Michelle A. Leppert, CPC, is a senior managing editor specializing in coding for JustCoding.com, which provides coders, coding supervisors, and health information management (HIM) directors with educational resources to test their coding knowledge, employ correct coding guidelines, and stay abreast of CMS transmittals.
In addition, she writes and edits the HCPro Inc. publications, Briefings on Coding Compliance Strategies. Email her at firstname.lastname@example.org.
Everybody likes a shortcut. We want to get things done faster, arrive home sooner, finish tasks more quickly. Generally shortcuts are good. Unless you’re talking about physician documentation. Then shortcuts are very bad.
One of the advantages to ICD-10 coding is the increased specificity of the codes. However, without complete, accurate documentation, more specific codes do us no good. We don’t need physicians to write a dissertation, but we do need them to give us some specific information.
We know physicians don’t document well now. We know we often default to unspecified codes. How many times do you report 250.00 for a diabetic patient or 715.90 for osteoarthritis (that’s not specified as generalized or localized, site unspecified)? Neither one of those ICD-9-CM codes tells us (or the physician) much about the patient’s condition.
So how do we defeat the Second Horseman of the ICD-10 Apocalypse? First, look at the current state of your physician documentation. You can’t fix it if you don’t know it’s broken (or in this case deficient). What’s missing in ICD-9? What will be missing in ICD-10?
Next, talk to your physicians. Show them their documentation, then show them the codes. Show them where their actual documentation is lacking. Don’t use generic examples. Use their actual documentation. Some of your physicians will be better than others.
Most physicians are not lining up to learn how to code in ICD-10 and that’s okay. We don’t want to teach them how to code (feel free to tell them that). We just want them to provide a clear picture of the patient’s illness or injury, their thought processes, and what they did to make the patient better. Really, they should already be doing that.
Remind them that good documentation does affect patient care. If we didn’t need clinical specificity, physicians could just say, the patient is sick. How sick is the patient? What specifically is wrong with the patient? And how can you treat the patient if you don’t know what’s wrong?
If physicians document accurately and completely, they will better represent the patient’s severity of illness and risk of mortality (and they can show that their patients are actually sicker than everyone else’s). They can better track the patient’s disease process. Is the patient better or worse? Has the disease progressed or is the treatment working?
It will also make for better communication between physicians treating the patient. If Dr. Jones documents that Jane suffers from monoplegia after a stroke, Dr. Smith doesn’t really know which limb is affected. If Jane comes in and is unable to move her right arm, Dr. Smith may think that’s a late effect from the stroke.
However, if Dr. Jones documents that Jane suffers from monoplegia of a lower limb following nontraumatic intracerebral hemorrhage affecting the right dominant side (I69.141), Dr. Smith knows immediately that a paralyzed right arm is a new problem.
You can also point out that better documentation leads to fewer queries from clinical documentation specialists and coders. That frees the physicians to spend more time treating patients, which is why most of them went to medical school in the first place.
Ah, Black Friday, when shoppers go doorbusting for deals and then head to the Fix ‘Em Up Clinic for repairs.
Today’s first wounded bargain hunter, Stephanie, comes in complaining of pain in her left elbow. She apparently elbowed another shopper out of the way to grab the last sweater on the rack. She is also experiencing difficulty when trying to bend her elbow and is wheezing when breathing.
Dr. Donner performs an initial evaluation and then sends Stephanie off for x-rays of her elbow. Dr. Prancer’s x-ray report states that Stephanie suffered a posterior subluxation of the left ulnohumeral joint with no broken bones.
Dr. Donner notes the findings in her notes and also notes that Stephanie has a history of asthma. Dr. Donner documents that Stephanie’s wheezing is due to mild, intermittent asthma without complications.
That gives us the following codes:
- S53.122A, posterior subluxation of left ulnohumeral joint, initial encounter
- J45.20, mild intermittent asthma, uncomplicated
Our next patient, Jane, comes in with a bleeding hand. Apparently Jane snagged the last iPad and another shopper wanted it enough to bite Jane’s hand. Ouch.
Dr. Donner cleans and dresses the wound and documents an open bite of the right hand. She also notes that the bite was a deliberate assault which took place in a shopping mall. That gives us codes:
- S61.451A, open bite of right hand
- Y04.1XXA, assault by human bite, initial encounter
- Y92.59, other trade areas as the place of occurrence of the external cause (the mall)
- Y99.8, other external cause status
If Jane sustained simply a superficial bite, we would have reported code S60.571A (ather superficial bite of hand of right hand, initial encounter).
Our final patient of the day came in for some non-shopping, but still holiday-related, setbacks. Ralph, a 23-year-old, apparently enjoyed Thanksgiving dinner at Mom’s house so much he had seconds, and thirds, and fourths. He now is suffering from a severe stomach ache and feels bloated. He complains of nausea but denies vomiting. He also complains of chest pain and has no history of heart problems.
Dr. Donner orders an EKG to rule out a heart problem as the cause of the chest pain. Dr. Donner rules out an ulcer and cancer as causes of Ralph’s pain. She also rules out a hernia. Dr. Dasher’s EKG report shows no signs of heart problems.
Dr. Donner diagnoses Ralph with functional dyspepsia and advises him to moderate his food intake in the future. That gives us code K30 (functional dyspepsia).
I think it’s time to grab some leftovers and start my Cyber Monday shopping. Happy Thanksgiving!
Farmer Brown came in today to see Dr. Gobbler for some injuries sustained when he tried to prepare his Thanksgiving main course. It seems Farmer Brown’s turkey wasn’t interested in joining him as dinner.
Farmer Brown first attempted to wring the turkey’s neck. After a considerable chase around the yard (and multiple trips and falls by Farmer Brown), he eventually captured the reluctant bird. Unfortunately for Farmer Brown, the turkey pecked his hands until he let go.
That gives us our first code for this encounter: puncture wound of the hand. The question is, which hand? In this case, it’s both hands, so we would report two codes:
- S61.431A, puncture wound without foreign body of right hand, initial encounter
- S61.432A, puncture wound without foreign body of left hand, initial encounter
If the turkey had left part of his beak behind, we would have used the code for puncture wound with foreign body (S61.441A or S61.442A) for the appropriate hand. Notice also that we need a seventh character to denote the encounter.
We can also add some external causes codes to round out our story:
- W61.43XA, pecked by turkey, initial encounter
- Y92.73, farm field as the place of occurrence of the external cause
- Y93.K9, activity, other involving animal care
- Y99.8, other external cause status
Farmer Brown’s attempts don’t stop with interrupted wringing. Next, he tried an axe to cut off the turkey’s head. A chase again ensued, with the turkey neatly tripping Farmer Brown, so he dropped the axe. Unfortunately for Farmer Brown, he dropped the axe onto his foot and crushed two toes.
We need to know which toes and which foot was involved. We scan Dr. Gobbler’s notes and find that Farmer Brown crushed two lesser toes on his left foot. That gives us code S97.122A (crushing injury of left lesser toe[s], initial encounter). Regardless of how many lesser toes (one to four) Farmer Brown crushed, we only report one code.
If Farmer Brown had crushed his great toe and one lesser toe, we would need two codes. The great toe has its own code (S97.112A, crushing injury of left great toe, initial encounter).
Farmer Brown’s third and (thankfully) final attempt to bag the bird involved a really big kettle of boiling water. The turkey decided to enlist the aid of a few friendly cows (who never liked Farmer Brown anyway) to tip over the cauldron and drench Farmer Brown.
Fortunately, he suffered only minor first-degree burns to his right leg. We do need some additional specificity in order to code this injury, namely the location of the burns. The following sites on the leg can be coded specifically:
- Lower leg
- Multiple sites
Dr. Gobbler documents burns to the right lower leg, so we would code T24.131A (burn of first degree of right lower leg, initial encounter).
We also spot the note under code T24.1 directing us to report additional external cause codes to identify the source, place, and intent of the burn (X00-X19, X75-X77,X96-X98, Y92), so we also report:
- X12.XXXA, contact with other hot fluids (boiling water), initial encounter
- X98.2XXA, assault by hot fluids, initial encounter (We could make a case that the turkey was only defending himself when he coerced the cows into tipping over the cauldron or that it was an accident.)
- Y92.73, farm field as the place of occurrence of the external cause
Following his latest setback, Farmer Brown has decided to let the turkey live and take his wife out for dinner instead. Let someone else deal with uncooperative poultry.
You may remember that the American Medical Association (AMA) has been pushing for an end to ICD-10 since 2012. During its recent House of Delegates meeting, the AMA reinforced its position that ICD-10 implementation should be delayed by two years. It initially put forth that resolution in June.
The AMA already helped force CMS to delay implementation of ICD-10 from October 1, 2013, to October 1, 2014. That may not have been a great thing for physicians, according to Paul Weygandt, MD, JD, MPH, MBA, CCS, vice president of physician services for J.A. Thomas and Associates in Atlanta.
“The worst thing for physicians was that the AMA delayed ICD-10 by one year,” he told AHIMA Convention attendees. That gave physicians the idea that the AMA will stop ICD-10 implementation again.
How do you get physicians on board for ICD-10 when the AMA is not? Remind them that ICD-10 doesn’t change the way they practice medicine. They will still treat patients the same way they do now. We’re just asking them to document a little more.
Physicians are likely documenting much of the necessary information already, such as laterality, because it’s good patient care. The physician wants to know where an injury occurred so when the patient comes back for a follow up, he or she is checking the correct area.
ICD-10 is also written in more clinical terms and less coder speak, which means docs will need to learn less than coders. For example, many pulmonologists already describe asthma as:
- Mild intermittent
- Mild persistent
- Moderate persistent
- Severe persistent
ICD-10-CM now uses those terms.
For myocardial infarctions, physicians have been documenting STEMI and non-STEMI for years, Weygandt says. In ICD-10-CM, coders will be able to report it that way.
Don’t tell physicians what they need to document. Tell them what they aren’t documenting. Give them a (figurative) pat on the head for the things they are doing correctly. And ask them if they would accept their documentation if it came from a resident.
“Good documentation for ICD-10 is what we should be teaching residents because it’s good clinical care,” Weygandt says.
ICD-10 is coming, whether the AMA wants it to or not. Work with your physicians now so you are all ready for the change.
Conquest, War, Famine, and Death have nothing on the Four Horsemen of the ICD-10 Apocalypse:
How do you tame these terrors? With planning and practice (and some luck thrown in as well). Over the next few weeks, we’ll look at each of these areas in more detail. Today, let’s start with accuracy.
By now, most coders should have started ICD-10 training of some sort. Coding managers should know how well their coders know anatomy and physiology and medical terminology. Ideally, coders should also have started training on the code sets. Some organizations are done training and are onto the practice stage, while others are just compiling the results of their coder assessments.
Dual coding has been a hot topic in HIM circles for at least a year now. People are trying to figure how and when to start dual coding, which method to use, and how many records to code. Some people think that dual coding or coder training and practice is all they need to get their coders ready for ICD-10.
Not quite. It’s not enough to be able to pick a code in ICD-10. You have to assign the correct code. How do you know if you are assigning the correct code? That’s where things can get tricky.
If you are coding actual records from your facility, you don’t have an answer key to look at to know you came up with the right answer. If you are the only one at your facility coding a particular record in ICD-10, you don’t even have anyone to ask.
Rachel Chebeleu, MBA, RHIA, and her colleagues at the Hospital of the University of Pennsylvania (HOP) came up with a solution. A group of superusers at HOP took a group of actual hospital records and each person coded each record in ICD-10. Then they compared notes to see whether they all arrived at the same codes. If they did, great. If they didn’t, they discussed the case and why each person came up with a specific code.
Consider creating a coding roundtable at your facility to make sure you not only practice coding in ICD-10, but you end up with the correct codes.
One horseman down, keep an eye out for Documentation.
A feral flock of wild turkeys has invaded New York City. Seriously. And with them, they bring all sorts ofcode-ready diseases and mishaps.
When we look up chlamydiosis in the ICD-10-CM Alphabetic Index, we are directed to see chlamydia. That doesn’t sound good.
On a clinical note, however, chlamydiosis in birds is different from the human venereal disease chlamydia. Patients who contract chlamydiosis from birds often experience fever, headache, and loss of appetite. They may also experience painful or difficult breathing.
Chlamydiosis in birds, such as our Big Apple party crashers, is caused by a bacterial organism, Chlamydophila psittaci. And it just so happens we have a specific ICD-10-CM code for it: A70 (Chlamydia psittaci infections).
We all know not to eat raw eggs and I certainly don’t want to fight a wild turkey for one (the grocery store ones come with much less hazard to my hands). Odds are, we won’t contract salmonellosis from the NYC flock.
Colibacillosis is caused by our old friend Escherichia coli. E. coli can cause all sorts of unpleasant conditions, including:
- A04.0, enteropathogenic Escherichia coli infection
- A04.1, enterotoxigenic Escherichia coli infection
- A04.2, enteroinvasive Escherichia coli infection
- A04.3, enterohemorrhagic Escherichia coli infection
- G00.8, meningitis due to Escherichia coli
- J15.5, pneumonia due to Escherichia coli
- P36.4, sepsis of newborn due to Escherichia coli
Those all sound like awesome reasons to avoid the walking turkeys and their droppings.
But what happens when a turkey tries to cross Broadway? Well, if the bird is unlucky, it will get run over. New Yorkers stop for no fowl.
However, if the birds are smart enough to cross en masse, they could cause some serious traffic disruptions and possibly some traffic accidents. We’ll leave the specific injuries (contusions, lacerations, broken bones, etc.) out of the equation for the moment and look at the External Causes codes associated with traffic accidents.
ICD-10-CM includes multiple code possibilities for transport accidents in sections V00-V99. The introductory note for the section states:
This section is structured in 12 groups. Those relating to land transport accidents (V01-V89) reflect the victim’s mode of transport and are subdivided to identify the victim’s ‘counterpart’ or the type of event. The vehicle of which the injured person is an occupant is identified in the first two characters since it is seen as the most important factor to identify for prevention purposes. A transport accident is one in which the vehicle involved must be moving or running or in use for transport purposes at the time of the accident.
If the turkeys on the Great White Way caused an accident, we would likely head to V40 (car occupant injured in collision with pedestrian or animal). Our choices include:
- V40.0, car driver injured in collision with pedestrian or animal in nontraffic accident
- V40.1, car passenger injured in collision with pedestrian or animal in nontraffic accident
- V40.2, person on outside of car injured in collision with pedestrian or animal in nontraffic accident
- V40.3, unspecified car occupant injured in collision with pedestrian or animal in nontraffic accident
- V40.4, person boarding or alighting a car injured in collision with pedestrian or animal
- V40.5, car driver injured in collision with pedestrian or animal in traffic accident
- V40.6, car passenger injured in collision with pedestrian or animal in traffic accident
- V40.7, person on outside of car injured in collision with pedestrian or animal in traffic accident
- V40.9, unspecified car occupant injured in collision with pedestrian or animal in traffic accident
All of these codes require a seventh character to indicate the encounter type, so you’ll also need two placeholder Xs so the seventh character ends up in the seventh spot.
So if you’re flocking to New York for the holidays, watch out for those wild birds.
CMS created a mini tempest in June when announced it would not conduct end-to-end testing for ICD-10. CMS has not reversed that decision, but will require MAC to conduct front-end testing in March 2014.
This front-end testing is just that—testing between MACs and trading partners, such as healthcare organizations. It is not a full cycle test from claim submittal to remittance advice, denials, and refund requests.
The testing will run March 3-7 and will include live help desk support from 9 a.m. to 4 p.m. in the contractor’s time zone, according to MLN MattersMM8465.
MACs will set up a registration site or an email address and publicize it at least four weeks in advance (so no later than the beginning of February).
Providers and suppliers participating during the testing week will receive electronic acknowledgement confirming that the submitted test claims were accepted or rejected.
Healthcare organizations should take advantage of the test week because it will probably be their only opportunity to test with MACs.
If you haven’t already done so, look into doing actual end-to-end testing with your other vendors and payers.
The majority of respondents in our recent (unscientific) JustCoding poll identified physician documentation as their biggest concern heading into the ICD-10 transition.
No one should be surprised by that because we’ve been worried about (the lack of good) physician documentation for years. Our friends at the Association for Clinical Documentation Improvement Specialists (ACDIS) dedicate their professional lives to improving physician documentation.
Many on the HIM side are also worried about training physicians about ICD-10 and documentation. How do you get them to listen (and care)?
Here are some quick pointers I picked up from various speakers at the AHIMA Convention last month in Atlanta:
- Physicians really want to hear from other physicians. And it helps if that physician is in the same specialty. Neurologists aren’t interested in hearing from plastic surgeons and cardiologists don’t care what orthopedists have to say.
- Keep the information relevant to the physician’s specialty. Don’t talk to pulmonologists about gastrointestinal procedures.
- Tell physicians specifically what they aren’t documenting. If they are already documenting laterality, don’t mention it during discussions about documentation improvement. If you spend time telling them they need to document something that is already in their notes, they may miss when you point out what isn’t in the documentation.
- Give them concrete examples from their documentation. This takes a little bit of work, but it could pay off big time. Take the physician’s note, open the ICD-10 manual, and physically show them what information is in the documentation and what information the coders need.
- Remind the physicians that ICD-10 is not going to change the way they treat patients. It’s not about the practice of medicine. We know they provide great care, we just want them to document that care.
- Ask the physician if his or her documentation would pass muster in medical school. Would you accept this note from a resident?
- Last, but probably most important—tell them why they should care. And don’t talk about money. They don’t care. They’re still getting paid based on CPT® codes for their services. Besides, financial gain should never be the main goal of documentation improvement. However, their documentation will affect their quality scores and their profile. So it really is in their best interests to clearly document how sick the patient is, what the physician was thinking, and how the physician treated the patient.
Have any other tips or ideas? Post them in the comments below.
Mr. Jack O. Lantern underwent some significant surgery at Stitch ‘Em Hospital back on October 16. Dr. Carver removed Jack’s liver, stomach, large intestine, small intestine, appendix, and gall bladder. Unfortunately, Jack is suffering some complications from his surgery.
Dr. Carver first diagnoses Jack with a fungal infection of the skin due to Podosphaera xanthii. When we look up infection in the ICD-10-CM Alphabetic Index, we find quite a list of potential codes. We know Jack has a skin infection, so if we go to Infection, Skin, we find an entry for “due to fungus.” We actually have two choices:
- B36.8, other specified superficial mycoses
- B36.9, superficial mycosis, unspecified
Since we know the causative organism, we would use B36.8. Remember that other specified is not the same as unspecified. We know the cause, we just don’t have a code for it. Some antibiotics should clear that right up.
Sadly, that is not the extent of Jack’s woes. Dr. Carver also diagnoses him with gangrenous cellulitis, an infection of soft tissue that produces extensive tissue necrosis and local vascular occlusions. Basically, Jack’s rotting from the inside.
If we look up cellulitis, gangrene in the ICD-10-CM Alphabetic Index, we are directed to Gangrene. However, gangrene does not list cellulitis as one of the sub-terms. The main code for gangrene is I96 (gangrene, not elsewhere classified). When we look that up in the Tabular List, we find gangrenous cellulitis listed under I96.
Dr. Carver also documents gangrene of the abdominal wall, which also maps to code I96.
Dr. Carver determines that Jack needs some excisional debridement to remove the gangrene so he can return to health.
In ICD-9-CM procedure coding, we would code excisional debridement using 86.22. How would we code it in ICD-10-PCS?
In order to get to the correct ICD-10-PCS table, we need to know the section, body system, and root operation. Section is easy. We know we are coding for a surgical procedure, which gives us an initial character of 0 (zero, not a capital O).
Let’s skip the body system for a minute and figure out which root operation we’re going to use. When you look up debridement in the ICD-10-PCS Alphabetic Index, you find two choices:
- Excisional, see Excision
- Non-excisional, see Extraction
We need to do a little investigating in Dr. Carver’s operative report to find out exactly what she did, so we know whether we’re coding excisional or non-excisional debridement.
Excisional debridement involves removing or cutting away devitalized tissue, necrosis, or slough. It is always a surgical procedure. Physicians use a recognized sharp instrument, such as a cutting curette, laser, scissors, or scalpel, to perform the excisional debridement.
Dr. Carver documents removal of devitalized tissue using a scalpel, so we are coding an excisional debridement.
On to the body system. Where specifically is Jack’s gangrene? Hopefully, Dr. Carver is very specific in her documentation, otherwise we’ll need to query.
We know some of the gangrene is located in the abdominal wall. ICD-10-PCS classifies the abdominal wall as a general anatomical region, which leads us to table 0WB.
Abdominal wall gives us a fourth character of F.
Now we need the approach. We have three choices:
- Open (0)
- Percutaneous (3)
- Percutaneous endoscopic (4)
Because of the extent of Jack’s abdominal gangrene, Dr. Carver elects an open approach. We have no device choice, which makes the sixth character Z. Our final character can be either X (diagnostic) or Z (no qualifier).
For Dr. Carver’s excisional debridement of Jack’s abdominal gangrene, our code is 0WBF0ZZ.
What about the rest of the debridements? Again, we need to know where exactly Dr. Carver is performing the debridements, as well as how deep the debridements are.
ICD-10-PCS guidelines state that an excisional debridement that includes skin, subcutaneous tissue, and muscle is coded to the muscle body part.
Dr. Carver documents the following debridements:
- 13 sq cm of the left quadriceps, percutaneous
- 9 sq cm of the right flexor pollicis longus and pronator quadratus, percutaneous
- 10 sq cm of subcutaneous tissue in the right lower leg, percutaneous
- 8 sq cm of dermis on the scalp, external approach
For each debridement, we need a specific code. Each one will vary based on the body system, the body part, and the approach. All of the codes will include the same section, root operation, device, and qualifier.
If you don’t know where the flexor pollicis longus and pronator quadratus muscles are, check the body part guide in the back of your ICD-10-PCS Manual (For the record, they are arm muscles).
Our codes would be:
- 0KBR3ZZ, excision of left upper leg muscle, percutaneous approach
- 0KB93ZZ, excision of right lower arm and wrist muscle, percutaneous approach
- 0JBN3ZZ, excision of right lower leg subcutaneous tissue and fascia, percutaneous approach
- 0HB0XZZ, excision of scalp skin, external approach
Hopefully, Dr. Carver excised all of the dead tissue and Jack will soon be on the mend.
Our sister publication, Medical Records Briefing, is conducting a benchmarking survey on ICD-10 implementation, and we would appreciate your input. Please take a few moments to complete this survey.
Simply click here to answer the survey questions online. You should only need 5-10 minutes to answer the survey.
You know you have been bombarded with surveys about ICD-10, so we thank you for input!