Interrater reliability of the Pediatric Infant Parent Exam: Nursing screening as a component of well-baby visits
The purpose of the dissertation study was to determine the interrater reliability, overall impression, and scoring of typicality of the interaction and time required for nurses' use of the Pediatric Infant Parent Exam (PIPE) to screen the interaction of a convenience sample of 50 mother-infant dyads during a well-baby visit in a clinic setting. After IRB consent was obtained, the PI and three pediatric clinical nurses completed training based on the PIPE manual. The research assistant identified English speaking mother-infant dyads using the pediatric clinic as a medical home. Infants were between six and nine months of age. Exclusion criteria were infants with congenital anomalies, fever or illness. On reporting for the scheduled well-baby visits, Mothers who expressed an interest received further information about the study. Once they expressed an interest in participation, informed consent was obtained and the one time brief observations were carried out in the private exam room by the PI and pediatric clinical nurse. A book was presented to each dyad who participated. Raw scores were transformed with SAS, version 9.3 and analyzed using Cohen's kappa. The kappa was 0.809 and the weighted kappa was 0.840 for cumulative paired ratings of the various stages of the PIPE while the "overall synchrony" ratings did not meet the assumption of symmetric distribution making any kappa analysis questionable at best. As to the "judgement of typicality", kappa was 0.666. Two of the pediatric clinical nurses rated use of the Pediatric Infant Parent Exam as "somewhat easy" while one nurse rated the instrument as "easy to use". The time for scoring the observation ranged from 60 to 202 seconds with a mean of 90.57 seconds, a median of 67.5 seconds, and a standard deviation of 44.13 seconds.