Preliminary Work: Initial testing for reliability and validation of the CHIIP scale was performed. Seven infants (28 to 38 weeks adjusted gestation) were videotaped during painless or painful episodes. Twenty NICU nurses independently rated the videos. Reliability was estimated with Cronbach's alpha (r = .80) and the interclass correlation coefficient (r = .99). Construct validity was investigated by comparing change scores for infants in the NICU receiving pain medication. The clinical records of 20 infants were examined for pain scores prior to pain medication and one hour after pain medication. The average pain score decreased by 2 ¨ö points (p< .001). This poster presents additional construct validity data for the CHIIP scale.

Methods: The CHIIP Scale's ability to discriminate between pain in infants before and after pain medication is an indication of construct validity. Subjects are infants in the NICU who are receiving medication for pain. The infants are videotaped for one minute immediately prior to pain medication delivery and videotaped one hour after the medication. Twenty babies from each age group (< 28, 28 to 36, and °Ã 37 weeks adjusted gestation) will be videotaped. The videotapes will be randomized and scored by two trained independent raters. Change scores (from pre pain medication to post pain medication) will be computed for each subject, and each age group will be reported separately. Based on the preliminary work, a sample of 15 per group is sufficient for a power of .9 with alpha set at .01.