Internal Medicine Coding Alert

Payers Speak Out:

Avoid These Top IM Coding and Billing Mistakes

Is your internal medicine practice using the most up-to-date versions of CPT and ICD-9? Do your physicians know to accurately link the diagnoses listed on the encounter form with the proper procedure codes in order to support the services rendered? Do you have methods in place to ensure you get the correct insurance and demographic information from each patient?

If you answered yes to all of these questions, you are way ahead of many of your peers. We contacted claims processing personnel at health plans across the country, as well as several coding and reimbursement consultants, to find out the most common errors internists offices make when submitting their claims. Here are the concerns they listed:

1. Inaccurate reporting of diagnosis codes.

(A) Incorrect linking: Payers and consultants alike outlined problems with the way practices reported ICD-9 codes and the way they were linked with medical procedures.

A key issue is the failure to use specific diagnosis codes when they are available, says Catherine Brink, president of Healthcare Resource Management, in Spring Lake, NJ.

For example, a practice may report a V70 (general medical examination) when the encounter really involved an evaluation of chest pain, stomach pain or some other specific ailment, says Diana Seymour, manager of front-end claims systems for Blue Cross Blue Shield of Arizona. We see a lot of submissions of a routine diagnosis code when a medical condition is being treated. And, maybe that patient has a contract that doesnt pay for routine service. If the diagnosis had indicated a specific medical problem, the claim would have gone through, no problem.

This lack of specificity may also mean that the diagnosis code will not match up with the procedure code, which also usually results in the claim being rejected, adds Brink.

(B) Use of the fifth digit: In many cases, coders also tend to leave off a necessary fifth digit in the ICD-9 code, or, in at least a few cases, put on a fifth digit when none is necessary. This either causes the claim to be downcoded or sent back to the payer for correction.

Among the claims that we have to pull out and seek more information on, we see a lot of invalid diagnosis codes, says Seymour. What we usually see is an extra zero at the end. What we are finding is that, if we drop the zero, the codes become valid.

Seymour speculates that some practice information systems may have a space for five digits, and some coders are trying to fill in all of the spaces even though not every diagnosis code needs five digits.

Seymours department has coders that attempt to figure out why the claim is being rejected. If they can determine the problem, as in the above situation, they run the corrected claim back through. However, she notes, this significantly delays reimbursement. Other payers may simply send the claims back.

(C) Multiple diagnoses: Another problem Brink sees in a lot of practices is inaccurate linking of multiple diagnoses and procedures.

Practices have to show the medical necessity for the service that was rendered by coding the appropriate ICD-9, she explains.

To illustrate this point, Brink uses the example of a established patient who presents in the office with a bleeding ulcer (533.40, peptic ulcer with hemorrhage), but who also has hypertension (402.90, hypertension with heart involvement, unspecified). The patient may see the physician for the ulcer, but while in the office, the physician performs an EKG to evaluate the hypertension.

The physician selects CPT code 99214 for an established patient, since there were two medical problems, a detailed history, and a detailed exam that required moderate medical decision-making.

There are two diagnoses that should be submitted for the 99214 visit, Brink says. The bleeding ulcer and the hypertension which the physician should link on the encounter form in order of medical necessity.

In box 21 of the HCFA form 1500 there are four spaces for diagnosis codes. The coder should list the diagnoses here in order of medical necessity, she explains.

Here is how box 21 would look for this visit:

1. 533.40

2. 402.90

3. Blank. No other diagnosis

4. Blank. No other diagnosis

However, in box 24, there are two columns, D and E. In the D column, the coder should list the procedures performed. In the E column, the coder should list the number of the diagnosis code that the procedure is linked to. In this case, the coder would list 93000 (EKG, with interpretation and report) with the diagnosis number 2 (corresponding to the place in box 21).

This is how box 24 would look for this visit:


D E
99214 1,2
93000 2


If there is no distinction on the charge form from the physician, the coder or person doing data entry might put down both diagnosis codes for the procedure code or enter the incorrect ICD-9 which would not support the medical necessity for the EKG, she adds.

For the purpose of billing for the EKG, the payer is not going to care that I have a bleeding ulcer, Brink says. Thats not going to get the claim paid.

2. Inaccurate CPT coding. Use of outdated or deleted procedure codes is also a common error, says Sandy Brown, claims manager for Health Plus of Louisiana, a 35,000-member HMO in Shreveport.

Practices often submit codes that have been removed from the CPT or they use specific codes in error, particularly age-specific codes, adds Seymour. I think thats a problem for everybody.

Though it may be too expensive to buy new CPT and ICD-9 books for everyone in the practice each year, many practice managers and consultants advise internists to invest in a few new copies that are stored in a central location.

3. Misuse and nonuse of modifiers. This is a problem almost all payers and consultants mentioned. Internal medicine practices seem to have particular problems with modifiers -25 (significant, separately identifiable E/M service on the same day as a procedure), -59 (distinct procedural service), and -26 (professional component), according to several sources.

Using the -26 modifier incorrectly will result in the practice getting a lower reimbursement than it should because modifier -26 is attached to a procedure code when the provider performs only the professional component of the service, i.e., reading or interpreting a diagnostic test. If the physician actually performed the test and interpreted it with a written report, then the practice should bill the CPT code without the modifier in order to receive full payment.

A lot of [physicians] will bill -26 even though they are actually performing a test, notes Brown. Hopefully, we would notice and pay the full fee.

Brown works in an HMO that employs its physicians and has certain rules about tests that are performed in the office, she says.

However, Brink adds, in most cases, payers would not have enough information to realize the modifier was attached incorrectly. The providers would really be losing out.

Because many health plans dont recognize modifiers, or only recognize some, it is important for practice managers to contact each plan to find out which modifiers they accept and how they want to see them reported.

Modifiers, from what I hear, are really the biggest problem in getting accurately paid, Brink explains. Some practices are using them correctly, but the payers they contract with do not recognize the modifiers and the claims arent getting paid.

However, most coding experts advise practices to continue to use modifiers because they are correct coding and indicate exactly the type of service that has been provided. Practice managers should try to convince their major carriers to accept the modifiers. If they wont, it might be worth re-examining the value of the payers contract.

Note: In upcoming issues of IMCA, we will cover the modifiers most commonly used in internal medicine coding.

4. Inadequate or incorrect claim information. A problem that is unrelated to coding, but still results in a number of unpaid claims, is inaccurate patient or payer information, says Bette Warn, CMPE, principal of Medimetrix Consulting in Denver, CO.

What happens a lot of times is that they wont have the right demographic information, like the patients correct Social Security number, says Warn. Or, sometimes, the patient has switched insurance plans and the practice has submitted the bill to the wrong payer.

These situations can be particularly damaging for the practice because unlike claims rejected for incorrect coding, they will remain in limbo, and so will your money.

Providers should double-check patient information at each visit, and always make a copy of the patients insurance card.