ED Coding and Reimbursement Alert

Emergency ICD-9-CM:

Coding Signs and Symptoms in the ED to Justify Diagnosis Tests, Services

Its a common diagnosis coding dilemma in EDs across the country: A patient comes in complaining of severe chest pain. It might be a heart attack, or it might be a bad case of indigestion. To the patient, the symptoms can be the same. To determine the final diagnosis, the physician must perform an extensive physical exam and order an array of tests to determine whether the patient has a cardiac-related illness. The extensive workup reveals only severe heartburnreported clinically as gastritis (i.e., 535.00, acute gastritis without mention of hemorrhage).

But, should this diagnosis be the only one reported? Should the CPT Codes for the extensive tests and procedures performed in the emergency department be linked to a diagnosis code that many payers will consider non-emergent?

Many coders have been trained to report the most specific diagnosis availablewhich means that if a diagnosis is established (i.e., gastritis), then it should be the primary code reported. They feel that only in situations where a diagnosis is not known to the physician could a sign or symptom be reported (e.g., chest pain, other -
discomfort, pain, tightness in the chest 786.59
). Others contend that when diagnostic tests or evaluations are performed, the final diagnosis was unavailable to the physician, so only codes for signs and symptoms properly indicate the reason for the test or exam.

This is a very confusing issue for many ED coders, both for those that code for the facility and for the physician group, says Sharon Timms, RN, MSN, manager of compliance and quality management for LYNX Medical Systems, Inc., an emergency medicine coding, billing and software development company based in Bellevue, WA.

The American Hospital Associations (AHA) outpatient [diagnosis coding] guideline 12.1A specifically states for accurate reporting of ICD-9-CM diagnosis codes, the documentation should describe the patients condition, using terminology which includes specific diagnoses as well as symptoms, problems or reasons for the encounter, she explains. However, it goes on to say that codes that describe symptoms and signs, as opposed to diagnoses, are acceptable for reporting purposes when an established diagnosis has not been confirmed by the physician.

To Timms, this statement means that the AHA guidelines indicate that a final diagnosis should be used in lieu of codes for signs and symptoms if the final diagnosis
is available.

However, Caral Edelberg, CPC, a member of the American College of Emergency Physicians Coding and Nomenclature Advisory Committee (CNAC) and president of Medical Management Resources, Inc., an emergency medicine coding and consulting company in Jacksonville, FL, offers a different interpretation of that guideline.

Diagnosis coding guidelines adopted by Medicare state that the coder should report the correct ICD-9-CM code from 001.0 to V82.9 to indicate the diagnoses, symptom, condition, problem or complaint or other reason for the patient encounter, she says. The four Cooperating Parties that are responsible for the development of ICD-9-CM codingthe AHA, the American Health Information Management Association (AHIMA), the Health Care Financing Administration (HCFA), and the National Center for Health Statistics (NCHS)all developed and approved these guidelines.

Although these guidelines were developed for use on government claims (Medicare and Medicaid), most commercial carriers are also expected to adopt them if they have not already, Edelberg adds. (See box on page 43 for a list of the guidelines.)

ED coders should be sure to report the diagnosis code that most accurately depicts the reason that the test or service was performed, she says.

The AHA outpatient guidelines may not specifically indicate the validity of coding signs and symptoms to justify the medical necessity of certain tests and services, that AHAs participation in the development of the Medicare guidelines indicates that this is correct, she says.

Do Not Code Suspected Conditions

Both Edelberg and Timms agree that signs and symptoms can be coded to justify diagnostic tests and services when the final diagnosis is not known.

However, they both emphasize that it is incorrect to report suspected or rule-out diagnoses as if they were confirmed. For example, a cardiac diagnosis should not be linked to the CPT code for the EKG, until the physician has confirmed that the patients illness is related heart condition, they both say.

Instead, coders should only report diagnoses that the physician knows to be true. In many cases, that may be a sign or symptom, such as chest pain.

Edelberg and Timms differ in their opinion of whether signs and symptoms can be reported in the event a final diagnosis is eventually established.

Reimbursement The Heart of the Debate

Although all coders know that you should not code to get paid, how well you use coding can often mean the difference between financial viability and bankruptcy.

The difficulty with this issue arises with some insurance carriers payment policy for emergency services, says Timms. Payers frequently have edits for diagnoses that are considered emergent vs. non-emergent and deny payments for ED visits with non-emergent diagnoses listed.

This situation is even more complicated by the fact that different third-party payers have a variety of policies on what is considered emergent and what is considered non-emergent.

In addition, Medicare carriers have lists of acceptable diagnoses that justify the medical necessity for diagnostic tests such as laboratory tests, x-rays, and EKGs, she adds. Without an acceptable diagnosis from the list, the hospital may not be reimbursed for the test.

For example, the code for chest pain (786.50) is on most carriers list of acceptable diagnoses for an EKG. the code for gastritis (535.00), however, is not. Gastritis also might not be considered an emergent diagnosis by many payers, and the payment for the entire visit might be downcoded or denied.

As you can see, reporting only a final, established diagnosis can have drastic reimbursement consequences for both the ED physician and the hospital.

Physicians and the hospital are in between a rock and a hard place, particularly because they dont have a choice over whether to perform the extensive workup or diagnostic tests, continues Timms.

The Emergency Medical Treatment and Active Labor Act of 1986 (EMTALA) mandates that patients must be evaluated to determine whether an emergency medical condition exists, she says. When a patient presents with chest pain, an evaluation to determine whether an emergency condition exists might well include a cardiac workup based on the nature of the patients presenting symptoms. As you can see, coders are caught in the middle.

Note: Federal legislation has been proposed that would require all health care insurers to reimburse ED claims based on presenting symptoms rather than final diagnoses, but this bill has never been passed. However, federal law does require Medicare to pay ED claims based on presenting signs and symptoms, and some states have passed similar legislation for payers in their states. ED coders should check to see whether their state legislature has passed such a law.

Code Should Indicate Reason for Encounter

According to the guidelines, the diagnosis code chosen should reflect the diagnosis or problem that is the chief reason for the encounter, Edelberg emphasizes. In the ED, the chief reason for the encounter must relate to the physicians consideration of the patients signs, symptoms, and/or complaints, particularly if they indicate a potentially higher level of acuity than the final diagnosis.

To support her interpretation, Edelberg points out that Medicare policy on medical necessity for services and procedures indicates that ICD-9 codes should indicate the reason that the service or procedure was performed.
Furthermore, in keeping with the federal law, the Medicare carrier manual specifically states that determination of a medical emergency should be prospectively on the presenting signs and symptoms, not retrosectively on the final diagnosis, she states.

It is clear that they intend the diagnosis reported for a test or procedure to represent the condition known to the physician at the time the service is performed, not a condition that may be confirmed several hours later, she adds.

The best advice to offer is for ED management and coding staff to consider the different interpretations of the diagnosis coding guidelines, go over the available guidelines themselves (see box at left), and examine their payers policies and the laws in their state to determine a method for accurately and appropriate reporting their emergency department encounters.