The Basics of ICD-9 Coding
By Emily Hill, PA-C

Accurate ICD-9 coding helps ensure payment, and avoids any hint of fraud or abuse.

Submitting claims to third-party payors and documenting the care given to each patient is a fact of life in health care today. All medical procedures, supplies and diagnoses are identified by codes.

Current procedural terminology (CPT) coding is designed to numerically identify the cognitive, procedural and material services provided to the patient. International Classification of Disease, Ninth Edition (ICD-9) coding is the system of numerically identifying the disease or physical state that justifies the performance or provision of those services or procedures. In other words, CPT codes are "what you do" and ICD-9 codes are "why you do it." Practitioners must document and code both for each patient. Because reimbursement from third-party payors is largely dependent on the coding selection submitted by the physician, coding accurately is key to getting proper payment for services and avoiding allegations of fraud and abuse.

This month, we will look at the basics of ICD-9 coding. In future issues, we will explore the nuances of CPT coding to ensure that the services reported are accurate and allow for appropriate reimbursement.

What Are ICD-9 Codes?
ICD-9 codes are three-, four- and five-digit codes that describe the diagnosis, symptom, complaint, condition or problem for which a physician sees a patient. These codes are also used to indicate other reasons why a person has sought medical care, such as a routine health examination. The codes are set up to be specific; each digit gives important information about a patient. If the diagnosis can be defined by five digits instead of three or four, all five must be used. Many payors, including Medicare, do not pay for services unless the diagnosis has been carried to the highest number of digits.

When seeing a patient for monitoring and treatment of classical migraines, this descriptor alone would lead a physician to report the ICD-9 code as 346.0. However, it is possible to code to the fifth digit under this category. The fourth digit describes the particular type of migraine, while the fifth indicates whether or not the headaches are intractable. In this example, failure to use the maximum number of digits available in the category will result in denied payment for the office visit. To be sure the highest number is used, look over the codes in that category.

Why Are ICD-9 Codes Needed?
Although the actual payment amount is linked to the CPT code, a proper diagnosis is needed to show why a service was necessary. Each service performed becomes a line item (CPT code) on an insurance claim form, and an ICD-9 code must be linked to it to establish the "medical necessity" of the service. Establishing "medical necessity" basically means justifying choice of treatment by linking the service (CPT code) to the diagnosis, symptom or complaint (ICD-9).

For example, the patient who presents for her regular visit for monitoring of migraine headaches also complains of chest pain. As part of the first step in the workup, an electrocardiogram (ECG) is done in the office. On the claim form, however, the only diagnosis listed is migraines. The insurer will most likely not pay for the ECG because the medical need for the test is not clear. Linking the specific symptom of chest pain or angina to the ECG will indicate why the test was ordered.

If a Patient Only Has Symptoms, What Should Be Coded?
Code just the facts. Patients are often seen for ill-defined complaints, such as "chest pain" or "numbness." A specific condition may be suspected, and appropriate diagnostic tests ordered; however, until the results are known or until a definitive diagnosis is made, only the symptom should be coded. There are no codes for words like "suspected" or "probable"; labeling patients with a disease without certainty should be avoided.

What If There Are Several Facts Or Diagnoses for a Patient?
Prioritize. When more than one diagnosis applies to a patient, code the primary diagnosis first, followed by the second most important, etc. Rank diagnoses so that the one listed as primary is the one receiving the most emphasis at that encounter. There is no reason to code a diagnosis that does not affect patient care.

Up to four diagnoses can be submitted on each claim form and linked to each service. Make sure that an appropriate diagnostic code is specifically linked to each CPT code to show that the service is appropriate and medically necessary. Remember, some payors will read only the first diagnosis linked to an individual CPT code when processing the claim. It is therefore important that the prioritizing and linking of diagnostic codes are done accurately.

Coding appropriately can streamline the payment process from third-party payors. Even in managed-care environments, accurately identifying the reasons patients are treated is important for practice profiling and contract negotiations. In addition, a review of the reasons patients come into the physician’s office often provides valuable management information. Although the coding process may seem confusing at first, it is an important time-saving and management tool.

Emily Hill, PA-C is President of Hill & Associates, a national health-care consulting firm specializing in coding and compliance for physician practices. Her e-mail address is