Pulmonology Coding Alert

High on the federal governments hit list is pneumonia upcoding. As part of its workplan for fiscal year 2000, the Office of the Inspector General (OIG) of the U. S. Department of Health and Human Services (HHS) specifically mentioned pneumonia upcoding in the list of things for which it will be watching. Hospitals that insist physicians keep complete patient records with full documentation of medical decisions need not worry about passing OIG audits.

The pneumonia-upcoding project focuses on identifying hospitals where coders are assigning viral/bacterial pneumonia diagnosis codes when the condition is not present or is not substantiated by the medical record documentation. Originally, the project was concerned primarily with pneumonia that was coded 482.89 (pneumonia due to other specified bacteria; other specified bacteria) but has since widened its investigation to include 482.83 (pneumonia due to other specified bacteria; other gram-negative bacteria). Another issue added to the list is when a hospital coder uses 482.89 instead of 482.9 (bacterial pneumonia unspecified), according to William Haik, MD, a pulmonologist in Ft. Walton Beach, Fla., and director of DRG Review, a hospital coding consulting firm. Actually, there are codes for most types of pneumonia. Any particular strain of pneumonia-causing bacteria that is not classified in the ICD would be very rare indeed. The difference may be subtle, but the red flags potentially raised during an audit are not, says Haik.

The more serious strain of pneumonia (482.89 and 482.83) falls under the diagnosis related group (DRG) 079. The cases that rightly should fall under DRG 089 keep the pneumonia DRG-upcoding project on the front burners at HHS and the U.S. Department of Justice (DOJ).
According to the HHS OIG semiannual report released in March, The OIG has found that a small percentage of hospitals across the country have assigned a disproportional number of pneumonia cases ICD-9 codes that result in a diagnosis being assigned the higher-paying DRG. According to a spokesperson for HHS OIG, to date 17 hospitals have settled their liability for such coding by paying more than $21 million and agreeing to initiate integrity requirements. Earlier this year, three hospitals were forced to pay a total of nearly $5 million as a result of their pneumonia upcoding.

OIG specifically targeted pneumonia upcoding for the past few years, and for 2000 has upgraded its Pneumonia DRG Upcoding Project to priority status.
Haiks advice to coders is to query the attending physician to make sure that the organism is linked to the condition, and scrutinize the manual carefully to ensure that, before using 482.89, the identified organism is definitely not included in the classification list.

It is not enough simply to document pneumonia in the medical record. At the very least, the physician should specify the type of pneumonia in writing, and to be on the safe side, cultures and gram stains should be included as well.

How Does an OIG Hospital Audit Affect a Pulmonologist?

Although these issues may not seem to affect pulmonology practices directly, as physicians attempt to provide the most specific diagnoses possible they should keep several considerations in mind, says Susan Callaway-Stradley, CPC, CSS-P, an independent pulmonology coding consultant and educator in North Augusta, S.C. One consideration is that the financial stability of local hospitals can be harmed by assessments against them, causing a trickle-down effect within the community at large, as well as within the local medical community. Pulmonologists who cause financial loss to a hospital might find their potential earnings and their ability to provide the best possible care to their patients is affected. In addition, when hospital records are examined, physicians records are available for scrutiny as well, and that puts the physician at risk for audit for any number of coding issues.

Physicians must recognize that what they write on a chart, which medications and tests they order, and the clarity and completeness of their written record of test results and the information that led to their decisions has a direct influence on coding and billing and therefore an influence on OIG audit results as well as the hospitals bottom line.

Editors note: See Documentation is Key to Risk Level on page 61.

Documentation Is Key to Risk Level William Haik, MD, a pulmonologist in Ft. Walton Beach, Fla., and director of DRG Review, a hospital coding consulting firm, explains that problems often arise when a physician writes pneumonia in the patient record and doesnt specify what type of pneumonia he or she observed. Coders have to use codes that reflect patient-record narrative. Moreover, narrative about the diagnosis needs to be supported by test results, such as cultures or gram stains, in the medical record. When Haik reviews his clients charts he divides fraud risk into four levels, with definitions that range from a high risk of fraud charges, level 1, to full compliance, level 4: Level 1. The physician writes pneumonia in the chart. He or she doesnt specify the type of pneumonia, and there are no cultures or gram stains to support gram-negative bacterial pneumonia. If the coder assigns code 482.83 (pneumonia due to other specified bacteria; other gram-negative bacteria), that is clearly wrong. Level 2. The physician writes pneumonia in the chart without specifying the type. Again, there are no cultures, gram stains, or other information to support gram-negative pneumonia. The coder asks the physician if its gram-negative pneumonia, and the physician says yes but doesnt put that answer in writing. Coding 482.83 still could be considered fraudulent because theres no evidence of gram-negative bacterial pneumonia in the medical record. Editors note: Although the coder has taken a step toward compliance at level two, the federal government views level 1 and level 2 as basically the same. Level 3. Again, the physician writes pneumonia with no specificity. This time, however, there is clinical information in the medical record to support gram-negative bacterial pneumonia. For example, the patient is at high risk for gram-negative bacterial infections because he or she has cancer and is receiving chemotherapy. If possible, the record also could include a gram stain. The coder sees the chart information and queries the physician in writing, pointing out the supporting clinical information in the record and asking if the patient has gram-negative pneumonia. The physician answers in writing that the patient has gram-negative pneumonia. With this documentation in the record, Haik says, it would be more acceptable to code 482.83. Level 4. The physician writes on the chart that the patient has gram-negative bacterial pneumonia. There is clinical evidence to support that diagnosis. This is the only scenario that is completely risk-free, Haik says. Although Level 3 may be acceptable, it is not risk-free. If the coder has to ask the physician for documentation, theres always a chance of miscommunication. In addition, the OIG auditor could say theres no proof that the physician really thought about the record when he or she answered a coders query. The only 100-percent-sure scenario, Haik advises, is when the physician has actually written or dictated the information in the progress notes or discharge summary, on the date he or she made the diagnosis.