Inpatient coders are used to being able to code conditions documented as possible, probable, suspected, or rule out, as if they were in fact confirmed. Outpatient coders can’t do that. They need a confirmed diagnosis.
However, in one case, inpatient coders always must have a confirmed diagnosis in order to report the condition: HIV. Regardless of setting, coders can only code confirmed cases of HIV. You don’t need a diagnostic lab test for a confirmation. In this context, you just need the provider’s diagnostic statement that the patient is HIV-positive, or has an HIV-related illness.
Patients with HIV can suffer from a host of related conditions, so you need to know the sequencing guidelines. Here’s the good news: the guidelines in ICD-10-CM are the same as the guidelines in ICD-9-CM. They just look different.
So, a quick review of HIV coding guidelines.
If the patient is admitted for an HIV-related condition, report B20 (human immunodeficiency virus [HIV] disease) as the principal diagnosis, followed by additional diagnosis codes for all reported HIV-related conditions.
If the patient is admitted for an unrelated illness or injury, such as a traumatic fracture, sequence the unrelated condition as the principal diagnosis, followed by B20 and coders for any reported HIV-related conditions.
Remember that whether the patient is newly or previously diagnosed does not affect sequencing. You go by the reason for the admission.
What happens when a physician documents that a patient is HIV-positive, but has no symptoms? Report Z21 (asymptomatic human immunodeficiency virus [HIV] infection status). You would not use the Z code if the physician documents that the patient has AIDS, or is treating the patient for any HIV-related illness, or describes the patient as having any condition(s) resulting from his/her HIV-positive status. Z21 is only for asymptomatic cases of HIV.
Once a patient has developed an HIV-related illness, you will always assign code B20 on every subsequent admission/encounter. Never report R75 (inconclusive laboratory evidence of human immunodeficiency virus [HIV]) or Z21 for patients previously diagnosed with any HIV illness. Once a patient has HIV, he or she always has HIV. We haven’t cured it yet.
Brush up on your knowledge of cardiovascular system anatomy as you learn how to code cardiovascular diseases in ICD-10-CM during the live, 90-minute webcast Reduce the Fear of ICD-10-CM Cardiovascular Coding!
Join JustCoding at 1 p.m. (Eastern) Thursday, March 6, as expert speakers Gerri Walk, RHIA, CCS-P, and Laura Legg, RHIT, CCS, give you the inside track on codes and guidelines for cardiovascular conditions in ICD-10-CM. They will discuss what you need to see in the documentation in order to assign the most appropriate code.
Gerri and Laura will present a companion webcast, Reduce the Fear of ICD-10-PCS Cardiovascular Procedures, April 10 to give you the ins and outs of cardiovascular procedure coding in ICD-10-PCS. They will address documentation requirements and provide tips and takeaways to assist
inpatient coders in preparation for this new coding system.
For more information or to order, call 800/650-6787 or visit the HCPro Healthcare Marketplace.
Can’t join us live? No problem! You can order both webcasts on-demand, meaning you can watch whenever it’s convenient for you. And you can train your entire team, live, on-demand, or both!
I really want the t-shirt that says, “I only do what the voices in my head tell me” and its companion shirt, “The voices in my head don’t like you.” Sadly too many people I know might believe it.
It depends on why you’re hearing voices. People hear voices all the time, sometimes for totally benign reasons. For example, you may hear voices when you fall asleep. Not to worry, they’re just hypnagogic hallucinations, part of falling asleep.
ICD-10-CM includes six different codes for hallucinations, including one for auditory hallucinations (R44.0). You’ll find these codes under category R44 (other symptoms and signs involving general sensations and perceptions). You would only report these codes if the physician doesn’t know why the patient is hallucinating.
However, plenty of medical conditions also cause hallucinations. For example, people who abuse drugs and alcohol can experience hallucinations. For these patients, you would need to know whether their drug or alcohol use falls under:
This is a new concept in ICD-10-CM. Use, abuse, and dependence even have their own hierarchy. The ICD-10-CM guidelines tell us the following:
- If both use and abuse are documented, assign only the code for abuse
- If both abuse and dependence are documented, assign only the code for dependence
- If use, abuse and dependence are all documented, assign only the code for dependence
- If both use and dependence are documented, assign only the code for dependence
Dependence, when documented, always trumps use and abuse.
In ICD-10-CM, you’ll find the code broken out by substance causing the psychotic disorder (i.e. hallucinations). Our choices include:
- F10.251, alcohol dependence with alcohol-induced psychotic disorder with hallucinations
- F11.151, opioid abuse with opioid-induced psychotic disorder with hallucinations
- F12.951, cannabis use, unspecified with psychotic disorder with hallucinations
- F13.251, sedative, hypnotic or anxiolytic dependence with sedative, hypnotic or anxiolytic-induced psychotic disorder with hallucinations
- F14.951, cocaine use, unspecified with cocaine-induced psychotic disorder with hallucinations
We need documentation of the substance and the level of use, as well as the specific psychotic symptom.
Alternately, a patient may be suffering from a mental illness that is causing his or her hallucinations. Schizophrenia may be the mental illness that comes to mind. If the patient does indeed suffer from schizophrenia, we need to know the type in order to choose from these codes:
- F20.0, paranoid schizophrenia
- F20.1, disorganized schizophrenia
- F20.2, catatonic schizophrenia
- F20.3, undifferentiated schizophrenia
- F20.5, residual schizophrenia
- F20.81, schizophreniform disorder
- F20.89, other schizophrenia
Many of these codes feature a list of inclusive terms underneath them. For example, catatonic schizophrenia includes:
- Schizophrenic catalepsy
- Schizophrenic catatonia
- Schizophrenic flexibilitas cerea
Patients with major depressive disorder may also hear voices. In order to code this condition, we need to know whether the patient is suffering from a single episode with psychotic features (F32.3) or recurrent major depressive disorder with psychotic symptoms (F33.3).
Bipolar patients may also suffer from auditory hallucinations. ICD-10-CM breaks down the bipolar disorder codes to identify the type of episode the patient has:
- F31.2, bipolar disorder, current episode manic severe with psychotic features
- F31.5, bipolar disorder, current episode depressed, severe, with psychotic features
- F31.64, bipolar disorder, current episode mixed, severe, with psychotic features
Patients suffering from post-traumatic stress disorder (PTSD) may also hallucinate. The ICD-10-CM codes for PTSD do not specify whether the patient is suffering from psychotic symptoms. They are divided into:
- Acute (F43.11)
- Chronic (F43.12)
- Unspecified (F43.10)
A range of organic brain disorders, such as brain tumors, temporal lobe epilepsy, and viral encephalitis, can also cause hallucinations. If the hallucinations are integral to the disease process, we don’t code them separately. You may need to ask a clinician or clinical documentation improvement specialist if you’re not sure.
If you’ll excuse me, I hear a cookie calling my name.
CMS reversed course earlier this week and announced it will conduct end-to-end ICD-10 training with a sample of providers. Previously, CMS had stated it would not conduct any end-to-end testing.
First off, who will be involved: Providers can sign up with their MACs to volunteer to take part in the testing. Volunteering does not equal participating.The MACs and CMS will choose a statistically significant diverse group of volunteers to conduct testing. Look for information about volunteering on your MAC’s website in March. Participants will be chosen by MACs and CMS in April.
Second, when will the testing occur: CMS representatives stated late July. That doesn’t give providers much time to fix any technical problems uncovered during the end-to-end testing.
And speaking of problems, CMS representatives did not provide any information about what the agency will do with the results of the testing. Hopefully, if CMS sees a recurring problem, the agency will alert all providers and everyone can fix it before October 1.
CMS representatives twice reiterated that the ICD-10 implementation date remains October 1, 2014, and implementation will not be delayed again.
One of the things that drives me crazy about how media, Congress, and the AMA discuss ICD-10 codes is their focus on the External Causes codes. Granted some of those codes are silly or strange or seem pointless.
The ICD-10-CM guidelines point out that coders are not to report External Causes codes. Some states or payers may require External Causes, but reporting these codes is otherwise optional. However, the guidelines recommend reporting these codes anyway because they provide valuable data.
So you don’t need to report W58.11XA (bitten by crocodile, initial encounter) or X06.1XXA (exposure to melting of plastic jewelry, initial encounter). The codes are there, they help tell a better story about the patient, and they can be useful for data collection.
External Causes codes include some very detailed codes for injuries involving motor vehicles, such as:
- V39.50-, passenger in three-wheeled motor vehicle injured in collision with unspecified motor vehicles in traffic accident
- V43.42-, person boarding or alighting a car injured in collision with other type car
- V65.00, driver of heavy transport vehicle injured in collision with railway train or railway vehicle in non-traffic accident
Those aren’t the codes you need to worry about, though. Instead of focusing on the External Causes codes, coders (and the AMA, media, and Congress) should be looking at the injury and illness codes. For example, in Chapter 19 (Injury, poisoning and certain other consequences of external causes [S00-T88]), we find six choices in category S90.1- (contusion of toe without damage to nail):
- S90.111-, contusion of right great toe without damage to nail
- S90.112-, contusion of left great toe without damage to nail
- S90.119-, contusion of unspecified great toe without damage to nail
- S90.121-, contusion of right lesser toe(s) without damage to nail
- S90.122-, contusion of left lesser toe(s) without damage to nail
- S90.129-, contusion of unspecified lesser toe(s) without damage to nail
In order to be valid codes, each of those choices needs a seventh character:
- A, initial encounter
- D, subsequent encounter
- S, sequela
So in reality, we have 21 possible codes for a contusion of the toe without damage to the nail. In ICD-9-CM, we have one possible code: 924.3 (contusion of toe). We also have 21 codes for contusion of a toe with damage to nail. Still reporting 924.3 in ICD-9-CM.
If you want to talk about the increased number of codes and the increased specificity and granularity of ICD-10-CM, those are the kind of codes you should be discussing. Those are the codes that are clinically significant and describe the patient’s condition. Those are the codes we need better documentation to report. Those are the codes that will get us paid.
Worry about those codes and their details and leave the External Causes alone.
Maybe the AMA’s letter did the trick. Or maybe CMS just thought better of its decision not to conduct end-to-end testing prior to ICD-10 implementation.
According to MLN Matters® SE1409, CMS will select a “small sample group of providers” and conduct end-to-end testing in summer 2014. The testing will include submission of test claims to CMS with ICD-10 codes and the provider’s receipt of a Remittance Advice that explains the adjudication of the claims.
That’s a pretty vague plan, but at least it’s a plan. We don’t know yet how many providers will make up the small sample size or how CMS will pick the providers. CMS states that the providers will represent a broad cross-section of provider types, claims types, and submitter types.
In its letter to CMS urging recommending another ICD-10 implementation delay, the AMA recommended that CMS conduct end-to-end testing with at least 100 different physician practices of varying sizes and specialties.
Summer 2014 is also a vague timeframe. Is CMS going to conduct testing in May (summer officially begins June 21) or will it wait until September (the first day of fall is September 23)?
When CMS conducts the testing could be as important as the results. If CMS tests early (April-June), fewer organizations may be ready to test, limiting the potential sample size. If CMS waits (August-September), organizations may not have enough time to correct problems prior to ICD-10 implementation.
MACs will be conducting front-end testing March 3-7 with trading partners, such as healthcare organizations, according to MLN Matters MM8465. It is not a full-cycle test and the MACs will not adjudicate the claims. In MLN Matters SE1409, CMS stated it will consider adding additional weeks of front-end testing as well.
As nice as it would be for CMS to conduct end-to-end testing with every organization, that’s not very realistic. I’m interested to see what sample size CMS uses, which providers it chooses, and how the testing goes. I’d love to see CMS conduct end-to-end testing with some of the organizations that run into problems during the front-end testing.
CMS also did not state whether it would publish the results of the testing or call out common problems. So we’ll have to wait and see. At least CMS will perform some end-to-end testing and that’s a step in the right direction.
What did you get for Valentine’s Day? Flowers? Chocolate? Mono? It is the kissing disease after all.
ICD-9-CM includes only one code for infectious mononucleosis: 075. That code includes glandular fever, monocytic angina, and Pfeiffer’s disease. Prepare to be shocked: ICD-10-CM offers many more choices for mono. The new codes identify the cause of the patient’s mononucleosis:
- B27.0, gammaherpesviral mononucleosis (also known as the Epstein-Barr virus)
- B27.1, cytomegaloviral mononucleosis
- B27.8, other infectious mononucleosis
- B27.9, infectious mononucleosis, unspecified
Those four codes are not reportable codes. If you try to report B27.0, you will either hit an edit or get the bill back from the insurance company. Why? Because each of those four codes includes four additional subcategories of codes to denote:
- Without complications
- With polyneuropathy
- With meningitis
- With other complication
Not only do we need to know the causative organism for mono, we also need to know what complications are present.
Keep in mind though that we can report a code from the B27.9- series if the physician does not know the causative organism. You always want to report the most specific code, which would be from the B27.0-, B27.1-, or B27.8- series. However, the physician may be unable to determine the type of mono without additional tests. The ICD-10-CM Official Guidelines for Coding and Reporting tell us not to conduct medically unnecessary diagnostic testing in order to determine a more specific code.
Poor Paul, he just wanted to take his black Labrador Molly to the vet for her checkup. Instead he ended up needing a doctor.
Paul put down a sheet in his truck so Molly wouldn’t have to sit on the cold leather seat. To make sure the sheet didn’t slip, he tucked it in. And in the process, he injured the middle finger on his right hand.
Undeterred by the drooping finger, Paul took Molly to the vet, where she got a clean bill of health, then left her with his friend Bruce while he took himself off to Fix ‘Em Up Clinic.
Dr. Neefe examined Paul’s finger and diagnosed mallet finger. Mallet finger is an injury of the extensor digitorum tendon of the finger at the distal interphalangeal joint. Fortunately, the ICD-10-CM Alphabetic Index lists mallet finger with three potential codes:
- Acquired, see Deformity, finger, mallet finger
- Congenital, Q74.0
- Sequelae of rickets, E64.3
We know Paul suffered an injury, which makes it acquired mallet finger. So off we go to deformity. Not surprisingly, we have three code choices for an acquired mallet finger:
- M20.011, mallet finger of right finger(s)
- M20.012, mallet finger of left finger(s)
- M20.019, mallet finger of unspecified finger(s)
ICD-10-CM only requires laterality, not specificity for the individual finger. If Paul had developed mallet finger on more than one finger, we would still report the same code and only report it once.
You will need to check the record for some additional information, because mallet finger can simply be a tendon injury treated with a splint or it can be serious enough to require surgery.
In some cases, the tendon may pull a piece of the bone away when it breaks. If a patient has large fracture fragments or the joint becomes misaligned, the patient may require surgery.
In other cases, the nail may become detached. Fortunately, Paul only needs a splint and Dr. Neefe expects Paul will regain acceptable function and appearance with this treatment plan.
Molly was very happy to see Paul return, even if he has a funny cover on his finger.
In its continuing quest to halt ICD-10 implementation, the AMA is touting a new study by Nachimson Advisors that shows much higher costs to physician practices than initially estimated in 2008.
In addition, the AMA contends in a February 12 letter to HHS Secretary Kathleen Sebelius that ICD-10 will not improve patient care, but instead “could disrupt efforts to transition to new delivery models.”
In the initial 2008 study, Nachimson Advisors estimated that small physician practices would spend a median of $83,000 to implement ICD-10. The amount went up to $285,195 for medium practices and $2.7 million for larger practices.
In the updated 2014 study, Nachimson estimated the transition costs at:
- $56,639‐ $226,105 for small practices
- $213,364‐ $824,735 for medium practices
- $2,017,151‐$8,018,364 for large practices
The biggest potential costs are software upgrades and lost productivity. Those aren’t new concerns. Neither is the cost of ICD-10 implementation.
We’ve known from the beginning that ICD-10 implementation would cost a lot of money and it would slow down productivity. Everything new costs money and slows down work.
Providers had an extra year to implement ICD-10, thanks in no small part to the AMA’s complaints. Smart organizations and practices used that time to spread out their costs, test their systems, and train their coders, billers, front-end staff, physicians, pretty much everyone but the custodial and kitchen staff.
The ones who counted on the AMA forcing another delay—or stopping ICD-10 altogether—didn’t. And in the end, they are the ones who will likely see higher costs. If you can spread out payments, like a mortgage, you can keep the costs more manageable. If you are trying to do everything all at once at the last minute, costs go up. It’s also more difficult to find a big chuck of money in one year’s budget.
The AMA also points out that physicians and coders will be going from approximately 16,000 ICD-9-CM codes to 68,000 ICD-10-CM codes. True, but the AMA fails to acknowledge two things:
1. No coder or physician is going to use all 68,000 codes. Most coders and physicians will use a small subset of codes the same way they do now. Cardiologists aren’t going to be using neurology codes. That cuts down on the number of possible codes.
2. Some part of that increase in codes is due to laterality—you go from one ICD-9-CM code to three or four ICD-10-CM codes for a condition just by adding laterality. Physicians should already be documenting laterality. Coders just need to start looking for it.
What physicians will need to do is document better. That’s not a new concept. It’s good patient care. Isn’t that what physicians want?
The most interesting line in the report comes almost near the end: “A poorly executed ICD‐10 implementation effort will increase those risks [of payment disruptions] and expose practices to large costs in 2014 and beyond.”
The report also states that planning must take place now to mitigate risks and allow practices to continue to operate efficiently. If you are just starting to plan now, you are in big trouble. ICD-10 implementation isn’t going to happen overnight, but regardless of the AMA’s wishes, it is going to happen.
Winter weather forced CMS to cancel Friday’s planned eHealth Summit: Road to ICD-10. Instead CMS will host a 90-minute special event Thursday, February 20, from 12-1:30 p.m. Eastern to discuss CMS readiness for the transition.
If you registered to attend the eHealth Summit, CMS automatically registered you to attend next week’s event.