Last month's column discussed that the implementation of International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10) will not impose dramatic new documentation requirements on physicians, but also mentioned that there have traditionally been pervasive documentation deficiencies that negatively impact revenue, quality, outcome, and pay-for-performance measures.
The need to correct these problems will continue under ICD-10, and in this month's column, we will explore how the new system provides some additional important documentation opportunities.
Coma is a very important diagnosis that profoundly affects revenue and quality/performance measures for both ICD-9 and ICD-10. Unfortunately, the term “coma” is infrequently documented in the record, even when a profound alteration in the level of consciousness has occurred. Other clinical terminology will not suffice to assign the correct code for coma.
With the introduction of ICD-10, the Glasgow Coma Scale (GCS) takes on new importance. To accurately represent the patient's condition with the correct codes, the GCS should be calculated for all patients with an altered level of consciousness and responsiveness, whether the cause is traumatic or nontraumatic.
Technically, the GCS classification of altered levels of consciousness and responsiveness could be considered mild, moderate, or severe coma. A GCS of 8 or less is severe and can be considered an unequivocal diagnostic standard for coma. Other clinical terminology is suggested for GCS scores greater than 8 (Table 1). Whether a diagnosis of mild or moderate coma can be considered valid for the assignment of a code for “with coma” is controversial and as yet undecided.
The most severe components of the GCS can also be assigned an ICD-10 code having the same impact as the documentation of “coma” (Table 2).
These components can add severity (but not additional reimbursement) in cases in which coma has already been documented. Coders may assign the individual component codes based on a scoring sheet recorded by someone other than the treating clinician, including the emergency medical service prehospital records. It is not necessary to assign a code for the total GCS score if the individual components are coded.
Hepatic encephalopathy has been reclassified by ICD-10 as “hepatic failure.” Hepatic failure codes allow for greater causal specificity, such as acute/subacute, chronic, alcoholic, and drug-induced, but all of these, as well as the code for an unspecified cause, have the same impact on revenue and quality/performance measures. The coder may assign greater specificity to a diagnosis of hepatic failure based on other documentation in the medical record.
The coding of hepatic encephalopathy has also changed to reflect whether it is associated with coma during the current encounter. It appears that a specific diagnosis of “coma” would have to be documented in the current episode of care to assign the “with coma” code, which carries great weight for revenue and quality/performance measures, unlike the “without coma” code. However, if the GCS scoring sheet includes any of the components listed in Table 2, a code for that component would also be assigned, having the same impact as a diagnosis of “with coma.”
If the patient with hepatic encephalopathy is only described as obtunded or lethargic (Table 1), it appears that the “with coma” code cannot be assigned. Such a patient could be described as having moderate or mild coma, but whether this would be considered clinically valid for the assignment of a code for “with coma” is controversial
The codes for pancreatitis have been expanded to include some specific causes: idiopathic, biliary, alcohol-induced, drug-induced, other, and unspecified. Documenting this specificity is desirable but not a requirement, since the impact on revenue and quality/performance measures has not changed and is the same for all causes. And once again, the coder may pick up greater specificity from other information in the medical record.
While the coding of diabetes has changed substantially, there's very little additional documentation required under ICD-10 that wasn't already needed with ICD-9. One improvement is elimination of the controlled versus uncontrolled distinction. Whenever there is any physician documentation indicating an abnormally high blood glucose level, a code is assigned for hyperglycemia. Such documentation terminology might include hyperglycemia, elevated blood sugar, poorly or inadequately controlled, and uncontrolled. If diabetes is documented as controlled, then the code for hyperglycemia is not assigned.
Perhaps the most surprising and clinically inconsistent change in ICD-10 is the elimination of specificity for hypertension. All hypertension (other than maternal and that due to heart or kidney disease) is assigned to a single code, I10. There is now no code specificity to express the clinical distinction between well-controlled benign hypertension and potentially life-threatening hypertensive crises (formerly coded as malignant or accelerated in ICD-9).
Henceforth, malignant (emergent) or accelerated (urgent) hypertension will be classified the same as benign hypertension, having no severity impact on revenue and quality/performance measures.
In summary, the vast majority of common, crucial documentation challenges that have vexed doctors and hospitals for many years under ICD-9 have not changed with ICD-10 (Table 3). They represent by far the greatest opportunities for documentation improvement, having an impact on revenue, quality, outcome, and pay-for-performance measures for both physicians and hospitals. Nationwide, pervasive physician documentation deficiencies in these areas persist and will need attention going forward under ICD-10.