Assessment of pain in advanced dementia. Construct validity of the German PAINAD

The observation scale PAINAD (pain assessment in advanced dementia) is composed of five behavioral categories: breathing, vocalization, facial expression, body language, and consolability. The present study investigates the construct validity of the German version. We conducted a prospective one-dimensional observation study with repeated measurements (t(1)=pretreatment, t(2)=2 h posttreatment, t(3)=24 h posttreatment). The sample consisted of 12 verbally noncommunicative demented inpatients with severe comorbidity treated in three geriatric clinics. Their age was M=84.3 years (SD=4.4) on the average. Ten of them were female. Inclusion criteria were pain-related physical illness and observed pain behavior. Every patient was treated with analgesics after t(1). After t(2) the medication was discontinued in five patients until t(3). Nurses documented the PAINAD scores after an observation period of 2 min during routine care. Pain behavior at t(2) diminished considerably displaying a large effect size. Scores continued to be low at t(3) only in the sample with continued medication. Scores in the other part of the sample returned to initial values. The data demonstrate that pain medication strongly impacts the pain behavior of demented patients. The outcome supports the assumption that PAINAD really measures pain.