Pain behaviour and pain intensity in older persons with severe dementia: reliability of the MOBID Pain Scale by video uptake.

Advancing age is associated with high prevalence of dementia, often combined with under-diagnosed and under-treated pain. A nurse-administered assessment tool has been developed to unmask pain during standardised, guided movements, called Mobilisation-Observation-Behaviour-Intensity-Dementia (MOBID) Pain Scale. The aim was to examine intra- and inter-rater reliability of pain behaviour indicators, inferred pain intensity, and the overall MOBID Pain Score. Twenty-six nursing home patients with severe dementia and chronic pain, 11 primary caregivers and three external raters at the Red Cross Nursing Home, Bergen were included. During video uptake the patients were guided by their primary caregivers to standardised movements of different body parts. Pain behaviour indicators (pain noises, facial expression and defence) were registered for each movement with subsequent rating of pain intensity by external raters, who assessed and scored the videos concurrently and independently at day 1, 4 and 8. Facial expression was most commonly observed, followed by pain noises and defence. Repeated assessments increased the number of observed pain behaviours, but did not improve reliability. Inter-rater reliability was highest for noises, followed by defence and facial expression (kappa = 0.44-0.92, kappa = 0.10-0.76 and kappa = 0.05-0.76 respectively, at day 8). Mobilisation of arms and legs were rated most painful. Intra- and inter-rater reliability of overall pain were very good [intraclass correlation coefficient (1,1) ranging 0.92-0.97 and 0.94-0.96 respectively, at day 8]. Reliability of pain intensity scores tended to increase by repeated assessment. Using video uptake, MOBID Pain Scale was shown to be sufficiently reliable to assess pain in older persons with severe dementia.