Improving Specificity in ICD-10 Diagnosis Coding

Untitled Document

Valid ICD-10-CM/PCS (ICD-10) codes have been required for claims reporting since October 1, 2015. But ICD-10 diagnosis coding to the correct level of specificity—a more recent requirement—continues to be a problem for many in the healthcare industry. While diagnosis code specificity has always been the goal, providers were granted a reprieve in order to facilitate implementation of ICD-10. For the first 12 months of ICD-10 use, the Centers for Medicare and Medicaid Services (CMS) promised that Medicare review contractors would not deny claims “based solely on the specificity of the ICD-10 diagnosis code as long as the physician/practitioner used a valid code from the right family.”1 Commonly referred to as the “grace period,” this flexibility was intended to help providers implement the ICD-10-CM code set and was never intended to continue on in perpetuity. In fact, this CMS-granted grace period expired on October 1, 2016.1

Unfortunately, nonspecific documentation and coding persists. This is an ongoing problem, even though the official guidelines for coding and reporting require coding to the highest degree of specificity. Third-party payers are making payment determinations based on the specificity of reported codes, and payment reform efforts are formulating policies based on coded data. The significance of over-reporting unspecified diagnosis codes cannot be understated. In the short term, it will increase claim denials, and in the long term it may adversely impact emerging payment models.3,4 Calculating and monitoring unspecified diagnosis code rates is critical to successfully leverage specificity in the ICD-10-CM code set.

An ICD-10-CM code is considered unspecified if either of the terms “unspecified” or “NOS” are used in the code description. The unspecified diagnosis code rate is calculated by dividing the number of unspecified diagnosis codes by the total number of diagnosis codes assigned. Health information management (HIM) professionals should be tracking and trending unspecified diagnosis code rates across the continuum of care.5

Acceptable Use of Unspecified Diagnosis Codes

Unspecified diagnosis codes have acceptable, even necessary, uses. The unspecified code rate is not an error rate, but rather an indicator of the quality of clinical documentation and a qualitative measure of coder performance and coding results. Even CMS explicitly recognizes that unspecified codes are sometimes necessary. “When sufficient clinical information is not known or available about a particular health condition to assign a more specific code, it is acceptable to report the appropriate unspecified code.”6 It’s also important that coding professionals use good judgment to avoid unnecessary queries for clarification of unspecified diagnoses. The official coding guidelines provide explicit guidance for appropriate uses of unspecific diagnosis codes.7

Overuse of Unspecified Diagnosis Codes

Overuse of unspecified diagnosis codes is a problematic trend. Use of unspecified ICD-10-CM codes, ignored during the first year following implementation of the code set, is not improving. In this author’s experience, unspecified diagnosis code rates can range anywhere from 20 percent, on the low end, to over 40 percent. A diagnosis code rate over 30 percent requires investigation and appropriate corrective actions.

Widespread use of unspecified codes should be the exception, not the rule.8 High unspecified diagnosis code rates may be due to either clinical documentation or coding practices. One strategy to ensure unspecified codes are used appropriately is to monitor unspecified code rates and address any upward trends or outliers. HIM professionals should spot check the most commonly reported unspecified codes. Review the clinical documentation on a targeted sample to confirm that the reported codes align with the degree of specificity in the health record. Examples of some conditions most commonly reported with unspecified diagnosis codes are listed in the sidebar above.

HIM professionals should identify the most commonly reported unspecified diagnosis codes in their facility and review a sample of related encounters to confirm that unspecified codes were used appropriately. When overuse of unspecified codes is identified, solutions may involve improvements in the specificity of clinical documentation or process improvement to ensure coding professionals are coding to the highest degree of specificity that is available.

Conditions Commonly Reported with Unspecified Diagnosis Codes

  • Alcohol and drug use, abuse, and dependence
  • Alzheimer’s disease
  • Arthritis
  • Asthma
  • Atrial fibrillation
  • Atrial flutter
  • Cardiomyopathy
  • Cerebral palsy
  • Congestive heart failure (CHF)
  • Depression
  • Epilepsy
  • Intestinal obstruction
  • Migraines
  • Neoplasms
  • Non-pressure ulcers
  • Pneumonia
  • Pressure ulcers
  • Respiratory failure
  • Sepsis
  • Strokes
  • Traumatic injuries

Specificity in Physician Documentation

Coding specificity is a shared responsibility between the provider and the coding professional to create a clear clinical picture of the encounter. Providers have an obligation to document conditions to the full extent of their clinical knowledge of the patient’s health. Toward this aim, providers may need assistance—in the form of provider education or clinical queries—to recognize relevant clinical details that may impact the specificity of code assignment. Inpatient and outpatient clinical documentation improvement (CDI) efforts should address documentation specificity to leverage the details available in ICD-10-CM.

Coding to the Highest Degree of Specificity

While physicians are expected to document the most specific clinical diagnosis, it is equally important that coding professionals assign diagnosis codes to the highest degree of specificity documented. There is a disturbing amount of unspecified diagnosis code reporting when more specific diagnoses are documented in the health record. Coding professionals must continually train their “coder eye” to look for specificity in provider documentation. A finely tuned “coder eye” and attention to the level of specificity available in the ICD-10-CM code set will ensure the highest degree of specificity of the codes assigned and reported.

As a final step in the coding process, coding professionals should perform a final review of the diagnosis codes on an encounter. Diagnosis codes ending in the numbers zero or nine are often indications that an unspecified diagnosis code was assigned. A quick second review of the clinical documentation associated with these codes may reveal clinical details needed to derive a more specific diagnosis code. Supporting documentation, particularly imaging reports, may be used to code to the highest degree of specificity when the physician has already documented a condition. AHA Coding Clinic for ICD-10-CM and ICD-10-PCS, from the American Hospital Association, provides guidance for using documentation from imaging reports when a physician has already documented a condition, such as a fracture, stroke, or pain.9,10,11 The following coding examples demonstrate appropriate coding to the highest degree of specificity.

Coding Example A

For this example, the documentation in the outpatient setting is as follows: the emergency department provider documents a “non-displaced right talus fracture.” The right ankle X-ray documents a “non-displaced avulsion fracture of the right talus.”

The code S92.101A, Unspecified fracture of right talus, was initially assigned. After a final review, the code S92.154A, Nondisplaced avulsion fracture of right talus, is assigned based on the greater specificity found in the imaging report.

Coding Example B

For this example, the documentation in the inpatient setting is as follows: the history and physical and discharge summary document that the patient suffered an “embolic cerebral infarction with residuals.” The brain CAT scan report documents a “large embolus in the right middle cerebral artery (MCA) territory.”

The code I63.40, Embolic cerebral infarction of unspecified artery, was initially assigned. After a final review, I63.411, Cerebral infarction due to embolism of right middle cerebral artery, should be assigned based on greater specificity found in the brain CT report.

Addressing Nonspecific Documentation and Coding

Despite the importance of specific documentation and diagnosis code reporting, nonspecific documentation and coding persists. The solution lies in addressing both improvements in the specificity of clinical documentation and process improvement to ensure medical coding professionals are coding to the highest degree of specificity that is available. The challenge for medical coders is to achieve a balance in productivity and quality to ensure that reported codes align with clinical documentation. The goal is to report specific diagnosis codes when they are supported by the available medical record documentation and clinical knowledge of the patient’s health condition—and use unspecified codes only when they are the best choice to accurately reflect the healthcare encounter.

Notes

  1. Centers for Medicare and Medicaid Services. “CMS and AMA Announce Efforts to Help Providers Get Ready For ICD-10 Frequently Asked Questions.” www.cms.gov/Medicare/Coding/ICD10/Downloads/ICD-10-guidance.pdf.
  2. Centers for Medicare and Medicaid Services. “Clarifying Questions and Answers Related to the July 6, 2015, CMS/AMA Joint Announcement and Guidance Regarding ICD-10 Flexibilities.” www.cms.gov/Medicare/Coding/ICD10/Clarifying-Questions-and-Answers-Related-to-the-July-6-2015-CMS-AMA-Joint-Announcement.pdf.
  3. Eramo, Lisa A. “Don’t Deny the Denials: Experts Recommend Implementing a Strong Claims Denial Strategy to Offset ICD-10-based Coder Productivity Loss.” Journal of AHIMA 85, no. 6 (June 2014): 30-33.
  4. Fernandez, Valerie. “Ins and Outs of HCCs.” Journal of AHIMA 88, no. 6 (June 2017): 54-56.
  5. Stanfill, Mary H. “Call for Additional Coding Metrics.” Journal of AHIMA 86, no. 4 (April 2015): 56-57.
  6. Department of Health and Human Services. “Information and Resources for Submitting Correct ICD-10 Codes to Medicare.” MLN Matters. 2014. www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNMattersArticles/Downloads/SE1518.pdf.
  7. Department of Health and Human Services. “ICD-10-CM Official Guidelines for Coding and Reporting FY 2017.” www.cdc.gov/nchs/data/icd/10cmguidelines_2017_final.pdf.
  8. Lee, Christine. “Nothing Lasts Forever: How Physicians Can Prepare for the End of the One-Year Amnesty from Post Payment Reviews Due to Unspecified Codes.” ICD-TEN. May 2016. http://bok.ahima.org/doc?oid=301570.
  9. American Hospital Association. “Using the X-ray Report for Specificity.” AHA Coding Clinic for ICD-10-CM and ICD-10-PCS (First Quarter 2013). Chicago, IL: AHA Central Office, 2013: 28.
  10. American Hospital Association. “Use of Imaging Reports for Greater Specificity.” AHA Coding Clinic for ICD-10-CM and ICD-10-PCS (Third Quarter 2014). Chicago, IL: AHA Central Office, 2014: 5.
  11. American Hospital Association. “Use of X-ray to Determine Site of Pain.” AHA Coding Clinic for ICD-10-CM and ICD-10-PCS (Fourth Quarter 2016). Chicago, IL: AHA Central Office, 2016: 143.

Jacquie Zegan (jzegan@uasisolutions.com) is senior coding consultant at United Audit Systems, Inc.