Miranda DiLorenzo1, David Flora1, Rebecca Pillai Riddell1
1) Canada
Background/Aims: Measures of facial activity, such as NFCS,1 have been considered to be the most reliable and consistent methods by which to measure infant pain. However, to ensure accurate assessment of acute pain in infants, the construct validity of NFCS will be evaluated.
Methods: The data is part of an ongoing longitudinal study that followed caregivers and children from infancy to preschool. The facial expressions of infants were coded using 7 facial action units on the NFCS. The factor structure of NFCS immediately after the needle and 1-minute post-needle was estimated using data from the 12-month vaccination (n=546).
Results: A scree plot indicated that a two-factor model would adequately fit the data. The two-factor model accounted for 91% of the variance in item scores immediately after the needle and had a RMSEA of .14, indicating adequate model fit. Similarly, the two-factor model accounted for 94% of the variance in item scores 1-minute post-needle with a RMSEA of .14. At both time points, the first factor included 4 items with high factor loadings (>.93). Items were all mouth facial action units (e.g. open mouth). The second factor included 3 items with moderate to high factor loadings (>.56). Items were generally facial actions around the eyes (e.g. eye squeeze). The correlation between factors was .73 immediately after the needle and .75 at 1-minute post-needle. Despite the high correlation between the two factors, they were not represented by a higher-order construct.
Conclusions: The two factor solution indicates that the presence of the different facial actions may not reflect a single underlying construct of pain. To ensure valid measurement of infant pain, a re-examination of different summative procedures appears warranted.
Citations:
1. Grunau, R.V.E. and Craig, K.D. (1987). Pain expression in neonates: facial action and cry. Pain, 28.
Acknowledgements/Disclosures: Funding provided by the Canadian Institutes for Health Research (524432).
