Radiology Coding Alert

Getting paid for diagnostic studies often depends upon proper ICD-9 coding . Easy as that sounds, experienced radiology coders will tell you diagnosis coding is anything but simple. It requires an understanding of national Medicare policy, as well as local carrier and third-party payer guidelines. Plus, it necessitates dialog between coders and radiologists coders must educate physicians about payers requirements, and physicians should provide coders with the clinical specificity they need to optimize reimbursement.

Diagnosis coding has always been challenging, admits Thomas Kent, CMM, principal of Kent Medical Management, a medical office management and coding consulting firm in Dunkirk, Md. But we are all expecting it to come under even closer scrutiny in upcoming years. Payers will be even more exacting than they have in the past.

Three main problem areas cause the most confusion, Kent says. If coders can master these, they will find claims are processed more smoothly.

Problem #1: Rule out or Probable Diagnoses

Often, radiology coders will see physician orders noting that a study is being done to rule out or confirm a probable diagnosis. These terms and similar language like "suspected" and "questionable" do not support the medical necessity of outpatient services (although they are acceptable when billing for hospital inpatients). HCFAs guidelines for reporting outpatient services, in fact, explicitly state that radiology practices should not use the condition being ruled out as diagnoses. Instead they should code the condition(s) to the highest degree of certainty for that encounter/visit such as symptoms, signs, abnormal test results ...

Rule out and related terms, Kent says, often cause improper coding. Coders need to find out why the study was done. If the test has been ordered to confirm or rule out an illness or condition, something prompted the physician to suspect the disease in the first place. Coders should assign the appropriate codes describing those signs and symptoms.

For instance, if a chest x-ray (71020) was conducted to rule out tuberculosis (TB) in an individual who had been in contact with a TB patient, the coder may assign 786.2 (cough) if documented and V01.1 (contact with or exposure to communicable diseases, tuberculosis) to describe the symptoms. However, the coder would not report a code from the 011series (pulmonary tuberculosis).

Problem #2: Coding Final Diagnosis vs. Symptoms

Another point of confusion for radiology coders is whether a final diagnosis can be reported instead of the signs or symptoms that prompted the service. According to Kent, guidelines differ from payer to payer. Coding guidelines as published in the Medicare Carriers Manual (MCM), for instance, specifically state that the relevant final diagnosis made by a radiologist or other physician must be reported. In fact, section B3 4010 of the MCM contains a specific example of coding results of the x-ray, not the signs and/or symptoms. Other payers, however, may have different requirements, and coders must communicate with representatives to determine which apply.

Ultimately, coders should keep in mind that any diagnosis codes reported with radiology services must support medical necessity at the time the services were ordered. There have to be valid symptoms to justify the exams, points out Cheryl Schad, BA, CPCM, CPC, owner of Schad Medical Management, a billing and coding consulting firm in New Jersey.

Problem #3: Screening vs. Diagnostic Tests

Because many payers restrict reimbursement on screening studies (i.e., bone density studies), coders often wonder if a screening exam may be converted to a diagnostic exam if a positive diagnosis is established after the study.

The answer is no, Kent says. Routine screening exams, ordered in the absence of symptoms to determine if a patient is healthy, must be reported as a screening study even if a condition is uncovered.

If, however, the patient presented with symptoms that indicate a condition, the exam should have been ordered as a diagnostic study and can then be billed as such. For example, a diagnostic bone density scan can be reported for a female patient who complains of bone pain (733.90, other and unspecified disorders of bone and cartilage, disorder of bone and cartilage, unspecified) and has been clinically diagnosed to be estrogen deficient after a hysterectomy (256.2, postablative ovarian failure).

Communication Will Improve Pay Up

In nearly all scenarios when a diagnosis code is in question, more information virtually always helps coders and may pay off with increase reimbursement. Experts note three specific tactics coders may use to get the information they need:

1. Review the patients chart. According to Schad, a little digging may unearth a lot of valuable information. The superbill may contain only a brief explanation of the patients condition, but notes and documentation may mention symptoms that can lead to a specific diagnosis code.

2. Seek out the radiologist or referring physician. Just as chart notes may provide greater specificity, the radiologist may be able to shed additional light. If coders feel they dont have enough information to assign the right diagnosis code, they should ask the radiologist for more detail, Schad says. This information must also be documented in the patient medical record.

Wanda L. Tunstall, CPC, coding manager for Per-Se Technologies, an international firm that delivers comprehensive business management services, financial and clinical software solutions, and Internet-enabled connectively to the healthcare industry, based in Voorhees, N.J., also advises speaking with the referring physician, if necessary. The primary care physician may not realize that the radiologist may not get paid if complete diagnosis information is not included with the orders. Ive found that educating the PCP about what we need is very helpful, too.

3. Let the radiologist know about coding requirements. Sometimes, Schad points out, radiologists dont know specifically which diagnoses establish medical necessity for a particular service and which dont. The business end of healthcare has changed rapidly, and because physicians dont do the coding or billing, they dont realize that what worked at one time may not work now.

Tunstall agrees. These days, we have to be very precise when we report diagnosis codes. Even when some medical terms appear to be the same, payers will accept one, but not the other.

For instance, she says, some radiologists use edema (782.3) and swelling of a limb (729.81) interchangeably. However, one of our local insurers allows swelling of a limb to support a CT of the abdomen, but will reject edema. Physicians need to be aware of cases like these. If they write edema on the report, thats what we have to code. And a claim like this would be denied.

Another example, Tunstall says, occurs when physicians note abdominal distention as the diagnosis when ordering an abdominal CT. Coders may assign 787.3 (flatulence, eructation, and gas pain) because the notes in ICD-9 include abdominal distention within this code. However, distention may also mean enlargement or swelling, and could be coded with 789.30 (abdominal or pelvic swelling, mass, or lump). Code 789.30 would justify an abdominal CT, but 787.3 would not.

When issues like these arise, Tunstall says, she attempts to talk with physicians. If we can let them know about specific language that may cause a denial, they can be more precise in the future.