Health Information Management

Partial code freeze prompts concern about lingering problematic codes

JustCoding News: Inpatient, October 27, 2010

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

The decision to implement a partial code freeze for ICD-9-CM and ICD-10-CM/PCS may not have come as a total surprise when the announcement was made in September. But many in the provider community still have compelling concerns about codes that will presumably remain unchanged, particularly regarding stages of acute kidney injury, severity of coronary artery disease, and classifications of heart failure severity.

At the 2010 ICD-9-CM Coordination & Maintenance Committee Meeting last month, CMS representative Pat Brooks announced that the committee along with CMS and Centers for Disease Control (CDC) finalized the decision to implement a partial code freeze prior to implementation of ICD-10-CM/PCS in 2013.

“I was happy to see that they did not implement a total code freeze within the years preceding implementation in 2013 because it would really be unfortunate to not be able to capture new diseases and technology,” says Shannon McCall, RHIA, CCS, CCS-P, CPC, CPC-I, CCDS, director of HIM and coding at HCPro Inc., in Marblehead, MA.

The partial freeze will be implemented as follows:

  • The last regular, annual updates to both ICD-9-CM and ICD-10 code sets will be made on October 1, 2011.
  • On October 1, 2012, there will be only limited code updates to both the ICD-9-CM and ICD-10 code sets to capture new technologies and diseases as required by section 503(a) of Pub. L. 108-173.
  • On October 1, 2013, there will be only limited code updates to ICD-10 code sets to capture new technologies and diagnoses as required by section 503(a) of Pub. L. 108-173. There will be no updates to ICD-9-CM, as it will no longer be used for reporting.
  • On October 1, 2014, regular updates to ICD-10 will begin.

James S. Kennedy, MD, CCS, managing director at FTI Healthcare in Atlanta disagrees. “I am disappointed that the ICD-9-CM Coordination and Maintenance Committee opted to freeze the codes, given that ICD-9 and ICD-10 do not have the uniform specificity essential to severity and risk adjustment in physician value-based purchasing algorithms,” he says. “While I’m very grateful that ICD-10 is coming, it is not where it needs to be today, so I’m just sad that it will take a number of years to officially enhance the code set so that it benefits the doctor taking care of his or her patient.”

Kennedy notes, for example, that there is only one code for hyponatremia in ICD-10-CM. The normal range for serum sodium levels is 135–145 mEq/L. Although 133 mEq/L is just two points below normal, 113 mEq/L is practically fatal, according to Kennedy. “Yet there’s only one code, E87.1 (hypo-osmolality and hyponatremia), which doesn’t reflect the varying patient severity in these cases.”

If the code set doesn’t reflect the various stages of illness severity, physicians will be profiled inaccurately in Medicare’s physician value-based purchasing, medical home, and accountable care organization risk-adjustment programs, Kennedy says. “The freeze is for the convenience of the coder and the computer programmers,” he says. “It doesn’t serve the provider community, especially with landmark health reform legislation that has been passed, which emphasizes the need to profile and report provider efficiency.”

Update restrictions will prevent potential for added specificity

Even though the partial code freeze allows the possibility for updates to identify new diseases and technologies, McCall points out that as providers use the code set more, there could be instances for which updates would be beneficial. For example, what if there are problems with the translation between ICD-9 and ICD-10 codes? Or if it’s determined that an ICD-10-PCS code didn’t account for something (e.g., an approach was omitted), can minor changes be made to the PCS table that are not really for new technology?

“Hopefully we won’t encounter too many of those problems,” McCall says. But she questions whether the committee might consider remedying other types of situations that could cause coding dilemmas as well.

For example, consider the significance of the creation of code category 249 in 2008 for secondary diabetes mellitus, McCall says. Previously, there was only code category 250 for patients who had Type I or Type II diabetes. But there was a whole population who had diabetes mellitus due to an underlying condition or because they were drug-induced, for example.

“So it wasn’t really a new disease; it was a new way to classify a condition,” McCall says. “Are they going to consider that a new disease?”

Consider problematic codes that will remain

Stages of acute kidney injury. Consider the fiscal year 2011 Inpatient Prospective Payment System final rule. “For acute kidney injury, there can be a slight elevation of creatinine above baseline (e.g., 0.5 mg/dl), or it can be so severe (e.g., to above 4.0 mg/dl) that the patient requires acute dialysis,” Kennedy says. Yet there’s only one code in ICD-9-CM (548.9) and one code in ICD-10-CM (N17.9) reflecting acute kidney injury or failure, and neither code differentiates its severity. Because of that lack of specificity, Medicare changed ICD-9-CM code 584.9 from an MCC to a CC this year. “Hospitals that care for a higher volume of sicker acute kidney injury patients lose out,” Kennedy says.

Severity of coronary artery disease. Patients have three main coronary arteries beyond the left main artery: Left anterior, left circumflex, and right coronary artery. Which patient is going to have more problems with his or her heart disease—one with single-vessel disease or one with triple-vessel disease? Clearly, the answer is the patient with triple-vessel disease, as evidenced by a study of its natural history. But ICD-9-CM code 414.01 (coronary atherosclerosis of native coronary artery) and ICD-10-CM code I25.10 (atherosclerotic heart disease of native coronary artery without angina pectoris) don’t differentiate between patients with single, double, or triple vessel disease or even those with left main disease.

High specificity of the control of arrhythmias, such as atrial fibrillation or ventricular tachycardia. There’s a difference in resource utilization between atrial fibrillation that is in normal sinus rhythm due to a drug treatment (e.g., amiodarone), atrial fibrillation with a normal ventricular rate, and atrial fibrillation with a rapid ventricular rate, Kennedy says.

Although patients who have atrial fibrillation with normal ventricular response do not require inpatient admission, those with a rapid ventricular response do. However, there is only one code ICD-9-CM code (427.31) and one ICD-10-CM code (I48.0) for atrial fibrillation that is present in all these scenarios. Likewise, there is a difference in resource utilization and outcomes among ventricular tachycardia that is controlled with medication, nonsustained ventricular tachycardia, and sustained ventricular tachycardia. ICD-9-CM and ICD-10-CM have only one code, 427.1 and I47.2 respectively, to report these conditions.

Classifications of heart failure severity. Consumer Reports will soon publish The Society of Thoracic Surgeons (STS) profiles of cardiothoracic outcomes, Kennedy says, according to a recent news release. This is a clinically abstracted database that does not rely on ICD-9 codes. While expensive, nurses actually look at the records and apply their clinical knowledge to clinical interpret and report the patient’s condition.

STS stratified their patients with heart failure according to their New York Heart Association status (scale of 1 to 4). ICD-9 and ICD-10 databases will not have this advantage, given that these code sets will not differentiate the severity of heart failure. Some patients may have mild congestive heart failure (class 1), which the physician treats with medication. Then there are class 4 congestive heart failure patients who can’t even walk and are short of breath just sitting at rest. ICD-9-CM and ICD-10-CM both ignore these stratifications of heart failure severity, Kennedy says.

Medicare and private companies are increasingly profiling physicians and providers as to their 30-day mortality, 30-day readmission rate, and cost efficiency, he says. “Unless the code set can accurately reflect that patient’s severity of illness, physicians who take care of sicker patients will be inappropriately profiled,” Kennedy says. “If a patient has class 4 heart failure, it would be reasonable that it would cost more to take care of that patient than another patient who is class 1. But there is only one code assigned for both the very sick patient and the less sick patient.”

Freeze will restrict improvements to value-based purchasing programs

CMS is using coded data to profile hospitals and physicians, but one has to question how effective this can be when the code set does not translate the patient’s severity of illness, Kennedy says.

“Unfortunately, the CDC has chosen to freeze the ICD-9/ICD-10 codes with their current imperfections,” says Kennedy, who suggests that one potential solution would involve CMS enhancing their claims data using laboratory data according to algorithms that the Agency for Healthcare Research and Quality is developing under the direction of Irene Fraser, PhD. (To learn more, access her presentations discussing this concept.)

“If CMS would allow for laboratory data to authenticate the severity of anemia, hyponatremia, acute kidney disease or injury, and other similar diseases, then the laborious process of clinical documentation improvement can be ameliorated,” Kennedy says. (Access this article from the Journal of the American Medical Association to learn more about the enhancement of claims data to improve risk adjustment of hospital mortality.)

“My hope and prayer is that the physician specialty societies that are affected by value-based purchasing will take a second look at the ICD-9 and ICD-10 codes and work to advocate specificity so that value-based purchasing programs can be properly constructed,” Kennedy says. “And I hope the cooperating parties are receptive and that they solicit responses from physician specialty organizations for code sets that would lend themselves to ICD-10.”

Editor’s note: E-mail your questions to Managing Editor Doreen V. Bentley, CPC-A, at

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

Most Popular