A coding expert responds to a physician's strong reservations about ICD-10, one by one.
This article originally appeared in JustCoding.
ICD-10 implementation is coming in less than 50 days. Despite the imminent transition, many physician practices lag behind in preparation, likely due in part to ICD-10 resistance from the American Medical Association (AMA) and other state medical associations over the years.
CMS and the AMA recently came to an agreement over ICD-10-CM audit and payment flexibility for physicians, but the agreement doesn't necessarily mean an end to physician resistance to the new code set.
Coders and coding managers need to understand physician objections to ICD-10 in order to help them overcome them in order to prepare for the transition. JustCoding recently spoke with W. Jeff Terry, MD, of Mobile, Alabama, about his views on ICD-10-CM implementation and how he thinks it will impact the industry.
Terry has practiced urology in since 1985 and served as president of several state medical associations. He has been a delegate to the AMA since 1995; has served on the AMA Council on Medical Service; and has been chairman of the Alabama delegation to the AMA since 2008.
Terry authored an AMA resolution to delay ICD-10 in November 2011, which led to postponing implementation until October 1, 2014. In February, Terry testified for the American Urological Association before the House Energy and Commerce Subcommittee on Health and was the only witness out of seven witnesses to testify against the implementation of ICD-10. His opinions do not necessarily represent those of the AMA or other physicians.
JustCoding asked Shannon E. McCall, RHIA, CCS, CCS-P, CPC, CPC-I, CEMC, CCDS, director of HIM and coding for HCPro, a division of BLR, in Danvers, Massachusetts, to respond to Terry's comments and provide ways that coders can ease physician concerns about the transitions.
Documentation challenges
Terry: We will not really know how much documentation will be needed until we implement ICD-10 and then have the insurance companies deny our care Physicians document appropriately in order to take care of the patient. It is an entirely different situation as far as how we must document in order to support an ICD-10-CM code and get paid.
If I document "urosepsis" in the hospital, every physician knows exactly what I am talking about. This is a patient who has a urinary tract infection so badly that it got into his or her bloodstream. Patients can die from this and it takes aggressive treatment to cure these patients.
The insurance companies and Medicare will not accept "urosepsis" in the new ICD-10-CM system. You must say in the medical record, "urinary tract infection with septicemia." It is a big game for these folks, as well as the coders. It is a game that allows them to take payments away from physicians for care given and it has nothing to do with the quality of the care. We have used the term "urosepsis" for more than 100 years and now it is not acceptable in ICD-10-CM.
Another example is a surgical operation I do called a TURP (transurethral resection of the prostate). We've also used this term for more than 100 years. It is found in medical textbooks and is a standard term for this surgical procedure.
Now in ICD-10-PCS, this will not be accepted. I must say TUEP or transurethral excision of the prostate. If I don't use this term, then they will not pay me or will take payment back on review. If I use TURP and they do pay me, then the Recovery Auditors can charge me with fraud for improper billing. It all hinges on the words I use and not the care I give.
These are just two examples that I know about. There are thousands of other examples like this in ICD-10 that I do not know about and the insurance companies will just be waiting on me to make a coding mistake.
McCall: I think Dr. Terry's viewpoint is one of the primary reasons why physicians are so adamantly against ICD-10 and unfortunately he includes some truths, but more importantly, many untruths and inaccuracies. Comments like these are the reason when coders say the words "ICD-10" to a physician, they roll their eyes.
Although insurance companies may make the coding/billing world an ongoing "game," as he calls it, it is unfair to throw in the coders as those who consider it a "game" too. Coders just have to learn all the rules of the "game" for the providers so they can avoid penalties that can have financial implications such as fraud/abuse.
Much like the goal of a provider to provide quality care to their patients, the coding staff's goal is for the provider to be accurately paid for their services by assigning codes correctly. Coders are on the same team as the provider, not in the opposite dugout.
I recently conducted a training session for a group of specialty physicians and the common consensus at the conclusion was ICD-10 is not as different, big, and bad as they had read about in information from medical associations. The super inflation of how many codes exist is all that they focus on, but most were surprised by how many options already exist in ICD-9-CM, with the minor addition in ICD-10-CM of certain factors such as laterality. The creators of ICD-10-CM did not come up with 50,000-plus new diagnoses.
Urosepsis is already a problematic diagnosis in ICD-9-CM. Based on his comments, as a provider he may not have experienced it as often because in ICD-9-CM there is a default code for the term "urosepsis," 599.0, urinary tract infection (UTI), site not specified.
Based on his stated clinical definition, this is not merely a UTI, but guess how many times it may have been coded that way based on documentation of "urosepsis"? Probably a lot. This is certainly not a good way to reflect the severity of the patient's condition.
In an inpatient setting (which it would likely be considering he said "they can die from this") some coders may just assign 599.0 since it is considered a complication/comorbid condition, which can still impact facility payment by possibly grouping to a higher-paying MS-DRG. If it does not impact the reimbursement, they may not find it vital to clarify further. But it is probably more likely that a query for clarification should be warranted.
From a clinical sense, if this is generally meant as a UTI with sepsis, another code would be assigned for sepsis (038.x or A41.9 in ICD-10-CM) and can impact the principal diagnosis, which can change the MS-DRG from a UTI (689-690) to sepsis (870-872), which would lead to higher reimbursement.
From a physician standpoint, though, I'd think it's fair to say that he is only reporting an evaluation and management (E/M) code for his management of this condition for his patients. Physician services are paid based on the Medicare Physician Fee Schedule and it is highly unlikely that payment would be denied due to the diagnosis used for the E/M code.
In fact, quite the opposite. For a provider it can help justify a higher level of E/M service if the condition was urosepsis versus a plain UTI due to the need for complex decision making (since the patient can die as he stated). Documenting the condition clearly as sepsis due to UTI versus urosepsis benefits everyone.
I also don't follow his concern regarding TURP, which is still used in the CPT® Manual. After ICD-10 implementation, physician coders will still use CPT to report physician services.
It's also not a problem for ICD-10-PCS. Per the ICD-10-PCS Official Guidelines for Coding and Reporting:
Many of the terms used to construct PCS codes are defined within the system. It is the coder's responsibility to determine what the documentation in the medical record equates to in the PCS definitions. The physician is not expected to use the terms used in PCS code descriptions, nor is the coder required to query the physician when the correlation between the documentation and the defined PCS terms is clear. (Emphasis added)
He can continue to document TURP and the coder will translate this information to ICD-10-PCS root operation Excision since this procedure typically removes part of the prostate tissue. The intention is not to make providers start to use illogical acronyms for procedures they've performed countless times. To be honest, I think coders would be more confused if he documented TUEP.
Coding difficulties
Terry: In our offices, it will slow us down tremendously trying to find the correct code. There are 250 codes for diabetes. Physicians are individuals who want to get it right and who want to do a good job. They may honestly choose two or three different codes for diabetes in the same patient on different encounters and then the insurance company will want to know why. The hospital coders will be calling the doctors to be sure the codes match up and this will take time.
The coding industry has done studies documenting that it will take their coders a lot more time to do the hospital coding and they have also documented that having several coders code the same record can result in different codes. If ICD-10-CM is so specific that the coding experts don't agree, then how in the world are the physicians going to get it right? If ICD-10-CM was just used for statistics and epidemiological data as it was designed and intended, it would not be a big deal if we were off slightly in our coding. The problem is that we have made the ICD system part of our billing system and therefore any small mistake is a big deal for physicians. It should not be this way.
McCall: It is true that ICD-10-CM includes more options for conditions like diabetes but it's really just selecting a type of diabetes (1, 2, due to drug, due to underlying condition, etc.) and then identifying any manifestations. This isn't new to ICD-10-CM, so was he not doing it in ICD-9-CM?
There may be more codes with the addition of specific codes for types of secondary diabetes but that's the only major change. Pull down lists from EMRs may necessitate a two-part questions like: diabetes, type, then identify manifestations, if any.
As for coding taking longer for hospital coders, ICD-10 will certainly affect productivity at first. But I do think as coders get more comfortable with the code set they will revert back to their current production. The variations in codes for an inpatient record when coded by different coders is solely regarding the subjectivity of certain codes. The variations really have nothing to do with the code set used, considering this is an issue in ICD-9-CM as well.
For example, one coder may code a history of skin cancer as being relevant to the admission whereas another may not. These codes most times do not even impact payment. This variation is not considered a "mistake," it is considered subjective and anyone who reviews codes (payer, auditor) expects this will occur. It's also not likely to impact physician coding nearly as much, considering in a physician encounter there are limited diagnosis that would be considered relevant to the encounter.
Moving straight to ICD-11
Terry: My opinion of ICD-11 is also the opinion of the Association of Medical Directors of Information Systems (AMDIS) – an association of Chief Medical Information Officers (CMIOs) and other healthcare IT leaders throughout America.
ICD-11 is probably more complicated in some aspects, but also more clinically oriented in other aspects. It uses a coding system called SNOMED CT®, which I don't understand. It is designed to be used in the EMR era, while ICD-10 was developed in the pre-electronic era. Putting politics aside, AMDIS feels that it would be best to skip ICD-10 because ICD-11 will ultimately be better and the rest of the world will be moving to this in 2017.
McCall: Waiting for ICD-11 would be just another procrastination tactic, because although it may be "better," it is still based on ICD-10's fundamental format. We'd be in the same boat, just in the distant future, so what would we do until then?
Saying the rest of the world is moving to ICD-11 in 2017 may end up being the truth per the World Health Organization, but as we've seen from the past, other countries adopted ICD-10 15-20 years before the United States. If we follow this timeline for ICD-11, that would mean we might be waiting until 2037.
Email your questions to editor Steven Andrews at sandrews@hcpro.com.
This article originally appeared in JustCoding.