ICD-10 Coding Corner: Recap of Coding Clinic, First Quarter 2013

The following is a recap of the issue with an eye toward what it means for CDI specialists.

 

In this entry, Coding Clinic states to assign Z87.898 (Personal history of other specified conditions) to a patient with a past history of methamphetamine abuse. This entry serves as a reminder for CDI specialists to educate physicians about documentation of abuse vs. dependence. Physicians are often hesitant to document abuse as they often worry about being judgmental, Ericson says, but there are strategies to overcome this resistance.

 

“Physicians need to become more comfortable with using the term dependent—i.e., long term cancer patients dependent on opiods. Also, it is helpful if you involve the pain management team. They’re more likely to highlight that someone is on long term medication,” she says. “It is always helpful to interact with as many members of the healthcare umbrella as possible—they can become good advocates.”

 

Coding Clinic for ICD-9-CM states to assign code I51.3 (Intracardiac thrombus, not elsewhere classified) for a left atrial appendage thrombus. This entry highlights the need for a strong understanding of anatomy and physiology in mastering ICD-10 coding.

 

“This demonstrates the continued need for the partnership between CDI and coding, because CDI has clinical expertise to know that you wouldn’t equate a thrombus in an atrial appendage to a cardiac type event that’s similar to a myocardial infarction (MI). Coronary arteries aren’t the same as an atrial appendage,” Ericson says.

 

If your hospital has an internal Internet firewall that doesn’t allow for Internet use/search by employees, ask to have it disabled prior to the go-live date of ICD-10.

 

“If you search the Internet for left atrial appendage, you’ll see that it’s different from a left atrial thrombus,” Ericson says. “Staff members needs to have access to electronic resources for anatomy to expedite code assignment.”

 

In this entry, Coding Clinic states to assign I21.02, ST elevation (STEMI) myocardial infarction involving left anterior descending coronary artery, for an AMI unspecified as either STEMI or non-STEMI of the left anterior descending coronary artery (stated duration of four weeks or less). This piece of guidance reinforces that STEMI—a more severe type of AMI—is the default for AMIs.

 

“If you look at the coding guidelines, you’ll see a hierarchy that STEMI supersedes a non-STEMI,” Ericson explains. If a patient starts out as non-STEMI, and their heart attack progresses, you code to a STEMI. If they start as STEMI, get TPA, and their blockage clears, you still code it as a STEMI.”

 

However, one problem with this default hierarchy is that American Heart Association runs a program called “Mission Lifeline” which only tracks STEMIs. If physicians don’t understand that the default AMI is a STEMI—in other words, if they document “AMI” without specifying it as STEMI or non-STEMI—it defaults to STEMI and skews the Mission Lifeline data.

 

“You may have more Mission Lifeline failures—coding will be picking up more STEMIs than actually occurred,” Ericson says.

 

ICD-9 did not use STEMI vs. non-STEMI for coding purposes. In ICD-9, everything was an AMI. In short, ask your physicians to clarify non-STEMIs vs. STEMIs in ICD-10.

 

In this entry Coding Clinic for ICD-9-CM states that a coder must assign E13.10, Other specified diabetes mellitus with ketoacidosis without coma, for a patient with Type 2 diabetes with ketoacidosis. This advice highlights the unfortunate shortcomings of the ICD-10 code set.

 

“Type 2 diabetes should be an E11 code, but ketoacidosis rarely occurs in Type 2 diabetes,” says Ericson, noting that Medlineplus has a helpful reference for Type 2 diabetes with ketoacidosis. “So we need to use an “Other specified diabetes” E13 code.”

 

Coding Clinic for ICD-9-CM states that this issue will go to before the ICD-10 Coordination and Maintenance Committee for reclassification in order that patients with ketoacidosis can be properly identified.

 

This entry seems to be a welcome relief for coders (and CDI specialists) in that it allows coders to assign a more specific fracture code based on documentation in a radiology report, rather than clarifying with the attending physician. This was permitted in ICD-9-CM, but Coding Clinic for ICD-9-CM stated previously ICD-9-CM guidance will not apply to ICD-10. In this case, it will.

 

Ericson anticipates the guidance could prove troublesome. The question writer refers to “coding guidelines in ICD-10” rather than Coding Clinic for ICD-9-CM. It is important to remember that Coding Clinic for ICD-9-CM and the Official Guidelines for Coding and Reporting are different documents. Ericson also wonders if this advice is limited to radiology reports, or whether it can be interpreted to include any diagnostic imaging.

 

“Right now Coding Clinic for ICD-9-CM just says radiology reports, but computed tomography (CT) scans of the head are done for stroke. It would be nice to be able extrapolate the exact location of the blockage from the CT, and the affected artery,” she says. “Right now we don’t know if physicians need to document to this level of specificity, or if we can get this detail from the CT report.”

 

Ericson also wonders how Coding Clinic for ICD-9-CM considers laterality (which is needed to code fractures and other procedures in ICD-10). For example, is laterality part of the location, which has to be documented by the provider, or is laterality considered specificity, in which case it could be taken from radiological imaging?

 

“It may seem like I’m splitting hairs, but consider pressure ulcers. Can we obtain laterality from other sources that are already allowed to define specificity, like wound care notes? Do you really want to have to query physicians for laterality?” Ericson asks.

 

Organizations must decide if their CDI specialists will review/clarify procedures. Outside of debridement, CDI specialists typically do not review operative procedures in ICD-9 because they don’t add a complicating condition (CC) or major complicating condition (MCC). However, this issue of Coding Clinic for ICD-9-CM reinforces the complexity of coding under PCS, which may require CDI involvement.

 

“We don’t know how PCS will affect coder productivity, but it’s likely to have a significant impact, much more than is being anticipated,” Ericson says. “Organizations may want to consider hiring CDI staff with surgical experience to assist coding with querying and/or accurately assigning PCS codes.”

 

For example, Coding Clinic for ICD-9-CM states on p. 27 that a coder should assign ICD-10-PCS code 031C0ZF (Bypass left radial artery to lower arm vein, open approach) for creation of a left proximal radial artery fistula. The question asks, “How does the coder know which direction the blood is flowing (‘from’ what body part, ‘to’ what body part)”—which is needed in order to correctly assign the 4^th character of body part and the 7^th character qualifier with the root operation bypass. A CDI specialist is more likely to know this information, bypassing the need for a physician query.

 

The elements of a PCS code are intertwined. The first three characters (e.g., section, body system and root operation) of a PCS code are necessary to identify the correct table that will be used to assign the additional characters; however, assignment of the root operation often depends uon the possible table values for body part and device. 

 

“The problem is the root operation is dependent upon knowledge of the body part and/or device, which forces you to work backwards,” Ericson says. “I can see where it may be problematic to accurately assign a PCS code if the encoder forces code assignment to proceed sequentially based on character order.”

 

This issue of clinical knowledge is further reinforced with Coding Clinic for ICD-9-CM’s advice on p. 27 regarding ICD-10 code assignment for injection of a sclerosing agent into an esophageal varix of the lower esophagus. The correct code is 3E0G8TZ (Introduction of destructive agent into upper GI), clarifying the use of “introduction” vs. “destruction.”

 

“This highlights that root operations will be a trial and error process. Again, this is why you might want CDI with a surgical background, because a physician doesn’t have to use root operation terminology per ICD-10-PCS coding guidelines —they just have to describe the procedure as such that a root operation can be correctly assigned,” Ericson says. “I think there will be a lot of struggle with PCS—you can’t bill a procedure without all seven characters and there are no default codes.

 

“Remember coder productivity is expected to decrease under ICD-10-CM/PCS so the coding department may not have the resources to query for PCS codes—so this job may default to CDI,” she adds. “Failing to assign a PCS code significantly impacts reimbursement, as surgical DRGs are typically paid at a higher rate compared to medical DRGs. Hospitals need to develop an action plan to address PCS querying.”