Objective: The specific objective for this research was to determine initial psychometric properties of the Faces Pain Scale (FPS) as a measure of pain intensity for use with the elderly.
Design: The study was descriptive correlational in nature, with nonrandom sampling. A total sample of 168 community subjects (30-121, depending on task completed), aged 65 or older, participated in the research protocol. To determine the validity, reliability, and scaling properties of the FPS, rating and ranking procedures, placement tasks, and test-retest methods were used.
Results: Response to six Likert-type items indicated that subjects agreed that the FPS represents pain: however, it is clear that the perception of the meaning of the faces can be influenced by the context in which they are presented. Rank ordering tasks for the individual faces demonstrated near-perfect agreement between the actual expected ranking and the ranking produced by the subjects (Kendall's W = .97, p = .00). When subjects placed individual faces along a 1-m-long red wedge indicating the amount of pain represented by each face, statistically significant separation of the faces in the anticipated equal interval position was demonstrated by the lack of overlap of the 95% confidence intervals when all faces were viewed and positioned simultaneously. However, when subjects placed faces independent of others, the expected placement fell outside the 95% confidence limit for three of the five faces placed. In addition, the actual intervals between the five faces placed by subjects demonstrated substantial variances from the 167 mm expected in several instances. Rating a vividly remembered painful experience about the degree of pain perceived using the FPS initially and again 2 weeks later, the FPS demonstrated strong reproducibility over time with a Spearman rho correlation coefficient of .94 (p = .01).
Conclusion: These results provide preliminary support for the construct validity, strong ordinal properties, and strong test-retest reliability of the FPS with a sample of elderly individuals. The equality of intervals in the FPS has not been fully supported in the older adult, but given the complexity of the task used, the results should not be considered to be refuted. Further evaluation of the FPS with experimental and clinical pain conditions and comparison with other standard pain assessment instruments in the elderly population are warranted.