[Retrospective assessment of the Dutch version of the Discomfort Scale--Dementia of Alzheimer Type (DS-DAT): is estimation sufficiently valid and reliable?]. / Retrospectieve afname van de Nederlandse versie van de Discomfort Scale--Dementia of Alzheimer type (DS-DAT): is inschatten achteraf voldoende valide en betrouwbaar?

The Discomfort Scale--Dementia of Alzheimer Type (DS-DAT) measures discomfort in severely demented patients with scores on nine items with behavioral descriptors. Direct observation of behaviour is the preferred method in severely demented patients, but is not feasible for some types of research. Alternatively, a patient's score may be assessed 'retrospectively', scoring an overall picture of the patient. To assess validity and reliability of such retrospective assessments, five observers--three nursing home physicians and two paramedicals--gave a DS-DAT score for 77 nursing home patients by direct observation, and, two weeks later, retrospectively. The mean score of the five observers was not different. The Intra-class Correlation Coefficient for intra-observer reliability was 0.50 for the five observers, and 0.55 for the three nursing home physicians. Our study demonstrated the possibility of a reasonably valid, but moderately reliable retrospective assessment of the Dutch version of the DS-DAT. Appropriate training of nursing home physicians who know their patients well may be required for this.