Long-Term Care Survey Alert

Survey Management:

DONT LET MDS MISTAKES DRIVE QUALITY SCORES OUT OF BOUNDS

No nursing facility would purposely report its quality outcomes as being more negative than they actually are.

Yet inaccurate or overzealous coding on the minimum data set can have that effect, flagging a facility as having a potential problem where none exists.

Experts share with Eli the most common clinical areas where the MDS not your patient care is the problem.

  • Dehydration. The Centers for Medicare and Medicaid Services views dehydration as a sentinelevent, which it defines as a problem that should occur infrequently, if at all. So even one resident flagged as having dehydration means surveyors will be all over your care plans and clinical protocols.

    Yet "dehydration is probably one of the most prevalent over triggered conditions on the MDS," cautions Beth Klitch, principal, Survey Solutions in Columbus, OH, who spoke at the recent American Health Lawyers' conference in Phoenix.

    To trigger the dehydration QI, the MDS staff must include a diagnosis of dehydration (ICD-9 CM 276.5) in Section I3 or code J1c when the resident has two or more of the following during the look back period:
    -  A fluid intake usually less than the recommended 1,500 ml of fluids daily (including water or liquid beverages and the water in foods like gelatin or soup);.
    -Clinical signs of dehydration; and
    -A recorded fluid loss exceeding his fluid intake (e.g., fluid replacement hasn't kept up with the resident's diarrhea or vomiting).

    To avoid triggering the QI inappropriately, make sure Section I doesn't include an out dated dehydration diagnosis from the hospital stay, for example, that's no longer an active diagnosis, Klitch advises.

    Also keep in mind that CMS has lowered the recommended fluid intake for coding J1c to 1,500 ml a day (rather than the previous 2,500 ml cutoff). The clarification appears in the updated Resident Assessment Instrument user's manual, which went in to effect Jan.1, 2003.

    And before you code that the resident's output exceeded his intake, double check the math or look for any missing cc's. "The busy CNAs may not have included fluids at med pass or as part of supplemental nourishment," notes Jan Stewart, consultant for QUnique Corp. of Carroll Valley, PA, and a master teacher and board member for the American Association of Nurse Assessment Coordinators.

  • Urinary tract infections. Don't code urinary tract infections (UTI) at I2J unless the diagnosis is confirmed by "significant laboratory findings" in the clinical record, advises Cheryl Field, director of clinical and reimbursement services at LTCQ Inc. in Lexington, MA. "The lab findings can include a culture or even a dipstick that shows protein in the urine," she notes. Yet even though you don't code instances where the physician treats based on symptoms or lab results are negative, it's a good idea to track the information through your internal quality assurance and infection control, Field suggests.

  • Fecal Impaction. Since CMS views fecal impaction as a sentinelevent, make sure the resident doesn't really just have a severe case of constipation. The RAI user's manual provides some key ways to distinguish between these two conditions for example, you diagnose an impaction through a digital exam or an abdominal X-ray or CAT scan. Some residents with an impaction will have symptoms, such as a fever, an acute abdomen, nausea and vomiting and a thin, watery stool. Fecal impaction usually requires digital removal.

  • Incontinence.To avoid artificially inflating your incontinence QIs, code the data collection for incontinence over the entire14-day look back period on all three shifts, suggests restorative nurse B.J. Collard, president of CTS Inc. in Denver. "Staff should mark each episode of both bowel and bladder incontinence," Collard suggests. "Then the MDS staff should evaluate the data and request further clarifications from staff before coding the MDS section." Don't forget to note and code incontinent residents who have a toileting plan.

    Restraints. The updated Resident Assessment Instrument user's manual has lightened up on what CMS counts as a restraint for coding at Section P4. For example, normally facilities would code chairs that keep the resident from rising as restraints at P4e. But the updated RAI manual says immobile residents confined in such due to a neuromuscular condition should not be coded as restrained.

    CMS also says the Merry Walker Ambulation Device and similar devices should not be "categorically" classified as a restraint in every case. So if the resident can easily open the front gate and exit the device, don't code the device as a restraint at P4e but rather at G5a as a cane/walker/crutch.

  • Fever. Chronic-care residents checked as having a fever at J1h will show up in the facility's infection measure reported as part of the Nursing Home Quality Initiative. So make sure the resident really has a fever as defined by the RAI user's manual (e.g., when the resident's temperature is 2.4 degrees Fahrenheit higher than his base line temperature).

  • End-stage disease. Several of the publicly reported quality measures are risk adjusted for residents with end-stage disease .Also, pressure ulcers in residents with end-stage disease aren't considered a sentinelevent. So check J5c on the MDS if the resident has any terminal condition with a life expectancy of six months or less. One caveat: The updated RAI manual clarifies that facilities must have physician documentation to that effect in order to code J5c.

    If your facility has trouble obtaining the required documentation, you might ask the physician to provide the statement via a telephone order, suggests Rena Shephard, a nursing consultant and MDS expert with RRS Healthcare Consulting Services in San Diego. Then follow the state requirements for having the physician sign the documentation in the resident's medical record.

     

  • Other Articles in this issue of

    Long-Term Care Survey Alert

    View All