Some days I swear I have the attention span of a hyperactive hummingbird or Dug the talking dog from the movie “Up.” Maybe what I really have is attention deficit disorder (ADD). How would you code ADD in ICD-10-CM?
If you look up “attention” in the ICD-10-CM Alphabetic Index, you will indeed find an entry for “deficit disorder or syndrome” and code F98.8 (other specified behavioral and emotional disorders with onset usually occurring in childhood and adolescence).
F98.8 is a pretty general code. In the Tabular List, F98.8 does not specify ADD, but does provide these conditions:
- Excessive masturbation
I know a lot of people who bite their nails (everyone I know has given up thumb-sucking and nose-picking though).
I wonder if ICD-10-CM might have a better code for a short attention span. It doesn’t really seem to fit with the other listed conditions.
Back to the Alphabetic Index. Under “attention, deficit disorder,” we find a subentry specifying “with hyperactivity.” For patients with attention deficit hyperactivity disorder (ADHD), we are instructed to “see Disorder, attention-deficit hyperactivity.”
When we get to that listing in the Alphabetic Index, we find out that ICD-10-CM includes codes for several types of ADHD:
- F90.0, attention-deficit hyperactivity disorder, predominantly inattentive type
- F90.1, attention-deficit hyperactivity disorder, predominantly hyperactive type
- F90.2, attention-deficit hyperactivity disorder, combined type
- F90.8, attention-deficit hyperactivity disorder, other type
- F90.9, attention-deficit hyperactivity disorder, unspecified type
In the Tabular List (remember we should never code from the Alphabetic Index alone), we also find two notes.
The Includes note tells us that attention deficit disorder with hyperactivity and attention deficit syndrome with hyperactivity fall under category F90.
The Excludes2 note tells us that the following conditions are not included in F90:
- anxiety disorders (F40.-, F41.-)
- mood [affective] disorders (F30-F39)
- pervasive developmental disorders (F84.-)
- schizophrenia (F20.-)
Because this is an Excludes2 note, we can code both ADHD and these disorders in the same patient as long as the physician documents both conditions.
After doing a little more research on ADHD, I’m pretty sure that’s actually not my problem. Well, at least I…oh, look, a squirrel!
CMS is currently hosting an ICD-10 “Code-a-thon” (a title which instantly brings to mind all of the PBS pledge drives I’ve unwittingly watched). One of the questions that has come up repeatedly is when to use A as the seventh character in an ICD-10-CM code.
I think part of the confusion comes from the quick definition of A: initial encounter. You will use seventh character A for more than the first time the patient sees a physician for treatment of an injury.
According to the ICD-10-CM Official Guidelines for Coding and Reporting, coders should use seventh character A when the patient is receiving active treatment, including:
- Surgical treatment
- ED encounter
- Evaluation and treatment by a new physician
James comes into the ED after breaking his arm. Dr. Bones evaluates James and diagnoses a nondisplaced fracture of the neck of the right radius. Dr. Bones stabilizes the injury and sends James home with instructions to see an orthopedist.
ED coders would report S52.134A (A for the initial encounter).
The following day, James sees Dr. Stetter, an orthopedist. Dr. Setter determines that James does not require surgery on his arm and instead immobilizes it in a cast. Dr. Setter’s coder will report S52.134A, exactly the same code that the ED coder reported.
Instead of thinking A for initial encounter, think of it as A for active treatment. If the patient is being actively treated and is not in the healing phase, use seventh character.
Ebola has been in the news quite a bit recently and it’s actually a good global case study for why we should be using ICD-10-CM codes.
ICD-9-CM does not include a specific code for Ebola. It gets lumped in with other specified viral infections in code 079.89. So in the U.S., we have no idea how many people suffer from Ebola (I think the last confirmed count was three).
In many other countries, they can monitor the rates of Ebola, along with a host of other infectious diseases. The code for Ebola in ICD-10-CM is A98.4, in case you’re wondering.
You may be thinking, who cares? Three cases of Ebola aren’t many. Why do we need to know?
The world is becoming a smaller place every day. Travel is faster and cheaper, and while security will pat you down looking for weapons, nobody checks to see if you are infected with a potentially fatal disease.
Still don’t think it’s important to know what’s making people sick? Harken back to the glory days of 2002 when a new virus know as severe acute respiratory syndrome (SARS) hit the world stage. SARS took only a few weeks to infect people in 37 countries.
The fatality rate for SARS was a little under 10%. Ebola is fatal in more than half of the cases.
More specific codes allow for better public health monitoring and better disease tracking. Without effecting tracking and monitoring, we can’t effectively identify and try to contain an outbreak.
We also can’t tell how well a particular treatment is working or even if it is working at all. And you certainly can’t determine how many people are dying from a disease like Ebola.
Maybe you don’t care about swapping stats with other countries (I’ve heard that from a few people who are not in favor of moving to ICD-10). Personally, I’d like to know what’s out there that could kill me so I know what to avoid.
The truth is often stranger than fiction because fiction has to make sense.
Apparently a panda in China figured out that pregnant pandas get better treatment, so she pretended to be expecting.
Instead of hearing the pitter-patter of baby panda paws, her keepers discovered they were dealing with a master deceiver.
Fake pregnancies do occur in humans, but they are more likely to happen in animals. If we had a patient who was faking a pregnancy, how would we code it?
It depends on our patient’s symptoms.
A fake pregnancy in humans can be an outright lie, which could mean you are dealing with a patient who suffers from a mental disorder. In those cases, the patient may have a factitious disorder (deliberately and consciously acting as if you have an illness when you aren’t sick).
ICD-10-CM includes four code choices for factitious disorder:
- F68.10, factitious disorder, unspecified
- F68.11, factitious disorder with predominantly psychological signs and symptoms
- F68.12, factitious disorder with predominantly physical signs and symptoms
- F68.13, factitious disorder with combined psychological and physical signs and symptoms
Since the panda in question was only displaying pregnancy behaviors, we would code F68.11.
Sometimes, though, a woman may believe she is pregnant when she’s not. False pregnancy, clinically termed pseudocyesis, is the belief that you are expecting a baby when you are not really carrying a child. In ICD-10-CM, we would report F45.8 (other somatoform disorders).
The ICD-10-CM Alphabetic Index includes the term pseudocyesis, but it’s not explicitly listed as one of the included terms under F45.8.
Not every physician will diagnose a false pregnancy as pseudocyesis. Instead the physician may simply document the patient’s symptoms, such as lack of appetite, nausea and vomiting, enlarged breasts, milk production, and interruption of the menstrual period.
The physician may think the patient is pregnant, but orders a pregnancy test to be sure. If the physician orders a blood test, you won’t know the results when the patient leaves the office after the visit. If the physician documents “possible pregnancy” in the outpatient record, we can’t code pregnancy. We have to code the signs and symptoms.
ICD-10-CM includes two possible codes for lack or loss of appetite:
- R63.0, anorexia (loss of appetite)
- F50.8, other eating disorders (psychogenic loss of appetite)
For the nausea and vomiting, we would report R11.2 (nausea and vomiting, unspecified).
Make sure you are not coding the conditions separately, since ICD-10-CM also includes these codes:
- R11.0, nausea
- R11.10, vomiting
- R11.11, vomiting without nausea
- R11.12, projectile vomiting
- R11.13, vomiting of fecal matter
If our patient is an inpatient, we should know at the time of discharge whether she is pregnant. However, if the physician documents “possible pregnancy” for an inpatient, we would report the pregnancy code as if the pregnancy was confirmed.
You should already be used to that distinction because we have the same rule in ICD-9-CM. No coding of possible, probably, suspected, or rule out diagnoses on the outpatient. On the inpatient side, report them as if they were present.
On a happier note, another Chinese panda recently gave birth to triplets. Hopefully, all three will survive.
Labor Day marks the unofficial end of summer, and hopefully, the end of patients with picnic-induced problems at the Fix ‘Em Up Clinic.
Dr. Sunni Daze sees Sam and diagnoses Sam with a torn medial collateral ligament and a torn lateral collateral ligament.
Unfortunately, Dr. Daze didn’t specify which knee in the diagnosis. She did, however, document that she examined his left knee, so we have laterality. We also know she is seeing Sam for the first time for these injuries, so we would report:
- S83.412A, sprain of medial collateral ligament of left knee, initial encounter
- S83.422A, sprain of lateral collateral ligament of left knee, initial encounter
She refers Sam to an orthopedist for further treatment. When Sam sees the orthopedist for the first time, we will report the exact same codes, even the same seventh character. Seventh character A is used when the patient is receiving active treatment, including when the patient sees a new physician.
Our second patient, Jake, fell victim to a vengeful squirrel during his family’s picnic in the park. The squirrel made a move to steal Jake’s potato chip for a snack. When Jake reached for it, the squirrel got a mouthful of Jake’s hand instead of the tasty treat. The squirrel took off, leaving Jake with a bleeding bite wound.
Dr. Daze documents the following:
Patient presents for initial treatment of open bite wound on hand caused by squirrel. Cleaned and irrigated wound. Placed five stiches in left hand. Applied sterile bandages. Prescribed antibiotics. Sent patient home with instructions for keeping the wound clean.
Again, we know laterality and encounter, so we would report S61.452A (open bite of left hand, initial encounter).
If your payer requires external cause codes, we do have one for bitten by squirrel (W53.21). Don’t forget the placeholder and seventh character.
Our final Labor Day casualty is Megan, who was taking in her first baseball game. Unfortunately, she forgot to pay attention at all times when the game is going on. Instead of catching that foul ball, she took it on the chin—literally. She suffered a broken jaw as a result.
As you’ve probably heard, ICD-10-CM fracture codes are full of all kinds of details. For a jaw fracture, we first need to know upper jaw (see fracture, maxilla) or lower jaw (see fracture, mandible). Chin is the lower jaw, so we will head to the mandible fractures, where we find these choices:
- S02.600-, fracture of unspecified part of body of mandible
- S02.609-, fracture of mandible, unspecified
- S02.61-, fracture of condylar process of mandible
- S02.62-, fracture of subcondylar process of mandible
- S02.63-, fracture of coronoid process of mandible
- S02.64-, fracture of ramus of mandible
- S02.65-, fracture of angle of mandible
- S02.66-, fracture of symphysis of mandible
- S02.67-, fracture of alveolus of mandible
- S02.69-, fracture of mandible of other specified site
Notice two things—we don’t need laterality, but we do need a seventh character. In order to find out which seventh characters apply, we need to go back to the top of category S02 (fracture of skull and facial bones). We have some additional seventh character choices for this category:
- A, initial encounter for closed fracture
- B, initial encounter for open fracture
- D, subsequent encounter for fracture with routine healing
- G, subsequent encounter for fracture with delayed healing
- K, subsequent encounter for fracture with nonunion
- S, sequela
In addition to knowing the exact site of the fracture, we’ll need to know whether it’s open or closed.
Remember, too, that you can take the specific site from the radiology report as long as the physician documents that the patient has a fracture of the jaw.
Keep your eye on the ball at all times!
Summer is almost over, and so are the summer road trips. Clark, for one, will be very happy about that.
Things got off to a bad start when Clark’s wife Ellen ran over his foot with the family SUV in their driveway before the trip even started. Fortunately for Clark, Ellen wasn’t driving fast and he only suffered a badly bruised foot.
What do we need to know to code Clark’s foot injury? If you said laterality and encounter, move to the head of the line.
Our choices are:
- S90.30-, contusion of unspecified foot
- S90.31-, contusion of right foot
- S90.32-, contusion of left foot
We need a placeholder X and one of the following seventh characters to complete our code:
- A, initial encounter
- D, subsequent encounter
- S, sequela
The term initial is a little misleading, because we will use A as the seventh character when the patient is receiving active treatment, such as:
- Surgical treatment
- ED encounter
- Treatment by a new physician
So much for the foot injury.
During their weeklong odyssey, Clark and the family stopped at Super Fun World Amusement Park. By the end of the day, Clark was far from amused.
While on the roller coaster, Clark was struck in the head by a low-flying (but not slow-flying) bird. He ended up with a concussion, a bruise, and a really bad headache.
In ICD-10-CM, concussion codes live in subcategory S06.0X-. We have nine codes that specify whether the patient lost consciousness, and if so, for how long. We even have two codes to use if the patient dies.
However, Clark did not even suffer a second of unconsciousness following the assault by bird, so we would report S06.0X0A (concussion without loss of consciousness, initial encounter).
Note that we need a seventh character and, in this case, the placeholder is in the fifth spot. ICD-10-CM helpfully includes the placeholder in the Tabular List so we don’t forget it.
If Clark had suffered an open wound to the head or a skull fracture, we would code that separately, per ICD-10-CM guidelines.
We can also add some external cause codes (just because it’s fun and hey, the mainstream media likes to pick on our external cause codes):
- W61.92XXA, struck by other birds, initial encounter (Clark didn’t get the name of the bird that hit him)
- Y92.831, amusement park as the place of occurrence of the external cause
- Y93.I1, activity, roller coaster riding
- Y99.8, other external cause status (leisure activity)
On Day 4 of the trip, Clark and the family were driving through Death Valley (where summer temps routinely top 120°F) when the SUV decided to take a break. And so did the air conditioning. More problematic was the lack of cell reception.
Rather than wait for a passing Good Samaritan, Clark decided to hike back to civilization for assistance. Sadly, he forgot to stock up on water when he exited the vehicle and by the time he made it to the main road, he had a whopper of a sunburn and a case of heatstroke.
ICD-10-CM includes four codes for sunburn, based on degree:
- L55.0, sunburn of first degree
- L55.1, sunburn of second degree
- L55.2, sunburn of third degree
- L55.9, sunburn, unspecified
We don’t need to know the location of the burn, just the degree. Remember to code to the highest degree.
For heatstroke, we only see one choice—T67.0—but we need two X placeholders and a seventh character.
Clark wrapped up his week of woes with a nasty case of food poisoning after chowing down at a roadside taco stand. In order to code Clark’s food poisoning, we need to know the causative organism. Our choices include:
- A02.0, Salmonella enteritis
- A02.9, Salmonella infection, unspecified
- A05.0, foodborne staphylococcal intoxication
- A05.1, botulism food poisoning
- A05.2, foodborne Clostridium perfringens [Clostridium welchii] intoxication
- A05.3, foodborne Vibrio parahaemolyticus intoxication
- A05.4, foodborne Bacillus cereus intoxication
- A05.5, foodborne Vibrio vulnificus intoxication
- A05.8, other specified bacterial foodborne intoxications
- A05.9, bacterial foodborne intoxication, unspecified
Happy trails and safe travels! (And be careful what you eat.)
The ICD-10 transition has been nothing if not contentious. We’ve had delays mandated by both CMS and Congress, as well as ongoing attempts by the AMA to kill ICD-10 altogether.
Both 3M’s Donna Smith, RHIA, and AHIMA’s Angie Comfort, RHIA, CDIP, CCS, say determining the correct code isn’t a sure thing. Coders aren’t always ending up at the same code.
Why? Well, first of all, the system isn’t live so no one is really coding in it. We’re still doing some guess work.
Second, physician documentation is not where we need it to be even for ICD-9. As a result, some coders may be guessing or choosing an incorrect default code.
Third, not everyone is finding the same information in the record. In many cases physicians already document laterality, Donna says; it’s just that coders might not know where to look for it.
Fourth, we still don’t have a ton of guidance for the grey areas. We have 30 years’ worth of Coding Clinic advice for ICD-9. We have a few issues for ICD-10.
Many organizations are doing some type of dual or double coding. I’m not sure how many are actually checking to make sure coders are coming up with the correct answer. And that’s another problem. How do you decide who got the correct answer?
You need a plan, Donna says. Part of that plan should include identifying the top diagnoses and procedures at your organization. Pull actual cases that include those conditions or procedures and have all of your coders code the record.
Once you’ve done that, compare the results, Angie says. Did you all come up with the same answer? Probably not. Agreement rates are pretty low right now, according to Donna.
So you came up with one code and your coworker came up with a different one. Maybe a third coworker came up with something completely different. Now what?
Sit down and talk about it, both Angie and Donna say. No one knows everything about ICD-10 yet (no one knows everything about ICD-9 either and it’s been around way longer). Try to figure out why you came up with different codes. Did someone miss a piece of information in the documentation? Did someone make an assumption based on his or her knowledge of the physician’s habits? Is the physician’s documentation so vague that everyone was just guessing?
If you can’t come to an agreement among yourselves, ask Coding Clinic. Send the de-identified record to AHA and ask them how to code it. Coding Clinic loves real-life examples, Donna says. So send them in. The more actual documentation they can look at, the better they can answer questions for everyone.
ICD-10-PCS requires coders to possess strong clinical knowledge as well as a solid foundation in anatomy and physiology. Coders need to understand what physicians are actually doing in certain procedures. Without an understanding of what is being done, coders can’t assemble the correct ICD-10-PCS code.
A Billroth II procedure is a surgical procedure in which an anastomosis is created between the stomach and the proximal loop of the jejunum. In case you’re wondering, a Billroth procedure involves removing the pylorus and creating an anastomosis of the proximal stomach directly to the duodenum.
If you know that a Billroth II is also called a gastrojejunostomy, or if your surgeon documents gastrojejunostomy, you’re in luck. ICD-10-PCS includes two listings for gastrojejunostomy in the Alphabetic Index.
Of course, two is not as good as one, but it gives us a place to start. Our two choices according to the Alphabetic Index are:
- Bypass, stomach 0D16
- Drainage, stomach 0D96
The only difference between the first four characters of the code is character three, the root operation. We determine the root operation based on the intent of the procedure. So in a gastrojejunostomy, what is the physician doing? He or she is creating an anastomosis, basically a surgical connection between two structures.
Look at the definitions of Bypass and Drainage. Bypass involves altering the route of passage of the contents of a tubular body part. Drainage involves taking or letting out fluids and/or gases from a body part. Based on the intent of the procedure, Bypass is our root operation.
In ICD-10-PCS, we assign codes based on the body part bypassed from to the body part bypassed to. Confused? Don’t be. The fourth character in the code is where the bypass began, in this case the stomach. The seventh character is where the bypass ends up, in this case, the jejunum.
We also need to know the approach:
- Via a natural or artificial opening endoscopic
We also need to know what type of device was used. Remember, in ICD-10-PCS you only code devices that are left in the patient after the procedure. For an anastomosis, we are connecting one body part to another. The physician probably isn’t placing a device, but read the operative report to be sure.
For an open Billroth II without a device, we would report: 0D160ZA (bypass stomach to jejunum, open approach).
If you aren’t sure what a physician is doing during a common procedure, ask the physician to explain it or make friends with your clinical documentation improvement specialists. They can probably help you out as well. You still have time to figure out what clinical information you don’t know and what procedures you commonly code. Make the most of that time.
It’s a texting world, and more and more we use shorthand in our everyday lives. What about shorthand in a medical record? Can you code from it?
First, let’s look at what kind of shorthand you might see. Physicians may use a +, ↑, or ↓. Those symbols could mean positive (like a pregnancy test), increased level, or decreased level, respectively. The problem is, they aren’t very clear or specific.
Na↑ could mean hypernatremia (elevated sodium) or it could just mean the sodium level has increased. Maybe the patient has a low sodium level and the physician is simply indicating that treatment to raise the level is working. It could mean the sodium level is a little high or significantly elevated. Maybe the physician just documented a slight elevation so he or she remembers to have it rechecked during the patient’s next encounter.
Bottom line: you just don’t know, so you can’t code from it.
So that means it’s time to query the physician, right? Well, it depends. First, make sure you review the entire record and see if the physician documented the information anywhere else more specifically. Maybe BP↑ was just a note on a summary sheet and in the history of present illness, the physician stated, “Patient’s blood pressure elevated.” No query needed. We have the information.
Maybe the physician didn’t document it more completely elsewhere. Is the condition clinically significant?
Hypernatremia is a CC, so it could affect MS-DRG assignment and the patient’s care. You should probably query.
↓BP could also be clinically significant because a low blood pressure could cause other health problems, such as dizziness, weakness, and fainting. Again, probably worth a query.
↓low chloride is probably not worth querying, because it isn’t associated with any adverse health effects.
If you have providers who routinely seem to document using symbols, work with them to eliminate the symbols and document in clear words. (because we can’t code from symbols in any coding system). The better the physician documents, the better the story of the patient becomes.
Take it from today’s victims, er, patients, at the Fix ‘Em Up Clinic: not every idea is a good idea.
Jeff was so determined to claim the s’mores title that he ate a few marshmallows that were a little too hot. As in, they were on fire. And while fire eating is fine for professionals, for a kid at camp, it’s not such a great idea.
Dr. Sunni Daze examines Jeff and documents burns to the mouth, pharynx, tongue, and lips.
The burns of the mouth, pharynx, and tongue are easy. One code covers all three and it does not specify degree of the burn. Since this is Jeff’s initial visit, we would report T28.5XXA.
The lip burns require a little more information. We need to know what degree of burns Jeff suffered on his lips. Fortunately for him, Dr. Daze notes the burns are first degree, so we would report T20.12XA (burn of first degree of lip[s]).
ICD-10-CM does not include separate codes for the upper and lower lip, so T20.12XA covers one lip or both.
We also find the following note under pretty much all of the burn codes:
- Use additional external cause code to identify the source, place and intent of the burn (X00-X19, X75-X77,X96-X98, Y92)
We definitely need an X00-X19 code, which in Jeff’s case is X10.1XXA (contact with hot food, initial encounter).
The X75-X77 codes are for intentional self-harm. Overeating burning marshmallows doesn’t quite qualify as planning to hurt yourself. Jeff just got caught up in the moment.
The X96-X98 are codes for assault. Again, not applicable in Jeff’s case.
For our place of occurrence, we’ll use Y92.833 (campsite as the place of occurrence of the external cause). Notice we do not need a seventh character for this code.
Our second patient Grace tried to change the direction of the fan by grabbing the cage that houses the blades. Turns out that wasn’t the best plan. That “squeeze” caused the blades to hit the cage and Grace’s fingers.
Fortunately, the fan wasn’t going fast enough to sever any of Grace’s fingers, but she does have some wicked-looking lacerations.
ICD-10-CM finger laceration codes are very specific, as in you need to know which hand and which finger is involved. You also need to know whether the nail is damaged and whether any foreign body remains in the laceration.
Let’s see what Dr. Daze documented:
14-year-old female with lacerations to the palm side of the right index, middle, ring fingers, no nail involvement. No foreign bodies remain in wounds. Flexor digitorum profundus tendon of index finger severed.
The first injury we’ll code is the severed tendon, because it is the most severe injury. If you look under laceration (because the tendon was lacerated, right?), the entry for tendon directs you to see Injury, muscle, by site, laceration.
Okay, so off to Injury, muscle, finger we go. And find all sorts of code options. The codes are divided into the forearm and hand level. We don’t have a code for finger level, so we’ll go with hand level.
However, we can code to the specific finger. In Grace’s case, that’s the right index finger and code S66.120A (laceration of flexor muscle, fascia, and tendon of right index finger at wrist and hand level, initial encounter).
The other codes for Grace’s misadventure would be laceration codes for the middle and ring fingers:
- S61.212A, laceration without foreign body of right middle finger without damage to nail, initial encounter
- S61.214A, laceration without foreign body of right ring finger without damage to nail, initial encounter
Dr. Daze is done for the day and so are we. Remember to make sure your food isn’t on fire before you eat it.