Analysis of behavioural and physiological parameters for the assessment of postoperative analgesic demand in newborns, infants and young children: a comprehensive report on seven consecutive studies.

Abstract

Many different systems for the assessment of pain in newborns and infants have been tested for validity, rarely for reliability but never for sensitivity or specificity. We aimed to determine whether the assessment of an analgesic demand in the lower age group during the postoperative period is possible by observational methods only. In an repetitive and sequential prospective process for identifying observationable behaviour and measurable physiological parameters as indicators of a postoperative analgesic demand, 584 newborns, infants and young children were studied (7 prospective studies, 4238 observations). Twenty-six items were selected as suggested by current literature and for reasons of economy and practicability. The factor analyses resulted in a two-factorial solution with the behavioural items loading on one factor and the physiological parameters on the other (principal component analyses). The physiological parameters blood pressure, respiratory rate and heart rate were found to be unreliable and had no discriminant power to detect an analgesic demand during the postoperative period (discriminant analyses, ROC-curves). In newborns and infants, nine observational items were identified as equally selective, reliable, sensitive and specific to the assessment of postoperative analgesic demand, whereas in young children only five items could be identified (discriminant analyses, ROC-curves). For economic reasons, these five items (crying, facial expression, posture of the trunk, posture of the legs, motor restlessness) were chosen as the basis of an additional pain scale ranging from 0=no pain to 10=maximal (Children's and Infants' Postoperative Pain Scale, CHIPPS). Its internal consistency yielded values for Cronbachs' alpha with 0.92 for toddlers and 0.96 for infants. The coefficient for interrater reliability was 0.93. The scale was validated constructively by the intravenous administration of metamizol, tramadol, nalbuphine, piritramide and ketamine (repeated measures analysis of variance). The Toddler-Preschooler Postoperative Pain Scale and CHIPPS equally identified painfree situations or analgesic demand in 87.4%. In cases with definite pain, the score of CHIPPS was never below 4 points. Seventy-one toddlers gave verbal comments on their pain intensity: in 29 painfree situations the CHIPPS score was 3.0 and in 29 painful situations it was 5.7. The values for sensitivity and specificity of CHIPPS were calculated to be 0.92-0.96 and 0.74-0.95, respectively (discriminant analyses). We conclude that it is possible to determine postoperative analgesic demand in the low age group of children by using an observational system such as CHIPPS alone.