We want physicians to be very clear in their documentation about what’s wrong with the patient and what the physician did to make that patient better. CDI Specialists spend their professional lives working to get more precise, detailed, accurate, no-room-for-interpretation documentation.
Why all the fuss about specificity? Different people interpret things different ways. My colleagues sometimes hate me for playing devil’s advocate and saying, someone could take that to mean Y instead of X.
Relevant case in point: Earlier this week, AHIMA emailed its members imploring them to call members of Congress to oppose an ICD-10 delay. AHIMA’s Margarita Valdez, senior director of Congressional relations, heard that Chairman of the House Rules Committee Pete Sessions, R-Texas, was drafting (or looking to draft) language for another ICD-10 delay.
My colleague Steve Andrews reached out to Sessions’ office asking for verification. In very short order, a member of Sessions’ staff replied that the Congressman is meeting with physicians about their concerns regarding ICD-10, but no legislation has been drafted.
I also heard from a blog reader who told me she had spoken to a staff member for Rep. Rob Woodall, R-Ga. That staff member had spoken to Sessions’ staff, who told him that Sessions has no plans at present to introduce legislation for another delay of ICD-10. (That’s a massive amount of hearsay, because it’s at best fourth-hand information, but still, we’ll take what we can get for now.)
At first glance, that looks like a win. No legislation, no delay, right? Not exactly. Here’s where playing devil’s advocate comes in.
Sessions had not drafted legislation and has no plans to introduce legislation. However, no one claimed Sessions isn’t looking to add language to an existing (or future) bill to delay ICD-10. It’s nitpicking, but we are talking about politicians.
The whole discussion may have started with a letter that the AMA and 99 specialty societies sent to Andrew Slavitt, acting administrator of CMS, detailing a “number of concerns that do not appear to be addressed” by CMS’ current transition plan.
Fair enough, we’re all worried about certain parts of the transition.
The medical groups want CMS to release more detailed end-to-end testing results broken out by:
- Type and size of providers who tested
- Number of claims tested by each submitter, percentage of claims successfully processed
- Specific details about problems encountered
That seems reasonable to me. The more we know about what worked and what didn’t, the better we can prepare.
Things get a little murkier near the end of the letter when it talks about code specificity. The AMA has been beating this drum for a while now because the increased specificity requires more detailed documentation. The detail, though, is one of the main strengths of ICD-10. We can get a better picture of the patient’s health, and with ICD-10-PCS, we’ll actually know what specific procedure the physician performed.
According to the letter:
CMS officials have stated that, absent indications of potential fraud or intent to purposefully bill incorrectly, CMS will not instruct its contractors to audit claims to verify that the most appropriate ICD-10 code was used. There is also general concern about how physicians will be audited as they learn to use the new code set. We urge CMS to: 1) confirm and broadly educate stakeholders and contractors that claims will not be audited simply for code specificity; and 2) to instruct contractors that they are prohibited from engaging in audits that are only predicated on code specificity.
I was kind of appalled by the apparent unconcern for specificity. Then I reread it and found it’s not so much appalling as it is vague. Then I went back to the source of the “no audit” statement, the Government Accountability Report on CMS’ preparation for ICD-10 implementation. According to that report:
CMS officials stated that the submission of valid ICD-10 codes is a requirement for payment; however, when the presence of a specific diagnosis code is not required for payment then the claim would be paid even if a more appropriate ICD-10 code should have been used on the claim.
In a way that makes sense. You could go to the physician’s office for a cold or a bad cut or just for your annual physical. In order to use the CPT® code for an office visit, you don’t need a particular ICD-10-CM code. You just need the code that will support medical necessity for that service. So if a patient is seen for S50.311A (abrasion of right elbow, initial encounter), you probably won’t get away with reporting CPT code 99215 (level 5 established patient visit).
The GAO report further stated:
Additionally, CMS officials indicated that, absent indications of potential fraud or intent to purposefully bill incorrectly, CMS will not instruct its contractors to audit claims specifically to verify that the most appropriate ICD-10 code was used. However, audits will continue to occur and could identify ICD-10 codes included erroneously on claims which could lead to claims denials, according to CMS officials.
Just when we were starting to feel really good about ICD-10’s chances of being implemented, AHIMA has learned that Chairman of the House Rules Committee Pete Sessions, R-Texas, is looking to draft language to delay ICD-10. The chairman is seeking support from his colleagues in Congress, according to an email from Margarita Valdez, senior director of Congressional relations for AHIMA.
AHIMA is urging HIM professionals to call Rep. Michael Burgess, R-Texas, at (202) 225-7772 and state that you support ICD-10 implementation in 2015.
Here are some talking points from AHIMA:
- We need the code sets in 2015!
- A recent GAO report supports ICD-10 readiness.
- Small physician practices are expected to spend between $1,900 and $6,000 to transition to the new code set. This is much lower than previous reports. The study can be found on www.coalitionforICD10.org.
You can also mention the House Energy and Commerce Committee’s subcommittee on health’s ICD-10 Implementation hearing February 11. During the testimony, the subcommittee’s members heard from seven industry stakeholders, six of whom supported ICD-10 implementation in 2015. Many members of the subcommittee agreed that the U.S. needs ICD-10, and it should not be delayed any further.
Then call the other congressional leaders also. AHIMA helpfully provided their names and phone numbers:
- Sessions: (202) 225-2231
- Rep. Virginia Foxx (R-North Carolina): (202) 225-2071
- Rep. Tom Cole (R-Okla.): (202) 225-6165
- Rep. Rob Woodall (R-Ga): (202) 225-4272
- Rep. Steve Stivers (R-Ohio): (202) 225-2015
- Rep. Doug Collins (R-Ga): (202) 225-9893
- Rep. Louise Slaughter (D-NY): (202) 225-3615
If you tweet, you can also find their Twitter handles very easily online, for example, @michaelcburgess and @PeteSessions.
Make sure Congress knows we support ICD-10 in 2015!
CMS declared its end-to-end testing week from January 26 through February 3 a success. A total of 661 volunteers submitted 14,929 test claims, with CMS accepting 81% (12,149 claims).
CMS rejected 13% of the claims for reasons not related to ICD-10, such as:
- Incorrect National Provider Identifier
- Dates of service outside the range valid for testing
- Invalid HCPCS codes
- Invalid place of service
Three percent of claims contained an invalid ICD-10-CM and 3% had an invalid ICD-10-PCS codes, leading CMS to reject the claims.
More than half (56%) of the claims fell under professional services and 38% were institutional claims. Suppliers submitted 6% of the claims.
CMS identified zero issues related to professional and supplier claims and stated that none of the claims were rejected because of front-end submission problems.
CMS did find one system issue related to institutional claims, but it affected fewer than 10 total claims. It will fix the problem before the next end-to-end testing week April 26-May 1.
Editor’s Note: This article originally published on the ICD-10 Trainer Blog.
The differences between the ICD-10 and ICD-9 code sets are primarily in the overall numbers of codes, their organization and structure, code composition, and level of detail. ICD-10-CM contains about 70,000 diagnosis codes compared to approximately 14,000 in ICD-9-CM, and approximately 72,000 in ICD-10-PCS (procedure) codes compared to 4,000 ICD-9-CM codes. ICD-10 codes are also longer and use more alphanumeric characters.
Punctuation within both the tabular and alphabetic indices still plays an important role in code interpretation and selection. When coders need to find their way around the codes, they can use either the alphabetic or tabular index. The biggest variation is that the ICD-10 tabular index uses more tables to drive digit selection. While very similar to ICD-9, CDI specialists and coders should review the ICD-10 indices to ensure accurate code selection.
Each ICD-10-CM code is three to seven characters long. The first is an alpha character, the second is numeric, and the third through seventh can be either alpha or numeric, with a decimal after the third character. Alpha characters are not case-sensitive. ICD-10-CM uses a placeholder character “X,” as the fifth character to allow for future code expansion in a particular area.
In ICD-9-CM it takes multiple codes to fully describe the patient’s clinical picture. Take, for example, diverticulitis of the large intestine with perforation or peritonitis with bleeding. Under ICD-9-CM coding conventions, there are two codes to assign, one for the diverticulitis with hemorrhage/bleeding and one for the peritonitis. Under ICD-10-CM conventions, there is a combination code to capture this clinical event, K57.41.
Pressure ulcers are another category of codes that this difference can be seen in that ICD-10-CM provides a more precise and descriptive code through combining the site and stage of the ulcer all in one code category, L89. For example, a Stage III pressure ulcer of the heel codes to L89.603.
The ICD-10-PCS codes are used for hospital claims for inpatient procedures. These codes differ from the ICD-9-CM procedure codes in that they have seven characters that can be either alpha (non-case sensitive) or numeric. The numbers 0-9 are used. Letters O and I are not used to avoid confusion with the numbers zero and one. The codes do not contain decimals.
Because ICD-10-PCS boasts more than 155,000 possible code combinations (ICD-9 only has 17,000), the increased number of codes allows ICD-10-CM/PCS to be far more specific than its predecessor. So, to report the new codes, the medical record documentation needs to reflect not only the procedure performed, but also how the physician performed it.
For example, the fifth character identifies the surgical approach for the procedure. (See Figure 2.1.) This forces the coder to consider how the procedure was performed at a level of detail previously unnecessary. And, more importantly for CDI staff, physicians previously did not necessarily need to document that level of specificity either.
Editor’s Note: This excerpt comes from The Clinical Documentation Improvement Specialist’s Handbook, Second Edition by Marion Kruse, MBA, RN and Heather Taillon, RHIA, CCDS.
Implementation costs and plans took center stage at The House Energy and Commerce Committee’s subcommittee on health’s ICD-10 Implementation hearing yesterday. Seven witnesses testified before the subcommittee about the benefits of ICD-10 implementation. Members of the committee then asked witnesses a wide range of questions. Six of the witnesses were pro-ICD-10. The lone voice against came from William Jefferson Terry, MD, a practicing urologist from Alabama who represented the American Urological Association.
Although the Congressional committee members weren’t quite 100% pro-ICD-10, most conceded that the move to ICD-10-CM/PCS was indeed needed. Rep. Kathy Castor (D-Florida) did urge no more delays and Rep. Chris Collins (R- New York) said we should get ICD-10 sooner rather than later.
Not surprisingly, the cost of the transition came up repeatedly and the projected costs varied widely. Edwin M. Burke, MD, from the Beyer Medical Group in Missouri, told the subcommittee that his small practice (two physicians and three nurse practitioners) was already using ICD-10 and they had no problems making the switch. He also said they incurred no additional costs. Their software vendor handled all of the transition work.
“We did not have special training,” Burke said. “We did not spend any money in preparation. We did not see less patients and our practice did not suffer. As providers, it was not frustrating or scary. It just was.”
Rich Averill, director of public policy for 3M Health Information Systems, brought up the just-published Professional Association of Health Care Office Management (PAHCOM) survey that put the costs of implementing ICD-10 for a small practice at approximately $8,000. He also mentioned that two other recent studies put the costs even lower.
Here’s the thing. People will procrastinate. It’s human nature. If you don’t give people a drop-dead deadline, they put it off. Everyone on the subcommittee and all of the witnesses agreed we need ICD-10. So let’s end the suspense and make October 1, 2015, THE date.
Q: Our facility is considering having our coders and CDI specialists “go–live” with ICD-10-CM/PCS on July 1, for practice to help off-set the impact of ICD-10. The system will would then code backwards into ICD-9-CM for billing. Is this “backward mapping” method appropriate?
A: The biggest issue is how your ICD-10-CM/PCS coding will be “translated” into ICD-9-CM for claims processing until CMS accepts ICD-10-CM/PCS codes to process claims. If you the system is coding backward code, the resulting code will likely be based on GEMs, which CMS discourages.
CMS specifically states the GEMS are not for coding purposes, but rather to help build coding databases. Backward mapping—going from an ICD-10-CM/PCS code to an ICD-9-CM code—could be problematic, and could result in assigning many nonspecific codes, which may have reimbursement ramifications. The GEM mappings were not intended to be used for coding purposes, but rather to help build coding databases.
Many codes have a “one to many” ratio, resulting in either a nonspecific code or, in some cases, a “no map” option. Ideally, we would like to think if we convert documentation to ICD-10-CM/PCS that it would automatically backward map to the correct ICD-9-CM code, but it may not.
For example, if the documentation states “severe persistent intrinsic asthma,” the ICD-10-CM assigned code would be to J45.50 (severe persistent asthma, uncomplicated), because the term “intrinsic” in ICD-10-CM/PCS is now an included term not previously factored into category selection in ICD-9-CM. So, if you backwards map the code J45.50, it will translate to 493.00 or 493.10 in ICD-9-CM.
Unfortunately, this does not translate to a direct match: 493.00 is for extrinsic asthma and 493.10 is for intrinsic asthma. The ICD-10-CM/PCS code translates to two possible ICD-9-CM codes, and only one can be chosen. But it has to backward map to both because the code for severe persistent asthma does not identify extrinsic or intrinsic in ICD-9-CM.
Most organizations are dual-coding–coding in both ICD-9-CM so their claims can be reimbursed appropriately and in ICD-10-CM/PCS so they can practice the new code set and identify improvement opportunities. However, not all organizations have software that can “hold” both code sets simultaneously. If your software allows you to hold both codes, even though it is time consuming, the best suggestion is to native code in both ICD-9-CM and ICD-10-CM/PCS. If you rely merely on the backwards mapping, it may not achieve the desired result. The systems are not identical, and very few codes have exact maps.
Editor’s Note: Cheryl Ericson, MS, RN, CCDS, CDIP, AHIMA Approved ICD-10-CM/PCS Trainer, Associate Director for Education at ACDIS and CDI Education Director at HCPro in Danvers, Mass. and Shannon E. McCall, RHIA, CCS, CCS-P, CPC, CPC-I, CEMC, CCDS Director of HIM and Coding at HCPro in Danvers, Mass. contributed to this response.