Pain and distress caused by endotracheal suctioning in neonates is better quantified by behavioural than physiological items: a comparison based on item response theory modelling

Pain cannot be directly measured in neonates. Therefore, scores based on indirect behavioural signals such as crying, or physiological signs such as blood pressure, are used to quantify neonatal pain both in clinical practice and in clinical studies. The aim of this study was to determine which of the physiological and behavioural items of 2 validated pain assessment scales (COMFORT and premature infant pain profile) are best able to detect pain during endotracheal and nasal suctioning in ventilated newborns. We analysed a total of 516 PIPP and COMFORT scores from 118 newborns. A graded response model was built to describe the data and item information was calculated for each of the behavioural and physiological items. We found that the graded response model was able to well describe the data, as judged by agreement between the observed data and model simulations. Furthermore, a good agreement was found between the pain estimated by the graded response model and the investigator-assessed visual analogue scale scores (Spearman rho correlation coefficient = 0.80). The information scores for the behavioural items ranged from 1.4 to 27.2 and from 0.0282 to 0.131 for physiological items. In these data with mild to moderate pain levels, behavioural items were vastly more informative of pain and distress than were physiological items. The items that were the most informative of pain are COMFORT items "calmness/agitation," "alertness," and "facial tension."