MDS Alert

Expert advice gets to the heart of practices that lead to miscoding.

If you think ADL accuracy issues are old hat, consider this: Undercoding a resident's ADL index by a single point can leave a couple of grand in RUG money on the table. And that's enough to pay a half-time RN's salary for a month in some locations.

Do the math: Suppose a rehab resident has an ADL index of 6 and RUGs into RVA. But if that person had a score of 7, he would have gone into RVL instead, says Sheri Kennedy, RN, MSEd, principal of Knowledge Solutions in Denton, TX. In this scenario, miscoding the total ADL score by one point leads to a loss of $2,143.20 for 30 days or $71.44 a day, says Kennedy, who presented "$ection G: ADL Coding Accuracy" at the spring 2008 American Association of Nurse Assessment Coordinators meeting.

To keep your ADL scores on track and your facility in the black, Kennedy and other experts suggest targeting these common problems:

1. Using ADL flow sheets that capture the highest level but not the frequency of help provided during a shift for self-performance. Kennedy noted in her AANAC presentation that numerous facilities aren't asking the right questions and collecting the right data to code ADLs. "Many of them are asking the CNAs and nurses to give them a code or score of limited, extensive, etc., for the shift. Yet the RAI manual doesn't ask how many shifts a resident received help -- it asks how many times during the lookback period did the person get a specific level of help."

ADL labors lost: Cindy Fronning, RNC, CDONA, RAC-MT, recently did an audit of 44 residents' MDSs and found that the facility had lost $27,000 over three months due to undercoding in G1. The problem? Staff wasn't capturing the frequency of weight-bearing support in order to code extensive assistance, says Fronning, director of clinical reimbursement for Pathway Health Services in White Bear Lake, MN. (To review the definitions for ADL self-performance, see p. 79.)

Solutions: Facilities can implement paper or computerized systems that capture the actual number of times they provided various levels of ADL assistance. Or they can use an ADL exceptions reporting process, Kennedy suggests. Using that approach, CNAs' assignment sheets identify the resident's level of ADL help on the last MDS assessment for the four late-loss ADLs. The CNAs report exceptions to the nurse, including the level and frequency of assistance for the shift, Kennedy tells Eli. The nurse then charts anything more or less than the baseline. That documentation is available for the MDS coordinator to code ADLs on the MDS, Kennedy explains.

Overcome this hurdle: "Some people will say that they have six CNAs reporting to one charge nurse who doesn't have time to hear all of those people report," Kennedy observes. In that type of situation, she suggests elevating two of the CNAs to lead CNAs and having the other CNAs report to them. The lead CNAs can, in turn, report to the charge nurse, Kennedy says.

Example: A resident was coded as limited assistance on the last MDS for an ADL. Thus, if the CNA had to provide weight-bearing assistance, that would be an exception. The CNA would also report how many times she provided that level of assistance during the shift -- for example, she had to lift the resident off the toilet twice.

The documentation might also explain the assessed reason for the exception, such as an acute illness. "This strategy also gives you a way to easily identify a change in ADL status," Kennedy notes.

2. Lack of understanding about the components of an ADL. When people assess a resident's ADLs, they sometimes don't consider all of the aspects of an ADL described in the RAI user's manual, notes Susan LaBelle, RN, MSN, a consultant with LTCQ Inc. in Lexington, MA. Toileting, which includes several components, is a key example (for details, see the article on p. 82).

Don't forget: Eating does include intake of nourishment from tube feeding and/or total parenteral nutrition but not eating/drinking during med pass.

3. Assessing and coding residents as being independent when they really require supervision. That's where Kennedy sees facilities go wrong most often across ADLs. Staff provides cueing and encouragement to help a resident perform an ADL, but then fail to document and code that assistance. For example, this may occur in the dining room where CNAs are actually providing this level of ADL assistance to individual residents who are coded as independent.

Watch out: If the resident falls trying to transfer independently when he really requires supervision in the form of oversight, encouragement or cueing to transfer safely, the facility has a potential liability issue, warns Marty Pachciarz, RN, RAC-CT, a consultant with the Polaris Group in Tampa, FL.

"The same is true for transferring on and off the toilet," she adds, although that's only one subtask of the toilet use ADL. "So it's harder to simply look at the coding to see if someone is getting [supervision] for toilet transfers when he's at risk for falling." For example, the supervision could be for hygiene or readjusting clothing, Pachciarz says.

If a resident coded as independent in eating really requires supervision to eat enough -- and he loses weight -- expect surveyors to be on the facility's case.

Anything that the resident requires, including a certain level of ADL assistance, that the team doesn't identify, care plan or provide becomes a liability issue, especially if the resident suffers a negative outcome as a result, cautions Pachciarz.

4. Missing a two-person assist for bed mobility, toileting or transfer. For example, if the facility codes a person as totally dependent in bed mobility but requiring only a one-person assist to help him slide up in bed, that's probably not true, commented Christie Teigland, PhD, in a presentation at the March 2008 American Medical Directors Association annual meeting.

Remember: Code a "3" for ADL support if staff provided a two or more persons assist even once in the lookback.

5. Purposeful undercoding to counteract fear of overcoding. Kennedy finds that MDS coordinators tend to undercode ADLs because they want to avoid the possibility of being accused of fraud for overcoding. "The irony is that it's just as fraudulent to undercode as to overcode. Fraud is when you do something intentionally expecting a gain or rewards or to get something out of it," Kennedy says.

Take the coding cure: Simply code ADLs accurately, which requires collecting the required data 24/7 for the lookback period, Kennedy says. See the sample ADL flow sheet, p. 83.