Long-Term Care Survey Alert

ASSESSMENT:

Predictive Pain Model Flags Residents at Risk for Having Unrelieved Pain

If a resident coded without pain has these conditions, take a closer look.

To identify residents who may need more in-depth pain assessment or treatment, compare the pain coding in MDS Section J to certain diagnoses and conditions on the MDS.

That's what an MDS-driven computerized program provided by the New York Association of Homes & Services for the Aging (NYAHSA) does, although anyone can use the data to perform the double check manually, reported Christie Teigland, PhD, director of Health Informatics and Research at NYAHSA, in a presentation at the recent American Association of Homes & Services for the Aging annual meeting.

Home in on These Conditions and Factors

NYAHSA researchers identified conditions and factors associated with pain in a cognitively intact population of residents who could report they were having pain. The diagnoses and conditions include arthritis, hip fracture, other fractures, osteoporosis, pathological bone fracture, asthma, allergies, vertigo, infections (urinary tract infection, wound infection, viral hepatitis), edema, fever, vomiting, and a fall.

Also on the list of conditions associated with pain: Stage 2-4 pressure ulcers, stasis ulcers, and other skin conditions (burns, open lesions due to cancer, surgical wounds, and skin tears). Skin treatments can also cause pain (turning/repositioning, surgical wound care and applying dressings), Teigland noted.

Other conditions indicating that a person may be experiencing pain:

• Having a high body mass index. People admitted to nursing homes are heavier these days, Teigland cautioned conferees at her AAHSA presentation.

• Residents receiving physical therapy (PT).

• Urinary catheters, including an external catheter.

• Stability of conditions indicators in J5, especially end-stage disease.

• A bowel and bladder training program. (Residents on a scheduled toileting program are less likely to have pain.)

Key point: To use the NYAHSA pain predictive model effectively, "you have to capture all the conditions on the MDS," says Naomi Bell, RN, RAC-CT, MDS coordinator at one of the facilities using the NYAHSA pain prediction model.

Expanded Application of Model in the Works

Currently, nursing facilities are using the pain model for people with moderately severe to very severe cognitive impairment, says Teigland. "But the risk factors apply to everyone."

Thus, if a person has the risk factor(s), he should be assessed for pain. NYAHSA plans to extend the predictive tool to the entire population in the nursing home, in fact. Very few cognitively intact ndividuals with a zero on the MDS 2.0 Cognitive Performance Scale who have the pain risk factors aren't coded as having pain on the MDS, Teigland says. If those individuals don't have pain coded, it's because they are receiving effective pain management, she adds. However, the extended model could help identify individuals with mild cognitive impairment whom the facility isn't effectively assessing and managing for pain, Teigland notes.

Outcomes Show Impact of Better Pain Detection, Rx

Using a data-driven pain predictive model can be an eye opener. In one case, a resident who couldn't communicate was always crying and rubbing her leg, reports Teigland.

The resident had been receiving an antidepressant but no pain medication. When nursing staff assessed the resident more closely, however, they found she had diagnoses of a left cerebral vascular accident with peripheral vascular disease and neuropathic pain.

Once the resident received pain medication, the behaviors stopped, Teigland tells Eli.

In another case, staff used the predictive model to identify that a "frequent faller" had rheumatoid arthritis with PRN pain medications, which the nurse said translated into "patient receives none."

Once treated for pain, the resident had no falls for six weeks, Teigland says. In fact, "the facility saw a 26 percent reduction in falls over a six month period due to better assessment and treatment of pain in residents with dementia." The facility staff reported that the model identified 52 percent of residents who fell had undetected pain.

Model Results in More Accurate QIs/QMs, More Pain Remedies

One facility using the MDS-driven pain prediction model saw an increase in its QIs/QMs at first, including the pain QMs, because staff members were doing a more comprehensive assessment. But the facility also began providing more pain medication, as well as non-pharmacological remedies, reported its administrator Brian Ruede who presented in the AAHSA session.

"The resident may receive hand massages and aromatherapy," for example, Ruede says.

For people with dementia who are having behavioral symptoms that might be due to pain or "sundowning, etc., we have a Snoezelen room with soft lighting and music," Ruede tells Eli. (Snoezelen is a "multi-sensory environment" that you can use to soothe or stimulate people. For more information and pictures of the modality, go to www.snoezeleninfo.com/main.asp.)

Editor's note: To view sample reports generated by the computerized pain prediction model, see pages 6 and 7 of this issue. Also see page 7 for an article on Reiki as a nonpharmacological pain management and spiritual modality.