The pain experience of cognitively impaired nursing home residents: Perceptions of family members and certified nursing assistants

Abstract

Pain in cognitively impaired nursing home (NH) elders is difficult to detect. We report the results of the qualitative interview portion of a larger study that characterized the pain experience of cognitively impaired NH residents. Interviews were conducted with 16 family members or friends and 11 certified nursing assistants (CNAs) of 20 cognitively impaired NH residents experiencing pain. Analysis of the interviews yielded themes in family and CNA perceptions of pain in cognitively impaired NH residents. Family members and CNAs concurred that knowing the resident’s usual behavior and daily habits was essential to being able to detect pain in cognitively impaired NH residents. Although a majority of family members reported that their relative was “stoic” about pain expression prior to NH placement, personal care information and historical information were perceived as helpful by nursing staff members. CNAs used facial and eye cues to detect pain and pain relief in their assigned residents. In addition, CNAs reported specific pain management strategies for their residents. Major themes related to pain detection in cognitively impaired NH residents identified in this investigation include: (a) knowing the resident; (b) importance of family input about previous pain behaviors in knowing the resident; (c) CNA reliance on face and eye cues for pain detection, particularly with residents who were nonverbal; and (d) the prevalence of pain with caregiving activities. Information from multiple sources can improve pain management strategies for cognitively impaired NH residents.

Introduction

Pain remains a persistent problem in nursing home (NH) residents, with some estimates of pain prevalence as high as 80% (AGS, 2002). Residents with mild to severe cognitive impairment comprise approximately 50% of NH residents, and the wide variability of individual abilities complicates a standardized approach to pain assessment for persons with cognitive impairments. As an example, some residents with cognitive impairment can no longer communicate verbally about their pain, some give ambiguous reports of pain, and others can verbally report pain with the use of one of several pain assessment tools Feldt 2000b, Herr 2002.

In attempts to assess pain in NH residents with cognitive impairment, research has been conducted with proxy reports of pain from nurses or certified nursing assistants (CNAs) Cohen-Mansfield and Creedon 2002, Fischer et al 2002, Kovach et al 2000 and from family members Cohen-Mansfield 2002, Weiner et al 1999. Kovach et al. interviewed 30 nurses who identified signs or symptoms of pain in cognitively impaired NH residents. Facial grimacing and restless body movements were the most frequently identified signs of pain. Cohen-Mansfield and Creedon further confirm that facial expressions are a key indicator of pain in cognitively impaired NH residents. They reported that 75% of 29 NH staff members interviewed (i.e., registered nurses, licensed practical nurses, and certified nursing assistants) identified facial grimaces as an indicator of pain in residents with dementia. In a study of CNA proxy pain reports, Fischer et al. reported that CNA pain assessments of NH residents using a simple, 3-question proxy pain questionnaire were more strongly associated with the residents’ analgesic medication use than the Minimum Data Set (MDS) pain items. The CNAs’ assessments also generated higher estimates of pain prevalence than the MDS items did (48% versus 20%).

Family member proxy pain reports for cognitively impaired NH residents have not been systematically investigated. In a study of chronic pain in NH residents, Weiner et al. (1999) found that relatives’ and nursing staff’s ratings of the intensity of pain of NH residents were poor. Seventy-one percent of the nurses in this study reported that pain assessments are more difficult in residents with dementia than in residents without dementia. Cohen-Mansfield (2002) found that family members were less able to rate their relative’s pain when the resident was significantly cognitively impaired or when the relative had a longer stay in the NH. However, when relatives visited at least once a week, their pain ratings were significantly correlated with MDS scores, as well as with residents’, physicians’, and nursing staff’s ratings.

Controversy exists over whether the MDS pain indicators and physicians’ and nurses’ pain frequency ratings are accurate, particularly with cognitively impaired residents (Fischer et al., 2002). Most often, MDS and provider determinations of pain underestimate the prevalence of pain in elders (Cadogan, Schnelle, Yamamoto-Mitani, Cabrera, & Simmons, in press). Determining intensity is even more difficult to accomplish in this population. Further, most pain instruments designed specifically for cognitively impaired elders have been normed on cognitively impaired hospitalized elders Decker and Perry 2003, Feldt 2000a and not for frailer cognitively impaired NH residents. Therefore, the purpose of this portion of a larger investigation was to gather and evaluate whether information from family members and friends about a patient’s lifelong pain behavior and expression improves pain detection in cognitively impaired residents and to evaluate pain information from formal direct caregivers (i.e., CNAs) who cared for the relatives of these family members.