The adequacy of the minimum data set assessment of pain in cognitively impaired nursing home residents

This paper examines the validity of the Minimum Data Set (MDS), a comprehensive functional assessment used in most United States nursing homes, for recognizing pain in cognitively impaired nursing home residents. Eighty nursing home residents participated in the study. They were initially divided into four groups, along two dimensions: severe vs. mild/moderate cognitive impairment, and pain-medicated vs. non-medicated for pain. Three indicators of pain were derived from the MDS: pain frequency, pain intensity, and number of sites with pain. Geriatricians from outside the nursing home assessed the medical condition of the residents and their associated pain levels. Residents were asked to report on their level of pain. The three MDS indicators were highly intercorrelated. According to these indicators, 34-39% of the residents suffered from pain. Some correlations between self-report of pain and MDS ratings were significant among those with mild/moderate impairment, but not for those with severe impairment. Similarly, some correlations between MDS ratings and the geriatricians' ratings were significant for those with mild/moderate impairment but not for those with severe impairment. Persons with mild/moderate cognitive impairment were rated by the MDS as having more pain than those with severe impairment. The findings reveal that the MDS underreports pain in cognitively impaired residents. Given the centrality of the MDS in the U.S. as a tool for clinical decision-making in long-term care, there seems to be an urgent need to improve either the MDS pain assessment tools or the procedures used for completing them.