Additional technical information about resident-level risk adjustment.

September 20, 2009 by admin

The following information is provided for individuals who are seeking a more indepth understanding of how resident-level risk adjustment is applied to the pain and delirium quality measures that are presented on Nursing Home Compare. Resident-level Risk Adjustment for the Pain Quality Measure: The use of impairment in decision making as a risk adjustor (covariate) is a proxy for cognitive impairment. People who are cognitively impaired may have more difficulty expressing their pain, therefore a nursing home with a lot of cognitively impaired residents may report a lower rate of pain than is actually present in the nursing home. Therefore, if a facility admits a lot of residents with cognitive impairment, their reported Quality Measure score for pain actually goes up (is worse) than would be if we just reported the numerator divided by denominator. For example, if a nursing home admits a lot of residents with cognitive impairment, their MDS assessments may indicate very little pain. However, because we know a lot of residents are cognitively impaired, we assume that they are under-estimating the level of pain because their cognitively impaired residents can't communicate their level of pain to staff. We "adjust" the reported rate of pain to take this "under-estimating" by the facility into account, resulting in a higher reported rate (level of pain looks worse). Resident-level Risk Adjustment for the Delirium Quality Measure: If a resident is newly admitted who has a previous residential history (e.g. they were recently in this or another nursing home), it is more likely that he or she will develop delirium. Therefore, this risk adjustor (covariate) gives the nursing home "credit" for admitting a lot of residents with a prior residential history (who are more susceptible to suffering from delirium). For example, if a nursing home admits a large percent of residents with a prior residential history, the nursing home will probably look worse (i.e. have a high QM rate) if you just divided the numerator by the denominator. However, because they admit a lot of residents who are at higher risk for developing delirium (prior residential history), their reported QM rate is actually lower than it would be without the adjustment, making the nursing home look better. Additional information about how the data for the quality measures is compiled and calculated can be viewed on the Nursing Home Quality Intiative page on www.cms.hhs.gov.