After two one-year reprieves, it’s finally here: The International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM).
Nationwide, on October 1, 2015, all health-care business transactions in the United States converted to ICD-10 from the previous version of the ICD (ICD-9). If they did not convert, claims and electronic transactions will be rejected. The ICD-9 coding system has been used in the United States since 1979, meaning for most people working in health care, it is the only diagnostic coding system they have known.
The Long Road to ICD-10 Implementation
According to Daniel B. Martin, MD, from the University of Washington School of Medicine and Seattle Cancer Care Alliance, the main problem with ICD-9 is its age. Not only does ICD-9 fail to reflect progress in both disease and diagnostic knowledge, said Dr. Martin, but its practically ancient 1979 format does not lend itself to upgrades.1
Work on ICD-10 began in 1983, just four years after ICD-9 went into effect, and was completed in 1992. Most countries have already transitioned to ICD-10 (Canada, for instance, completed its five-year changeover in 2005).
ICD-10 is expected to improve “tracking of diagnosis trends, public health needs, and epidemic outbreaks,” Dr. Martin said, “and will enable more accurate payments for new procedures; reduce miscoded, rejected, and improper reimbursement claims; and provide a better understanding of the value of new procedures.”
All of this will lead, presumably, to systemic cost-savings, but the road to those savings is a costly one. To prepare for implementation, organizations have had to devote significant resources to information-technology upgrades, training, and payer contract renegotiations. Lost productivity also adds to the total expense: One Canadian hospital reported that the changeover to ICD-10 slowed overall productivity for more than a year.2
How Will the Conversion Affect Hematologists?
The greatest challenge to implementation of ICD-10 has been the changes in diagnostic code reporting from ICD-9. For instance:
- The total number of ICD-9 diagnostic codes is about 13,000. ICD-10 has about 68,000.
- The ICD-10 code set has been expanded from five positions (first one alphanumeric, others numeric) to seven positions. The codes use alphanumeric characters in all positions, not just the first position as in ICD-9.
- There is no clear mapping between ICD-9-CM and ICD-10-CM code sets. There are some one-to-one correspondences, but often there are one-to-many, many-to-one, many-to-many, or no correspondence at all.3
And, in most cases, the increase in codes represents an increase in the specificity of the codes available, rather than a radical reclassification of the diseases. In ICD-9, for example, myeloid leukemia has 24 codes (205.xx), whereas in ICD-10 it has 40 different codes (C92.xxxx). Some of those newer codes allow one to report the disease with the presence or absence of certain genetic markers (FIGURE).
On the brighter side, the trauma may be lower for hematologists and oncologists than for some other specialties with more “convoluted mappings” between ICD-9 and ICD-10, Dr. Martin said. Some changes will be simple, but changes to details in disease subtype classifications may seem “unpredictable.”
“In the ICD-10 system you need to be much more specific,” Paul Fishkin, MD, a hematologist from Illinois CancerCare in Peoria, Illinois, told ASH Clinical News.
Each patient’s progress note has to stand completely on its own in the medical record, Dr. Fishkin explained, offering the following example: “Simply designating ‘hypertension’ as a comorbid condition would be insufficient. Rather, primary or secondary hypertension has to be indicated, and if it is secondary to renal disease, the stage of renal disease must be noted also.”
Dr. Fishkin provided a bevy of further examples: “Designation of myelodysplastic syndrome in ICD-10 also requires more specificity than ICD-9, including indication of ‘refractory anemia with ring sideroblasts,’ while designation of anemia would require more differentiation such as in ‘iron deficiency anemia caused by malabsorption or blood loss’ or ‘chemotherapy-induced anemia.’” In multiple myeloma, coding is broken out to multiple myeloma, plasma cell myeloma, and solitary myeloma, with further refinement stating whether or not the condition is in remission, in relapse, or with remission not achieved.
One Center’s Experience with ICD-10
Preparation at Illinois CancerCare, where 70,000 to 80,000 physician visits occur annually, has been extensive and ongoing for several years, Dr. Fishkin said. Illinois CancerCare certified coders Ester Harn (auditor/educator) and Cheryl House (compliance specialist) agreed that it will take coders more time to code with ICD-10. Time management, then, will be a challenge.
About 80 percent of progress notes at Illinois CancerCare reflect some comorbidity and diabetes, a common comorbidity which used to encompass 20-30 codes, now has more than 50 in ICD-10. Dr. Fishkin noted further that elevated blood sugar has to be specified as either type 1 or type 2 diabetes, and hypertension designations are divided according to whether hypertension is drug-induced or stems from vascular disease. “I am not treating their diabetes, but in order to code properly and get payments through, we have to code all the comorbidities,” Dr. Fishkin said.
The coding categories requiring designations for tumor location, laterality, and extension, Dr. Fishkin observed, are much more concerned with anatomy than biology or histology. To be able to designate the quadrant of the tumor in a breast cancer patient who he had treated for more than a decade, he had to hunt down the original mammogram. “That’s not always easy,” he commented.
Information about the source of illness is being collected, an example being the need to indicate smoking status and body mass index, which might allow payers to charge higher rates for those with “self-inflicted” conditions, Ms. Harn and Ms. House speculated. The 24 certified coders (and five more in training) at Illinois CancerCare have been through “boot camps” and twice- or thrice-weekly training sessions over the last two years – amounting to 50 to 100 instructional hours per coder. The coders, in turn, are working to help the center’s physicians make progress notes more specific through one-on-one meetings, tip sheets, and audits of their notes.
Dr. Fishkin shared a few corrective notes he received from coders. He had dictated “classic Hodgkin disease, stage IIA.” The coder’s suggestion: “Stage IIA classic Hodgkin disease, right cervical lymph node.” Dictation: “Small cell lung cancer with liver and lung metastases.” Suggestion: “Small cell lung cancer, right upper lobe, with liver and lung metastases,” and a suggestion that further detail on the location of the metastases (e.g., mediastinum or pleura) be provided.
The Illinois CancerCare coders pointed out that in situations in which definitive information is unavailable, a designation of “unspecified” is possible, but the chances of a claim being rejected go up.
How will ICD-10 implementation play out? Dr. Fishkin relayed that one physician in his practice suggested that it will be “another Y2K,” a lot of fuss, but then, “Poof – it all will happen.” The other scenario: “If it all backs up, we could be totally out of cash in 30 days.”
- Johnson K. Implementation of ICD-10: experiences and lessons learned from a Canadian Hospital. Accessed September 9, 2015 from http://library.ahima.org/xpedio/groups/public/documents/ahima/bok3_005558.hcsp?dDocName=bok3_005558.
- Martin DB, Silas S, Covner A, et al. Development of a hematology/oncology ICD-10 documentation job aid. J Nat Comprehensive Cancer Network. 2015;13:435-440.
- Centers for Medicare & Medicaid Services. “ICD-10 changes from ICD-9.” Accessed September 13, 2015 from www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Data-and-Systems/ICD-Coding/ICD-10-Changes-from-ICD-9.html.
|FIGURE. Key points about the ICD-10 coding structure|
|C92.11: Chronic myeloid leukemia, BCR/ABL-positive, in remission