It is true that the number of diagnosis codes in ICD-10 increases by approximately a factor of five over ICD-9. Many physicians bemoan the fact that they will have to learn five times as many terms as they knew under ICD-9. Yet this simply reflects a lack of knowledge. Consider, for example, a non-traumatic subarachnoid hemorrhage. Under ICD-9, this diagnosis fell under the categorical single code for subarachnoid hemorrhage. In ICD-10-CM, there is a specific category for non-traumatic subarachnoid hemorrhage featuring 20 specific codes. Going from one to 20 codes is a big leap. Does this reflect a lot of new learning for physicians?

Hopefully not.

What the code expansion actually represents is huge advances in clinical medicine made in the 35 years since ICD-9 was drafted. Specifically, we now have technology (CT/MRI/angiography) that can identify the specific artery involved (middle cerebral, basilar, anterior communicating, etc.) as well as laterality for many of those blood vessels. It is basic, responsible practice of medicine to identify the nature of the non-traumatic subarachnoid hemorrhage, in today’s clinical world.

Cardiology is a specialty for which terminology has improved significantly. The “new” terms for acute myocardial infarctions in ICD-10-CM are STEMI and NSTEMI (terminology that, incidentally, has been used by cardiologists, internists, and others for the last couple decades).

One area of legitimate concern among cardiologists relates to specificity regarding type of heart failure (acute or chronic, systolic, diastolic, or combined). Many cardiologists historically have not made such a differentiation, but they should acknowledge that this terminology is not new under ICD-10-CM – it’s also the codeable terminology found in ICD-9.

Is there a high degree of specificity in ICD-10-CM? Absolutely. And it is perhaps nowhere better exemplified than in orthopedic diagnoses. Just as an example, there are a myriad of codes for fractures of the femur. That sounds extreme until one actually learns how ICD-10-CM works. The new coding system includes terminology that every orthopedic resident must learn: the Gustilo classification of open fractures, the Salter-Harris classification of epiphyseal fractures, as well as other essential information such as specific anatomic site, type of fracture (spiral, etc.), and of course, laterality.

For neurosurgeons and trauma surgeons, ICD-10-CM includes provisions for the documentation of attributes of head injuries such as duration of loss of consciousness for cerebral edema as well as Glascow coma scores.

A comment from an internist in Anchorage, Alaska put ICD-10-CM into sharp perspective. He said (and I agree) that ICD-10-CM is exactly what we should expect from a resident in a morning report. So should we expect any less from practicing physicians? This observation should be reassuring to most physicians. I will admit, however, that the challenge is not the same for all specialties. Generalists, whether they are trained in family medicine, internal medicine, or emergency medicine, treat patients with a broad scope of diagnoses and will require more assistance transitioning to ICD-10-CM.

I should also mention ICD-10-PCS, the procedural coding system. Here we have a far greater increase in the number of codes. The marked expansion in the number of codes in ICD-10-PCS is primarily due to the architecture of the PCS system. ICD-10-PCS codes capture information impossible to capture in ICD-9, such as specific body parts, types of operations, types of devices left in a patient, and more. While this seems like it might present a lot more work for the physician, in most instances a quality operative report will include all the necessary information for coding, particularly if physicians are educated on basic code structure. It is unnecessary and absolutely not recommended that physicians memorize any ICD-10-PCS codes.

Let me share an observation from dealing with hospitals and medical staffs from across the country. The strongest physician advocates of ICD-10 appear to be those working in academic medical centers. That is, perhaps, because specificity and accuracy of diagnoses and procedure documentation are key components of residency and fellowship training. Additionally, most academic physicians recognize the lack of utility of ICD-9 for clinical research or epidemiology. ICD-10 provides much better information for such purposes.

I will conclude with one final observation. ICD-10 presents different challenges to different constituencies. I am very sympathetic to coders. They are essentially learning a new language, and they need to learn it for all specialties, all diagnoses, and all procedures. And here is where some of the confusion may have arisen for physicians who have not as yet learned much about ICD-10. We, fortunately, do not face the same challenges. For the vast majority of physicians, we can constrain our focus to our specialty areas.

In so doing, we quickly realize that the information needed for ICD-10 generally mirrors that which is required for high-quality patient care.