Self- and proxy report for the assessment of pain in patients with and without cognitive impairment : Experiences gained in a geriatric hospital.

BACKGROUND: Pain assessment is a complex procedure in patients with different degrees of cognitive impairment. The challenge is to determine whether self-reporting tools are adequate and to identify the cases in which proxy ratings are necessary. PATIENTS AND METHODS: As part of an open, prospective observational study, 178 patients underwent a comprehensive pain assessment consisting of the Numeric Rating Scale (NRS), the Verbal Rating Scale with four and five items (VRS4 and VRS5) and the Pain Assessment in Advanced Dementia Scale (PAINAD). RESULTS: Even without prior knowledge of a patient's cognitive impairment, this toolkit can be used to reliably identify cases where self-rating is appropriate and where proxy rating becomes necessary. Inter-rater reliability: "good" agreement [Cohen's κ = 74.2% (p < 0.001) (95%CI, 54.6-93.8%)], test-retest reliability: "moderate" agreement [κ = 55.3% (p < 0.001) (95%CI, 28.5-82.1%)]. Furthermore, movement resulted in a higher correlation between the selected assessments. Self-report assessments are appropriate up to a mini-mental state examination (MMSE) value greater than ten. In comparison to NRS, VRS4 and VRS5 remain more stable as the degree of cognitive impairment increases. CONCLUSIONS: In the majority of the cases, our approach enables reliable appropriate pain assessment without the strict prerequisite of an upstream cognitive assessment.