The ICD-10 emperor has no clothes

The howling about the delay of ICD-10 was loud and fierce. It seems the quality of health care in the United States depends on our ability to use 68,000 diagnosis codes. The rest of the world has switched to ICD-10, and we alone insist on using an outdated coding system. Here’s a secret. The World Health Organization’s version of ICD-10 has about 16,000 codes, equivalent to ICD-9-CM. The rest of the world is not using ICD-10-Clinical Modification set, which has 68,000 codes. Only we, in the US, are considering that. The Canadian version of ICD-10 has about 16,000 codes, but the physicians do not use those codes for billing and reimbursement. They use a more limited code set of about 600 three-digit codes. Let me repeat this: The WHO version of ICD-10 that the rest of the world uses about 16,000 codes. Our version, developed jointly by the CDC and the American Hospital Association has 68,000 codes.

ICD-10-CM is going is going to add significant cost and complexity to physician practices without any benefit to the patient or physician. Perhaps facilities or payers need this level of detail, but we on the medical practice side do not. Selecting an ICD-10 code in an electronic health record will add 1-3 minutes to each patient encounter. Is that a reasonable use of physician time? If your mother or child is in the exam room, wouldn’t you prefer that the doctor spent that time with your family member, or you?

For years, I listened to the experts say that we needed greater granularity and detail in our diagnosis coding. The transition from ICD-9-CM to ICD-10-CM increases the number of diagnosis codes from about 14,000 to 68,000. That is significant additional granularity. But, when I studied ICD-10-CM in order to teach it to physicians and coders, I realized the ICD-10 emperor has no clothes.

My objections to ICD-10-CM are that it includes needless specificity, absurd adherence to taxonomy, unnecessary detail about injuries and insufficient additional information about chronic illnesses to justify its use.

In ICD-9, there are about a dozen codes for acute or chronic conjunctivitis. There are about 50 in ICD-10. Here are four: unspecified chronic conjunctivitis, right eye, unspecified chronic conjunctivitis, left eye, unspecified chronic conjunctivitis bilateral eyes, unspecified chronic conjunctivitis, unspecified eye.

Coding for gout explodes from about a dozen codes in ICD-9 to over 150 codes for gout in ICD-10-CM, differentiating chronic gout, lead induced chronic gout, drug induced chronic gout, chronic gout due to renal failure, other secondary chronic gout, acute idiopathic gout, acute lead induced gout, drug induced chronic gout, chronic gout due to renal impairment, and other secondary chronic gout. Each is reported by joint and acute gout by with or without tophus. Idiopathic gout right knee, idiopathic gout left knee, idiopathic gout unspecified knee. Or, chronic gout, unspecified. You don’t believe me do you? Get out your ICD-10-CM book and compare the codes using the search function on the WHO website for ICD-10.

Of the 68,000 codes over half are for injuries and accidents. Is it a laceration with or without a foreign body, which side, initial or subsequent encounter? Is the fracture at the upper or lower end of the ulna? What type of fracture is it? The mainstream media focused on the external cause codes “struck by a parrot.” But, these external cause codes are the least of our worries as we attempt to use ICD-10-CM in medical practices. And, I assure you, “struck by a parrot” is not in the WHO ICD-10 code set.

Now I’m sure that in response to this post you will show me the chronic care codes that have increased specificity and provide additional information for physicians health systems and payers. I know they exist. Great, let’s use a version that includes those codes without all of the other detail.

If Stark was the full employment act for lawyers, and HIPAA was the full employment act for consultants, then ICD-10 is the full employment act for coders. Much of the outrage against ICD-10 came from my fellow coders and consultants. I am sure that some of them have a deep-seated belief that ICD-10 is better. And, coding is the job coders have selected to do, and coding in ICD-10-CM is a fun, interesting activity for coders. Not so much for physicians.

In the 1986 movie, “Star Trek IV the Voyage Home” Dr. McCoy says, “the bureaucratic mind set is the only constant of the universe.” You remember the scene don’t you? Our heroes are in a shuttle … well, maybe you don’t remember the scene. Dr. McCoy wasn’t talking about ICD-10-CM, but we can imagine his reaction to it.

“Dammit Jim, I’m a doctor not a coder.”

But, in medical practices today, the physician typically selects the CPT code and the diagnosis codes that were the reason to provide the service and the diagnosis code that is the indication to order tests. ICD-10-CM will do little more than increase the bureaucratic side of physicians’ lives at the expense of their being doctors.

Betsy Nicoletti is president, Medical Practice Consulting and author of Auditing Physician Services. She blogs at Nicoletti Notes.

Comments are moderated before they are published. Please read the comment policy.

  • betsynicoletti

    Well, the AHA and the CDC developed it–I guess that means they take responsibility for it. The hospitals love it. But, the hospitals have coder who do all the coding. And, if they are sending queries to physicians now, just imagine the number of queries doctors will get after ICD-10.

    • Bob

      ICD codes are like IRS taxes in that you can have all the CPA’s with the same data, all come up with different tax amounts, so give every coder the same data and I’ll bet the computer inputs differ too. Don’t you? No? Try it!

      • betsynicoletti

        I think that if the coders have complete info (op report for example) they can probably select the same ICD-10 codes for a case. When dealing with less specific documentation,– well– if you get a lot of queries from coders now about your documentation, just imagine how many you’ll get when there are 5 times the codes. Maybe 5 times the queries.

  • ninguem

    As I’ve said for a long time.

    The ONLY people in favor of ICD-10 are those who stand to profit from it.

    Nice, for a change, to hear it directly from someone connected to the coding biz.

    • betsynicoletti

      Thanks. I would be happy to use the WHO version of ICD-10, 16,000 codes, if that is better. But why we need 68,000 codes when the current WHO version is so much smaller escapes me.

      • ninguem

        What I’ve noticed in international literature, is ICD-10 degrades performance, and it’s NOT a learning curve.

        “Learning curve” meaning you’re slow for a while, then back to normal as you get used to the new thing you had to learn.

        The degraded performance is permanent. That’s doctors, AND back-office personnel such as coders. Even in a “socialized” medical system (whatever that means) there still has to be some sort of “billing” internally, so people get credit for the work they’ve done.

        So efficiency is degraded. Fewer patients are seen, fewer claims are processed. You have to hire that many more medical personnel, that many more administrative and clerical personnel, and healthcare is that much more expensive.

        Hope it’s worth it.

        BTW, have there been outcome studies? Some great thing that happened because of ICD-10, that would not have happened under ICD-9.

        Because the only outcome study I can see, is the people who organize ICD-10 training courses get to buy a BMW.

        I drive a Subaru..

        • southerndoc1

          “ICD-10 degrades performance, and it’s NOT a learning curve”

          Just like EMRs!

  • betsynicoletti

    I don’t know. Hospitals are paid based on the diagnosis coding, while in fee-for-service medicine, physicians are paid based on CPT code. How this will change if we get value based/outcome based payment for physicians: that’s above my pay grade. It must be good for hospitals, or else, they are so invested in it at this point, they can’t go back. I have a lot of sympathy for groups and organizations that have invested heavily, but….

    • Bob

      Check out all the fraud cases at and see that after pharmaceutical Medicaid Rebate frauds, Hospitals upcodings are the next leading in fraud settlements.

  • betsynicoletti

    What is missing from the system that would be helpful to you? I think that rather than getting doctors on the same page, there are so many codes that policy makers, insurers, etc will need to map backwards from the specific to more general in order to have meaningful data.

  • betsynicoletti

    I agree with that. In the end, the doctor has to finish the note in the EMR, get paid and move on to the next patient.

  • doc99

    Why not simply skip ICD 10 altogether and move to ICD 11?

    • Ava Marie Wensko George

      Our infrastructure is not ready for ICD-11. The process is more than just flipping a switch. There are many areas that have to be remediated before we can even get to ICD-10. It’s like building a sailboat…..You can’t set the mast before building the boat.

      • Bob

        So you project as I do, 2053 before ICD-11′s are approached?

    • Bob

      ICD-11 includes mental illness and computer reforms, that haven’t been screwed up too much yet and can’t be accommodated as there aren’t enough mental illness facilities and psychiatrists to treat even a slightly larger number of patients. look to about 30 years from now just as it took the last 30 to switch from ICD-9′s to 10′s.

  • Ava Marie Wensko George

    Your article is very interesting and points out a very important fact – Physicians should not be expected to code. Quite frankly, physicians do not have the professional skill set. Yes, I do want physicians to treat patients, but I also want them to thoroughly document the entirety of the patient condition. Just because a physician know what is wrong with the patient, doesn’t mean that the next clinician does. It is part of what is required of physicians to clearly and concisely document, but most do not…..and yes, as a patient I want it clearly documented right or left, what kind of organism I am fighting, and how my accident happened. I just don’t think physicians should have to code it. ICD-10 is needed. I believe that it will happen. We cannot fall back and put our collective heads in the sand bemoaning and lamenting ICD-10. Just do it – Get ready.

    • Bob

      Yes Ava, I too believe ICD-10 will happen and will lead to the rapidly disintegration HIPPA, financial controls and lead to the end of PPACA as there will be no controls except ICD-10′s.

  • Bob

    But you left out all the good parts: the major increases in waste, frauds and abuse which obviously increase the current $1.2 trillion!

    • betsynicoletti

      How do you see ICD-10 doing this?

  • betsynicoletti

    Good point. Not all hospitals are enthusiastic about it, and I think smaller hospitals had less resources to spend. I think most hospitals assume they will need to staff up for ICD-10 and add coding staff–that’s what I hear, I don’t have citation for that. And that means more overhead, more non-clinical jobs.

    • SteveCaley

      Let’s go back to the fundamental premise. The principle in manufacturing the International Classifications of Disease is to allow the creation of an artificial language, if you will, that only has symbols for certain nouns, things called disease. That allows them to be shared across natural languages without requiring familiarity with the rest of that natural language. Chinese public health can be readily compared with Polish public health, in the area of Tb diagnosis and treatment, without having to learn ideograms or Polish.

      As long as they can be used as approximations to the natural ideas and language of medicine, they are innocuous. But, like Orwell noted, the ability to control a language represents the ability to control thought.

      Once the vast concept of Disease can be shoehorned into a digital meme, one can then assert that there is, for instance, a patient with M10.9/274.9 The patient needs M04AA01 or a similar M04AA class drug, to follow Best Practice.
      That sentence should be viewed as gibberish. It benefits the IT systems at the expense of treating the patient. But because of our fascination with IT, we will follow such Newspeak until we no longer have words for sick humans, simply codes.

  • betsynicoletti

    Always good to hear opinions from proponents and opponents. Thanks, Steve.

  • SteveCaley

    We don’t have to worry about ICD-10 for more than three years. The next
    version, ICD-11, is due for rollout then. There is no interoperable
    compatibility between ICD’s 9, 10, and 11. I swear I am not kidding

    Coding is based upon fundamentally absurd principles as
    well as intellectual laziness. The foreign principle – never proven by
    the slightest whisper of evidence – is that medicine is disease-based,
    organ-specific, and is neatly groupable into taxons and sub-taxons.
    None of those assumptions are useful laws for the physician’s work. In a
    recent job, I was assured that the hypoxia of congestive heart failure
    had to be listed in an internal visit code as Pulmonary – and I was
    scolded for not using beta-agonist inhalers before oxygen!

    coding is not benign – it will lead to bad quality evaluations and
    utilization reviews. Every code has a cocktail, or will in a few years.
    The Independent PA’s will go down the checklist and transmit the
    canonical medicines to the pharmacy. The reality of patient care will
    be jammed into algorithms, which benefit nobody.
    It is intellectually dishonest and lazy – as most “IT” projects are, and why they are drains on productivity. It does not help doctors do what doctors do – it decreases productivity, leading to bogus savings.

Most Popular