness to them, they can only use what CMS has offered in the ICD-10 dataset. ICD-10 does not benefit our patients or our medical care of them. The idea that this is going to be safer because there's laterality is really specious. Is it best for patients to have physicians do this? That's a rhetorical question. I would say no.

ICD-10 might benefit our national clinical research enterprise and public health research, but not for decades. Understanding whether a patient had a macaw or a parrot bite might be valuable from a public health standpoint over years, but it's not going to help our patients in the short term. It certainly would be better to have heart failure with reduced ejection fraction (HFrEF) and heart failure with preserved ejection fraction (HFpEF) separated for coding so we can follow patients' data better. I think that would be very useful. And having granular data about an inferior STEMI vs general MI would clearly be more valuable for public health and research reasons. ICD-10 probably will have a negative impact on patient care in the short term because of the time and financial burden for providers as practices and healthcare organizations try to implement it.

Our response again, I would say, is that we have to capitulate and comply fully. We have to accept this as another federal mandate, which undermines physician autonomy and the patient experience. We have to share that with our patients. We have to be cautious about the impact this distraction has on patient care. Physicians have a finite capacity for multitasking. Coding these intricacies I think is going to jeopardize the care of patients; the diagnosis, the prevention, and treatment. We ought to be wary because our physician societies have really failed us; they stood by and allowed this to be released without things like these good search functions and appropriate compensation.

Good luck after October 1. We're in this together. We're being put upon again, but I hope our patients will be understanding, and I hope there aren't too many errors that result from the added burden in taking our eye off the ball of our mission to care for patients.

Until next time, I'm Seth Bilazarian. Thanks.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: