It seems like a near-daily occurrence these days that a blog, industry survey, or Centers for Medicare & Medicaid Services (CMS) announcement reminds us all of what a success ICD-10 was. It has been common for pundits to refer to the fear leading up to the event as a Y2K-esque false alarm. I am guilty as well, considering that things could have gone far worse, with loss of production and unforeseen costs being two of the most commonly discussed topics associated with ICD-10 challenges. Both the denial situation and the invalid code issues were not as bad as we thought, either.

So it was a complete and total success, right? Well, that depends on the definition of “success” you are using. It seems that the standard for this definition being used in the industry panders to the lowest common denominator; i.e. “the world didn’t end.” Using that standard, I would say yes, ICD-10 implementation was a resounding success. However, for those of us in the industry focused on making sure the reporting reflects the clinical truth of what was actually wrong with patients, ICD-10 seems decidedly less successful.

Let’s highlight some of the biggest ICD-10 failures. Going through the numbers:

Indexing Problems (Using a Code Book)

Distress, acute, respiratory (adult) (child): J80. ICD-10 Code J80 is an inappropriate code assignment for acute respiratory distress, as it includes conditions such as adult hyaline membrane disease and acute respiratory distress syndrome.

Delirium superimposed on dementia can be found in software groupers, but I cannot look it up in my code book under F05. Even worse, due to productivity concerns and formatting in software groupers, I have found that many coders do not read what is in the “includes” notes for codes such as this, meaning they are not aware of all the terms that should be reported with the correct code.

ICD-10 encourages consistent documentation, yet assigns the terms inconsistently for respiratory compromise, for example. Medical-related respiratory failure is termed as “failure” while surgical respiratory failure is termed “Insufficiency,” and no further guidance is provided nor defined for us in Coding Clinic or official guidelines. 

PVD unspecified, a diagnosis/description often used to describe a venous disease process, codes to I73.9, a code describing an arterial disease process.


Intractable vomiting, G43.A0, and cyclical vomiting, not intractable group to DRG 103, Headache w/o/MCC. What if the origin of the intractable vomiting was gastrointestinal or neurological, and not related to migraines and headaches? In the ICD-9 world, cyclical vomiting (not psychogenic) appropriately grouped to a GI DRG: 392 Esophagitis, gastroenteritis, and misc. digestive disease.

Obesity with hypoventilation syndrome and a gastric banding procedure groups to DRG 989, Non-extensive OR Procedure unrelated to principal diagnosis. In the ICD-9 world, obesity with a laparoscopic banding procedure appropriately grouped to DRG 621: OR procedure for obesity.

Cirrhosis with portal hypertension and bleeding esophageal varices receiving an EGD with band ligation groups to DRG 981, Extensive OR procedure unrelated to principal diagnosis. In the ICD-9 world, this combination grouped to DRG 432, Cirrhosis and alcoholic hepatitis. Perhaps even worse, Coding Clinic (apparently focusing on just coding assignment and not interacting with CMS regarding grouper decisions) specifically addressed this and declared it to be the correct reporting.

Pneumonia with bronchial alveolar lavage selected as “non-diagnostic” groups to DRG 165, Major chest procedures, in ICD-10. Many coders believe that working off a strict coding definition, the non-diagnostic modifier is the correct choice, while clearly DRG 165 is inappropriate. In the ICD-9 world this would have grouped to DRG 195, Simple pneumonia. 

MCC/CC Status

Some things associated with DRG assignment just don’t add up, and they were not fixed with ICD-10. For example, persistent vegetative state is a CC, while functional quadriplegia is an MCC. Calling this non-intuitive would be an understatement.

Coding Guidelines

Coding guidelines specify a definition of “pathological” as that which occurs as a result of minor trauma as long as that minor trauma would not result in a fracture in a healthy bone (in the presence of osteoporosis, malignancy, or some other disease process), yet many facilities are not applying this guideline and incorrectly identifying fractures reportable as pathological. Even worse, many facilities are still struggling with defining an initial encounter versus a subsequent encounter for fractures in general.

Coding guidelines prohibit the separate reporting of signs and symptoms or diagnoses integral to diagnoses already reported, but they then turn around and recommend that metabolic encephalopathy be reported in patients with confusion from diabetic hypoglycemia (while previously advising against metabolic encephalopathy during the post-ictal state). If one is integral to the diagnosis, isn’t the other?

Coding guidelines also direct us to utilize the Glasgow coma score to add additional severity to acute CVA, and as long-term sequela of CVA (when applicable), yet many facilities and physicians refuse to apply the Glascow coma scale to their patient population, focusing instead on the NIH scoring method.


ICD-10 brought with it a promise of getting complicating co-morbidity credit for atrial fibrillation in the form of “persistent atrial fibrillation,” yet almost no one can agree to a definition of persistent atrial fibrillation and when to report it.

ICD-10 provided us with a code for a drop in hgb/hct (R71.0) which is a CC. Great, right? Except that most either don’t’ know about it, don’t know how to index it, can’t agree on the clinical significance of the finding, or apply the same standards to this sign and symptom code as they would to the code for acute blood loss due to anemia (D62).

Unable to Be Reported in ICD-10

  • Acute cor pulmonale was reportable in ICD-9 but is not reportable in ICD-10 unless the patient has an acute pulmonary embolism.
  • Hepatic encephalopathy was reportable in ICD-9 but is not reportable in ICD-10.
  • Demand mediated/MI Type I-II: Also missing from ICD-9, the current proposal for new codes at the end of the code freeze holds the promise of new code entries to capture this common diagnosis.
  • Peptic ulcer disease with obstruction was reportable in ICD-9, but is not in ICD-10.
  • DKA in a patient with DMII is reportable in ICD-9, not ICD-10.
  • Esophageal hemorrhage is reportable in ICD-9, not ICD-10.

Both ICD-9 and ICD-10 have largely ignored the progression of CHF documentation by cardiologists, including the NYHA/ACC classifications. Coding Clinic rejects the coding of heart failure with preserved or rejected ejection fraction, a problem that could easily be resolved with new codes or an update to the indexing terms. This suggestion was made in the September 2015 ICD-10 coordination and maintenance diagnosis agenda. 

Listed above were all things that ICD-10 held the promise of correcting, yet no corrections came, and in several cases, new problems that we did not have in ICD-9 were created. Partly to blame is the five-year code freeze. CMS recently announced that it was opening up the floodgates for new codes this year, and the Centers for Disease Control and Prevention (CDC) has listed a group of proposals for roughly 6,000 codes. Unfortunately, a cursory review of the proposed codes only fixes a few of these issues (hepatic encephalopathy, DMII with DKA, and Type II MI, for example). Hopefully some of the indexing and grouping problems not specified under the category of “new codes” will also be corrected come Oct. 1.

Share This Article