[Respiratory and circulatory parameters as indicators of the postoperative analgesic demand in newborns and infants]

Due to immature cognitive functions, infants are unable to communicate their pain perception verbally. To assess postoperative analgesic demand, the anaesthetist has to rely on observational techniques. Generally, pain expression is considered to be a multidimensional phenomenon consisting of physiological, motor-reflex and behavioural patterns. The majority of observational approaches to pain assessment in infants use the behavioural dimension only, regardless of the fact that pain perception might contribute substantially to the stress response. The aim of this study is to evaluate, whether sensitivity and specificity of a behavioural pain scale (CHIPPS [1]) can be improved by adding physiological measures, especially those representing the stress response.

30 infants aged 0-12 months and scheduled for unilateral herniorrhaphy were studied prospectively. In addition to 9 behavioural items (crying, facial expression, wrinkling of the forehead, motoric restlessness; posture of fingers, arms, legs, toes and torso) the ratio of actual physiological measurements (heart rate, respiratory rate, blood pressure) and their respective preoperative baseline values were recorded by a single observer in 5 minutes intervals during the first hour after recovery from anaesthesia. Maximal efforts were made to achieve valid measures. Factor analysis was performed to determine the dimensionality of the complete item pool. For additional validity testing, receiver operating characteristic curves (ROC) were calculated using the independent opinion of an experienced clinician as an external criterion. Discriminant analysis was performed to assess the accuracy of a combined behavioural and physiological scale.

The factor analysis resulted in two independent dimensions: behaviour and cardiocirculatory measurements. The strong intercorrelations of all behavioural items and the fact, that the affective pain experience is expressed by a specific mimic behaviour, suggest the behavioural dimension to be regarded as pain expression. Because of the strictly orthogonal structure of the factor system, the circulatory and the respiratory dimension lack any relationship to pain experience. In addition to these statistical reasons, considerations on practicability disprove blood pressure and respiratory rate as useful pain indicators: Whereas the observer never failed to obtain a behavioural score, only 60% of the blood pressure measures and 80% of the respiratory rates were valid. In contrast, heart rate counts were obtained in over 99% and thus have to be considered as the only practicable physiological measurement in the early postoperative period. Corresponding to the results of the factor analysis, ROC curves suggest that the ability of the heart rate alone to assess pain is not substantially better than a random process, whereas the behavioural scale performs well. In addition the heart rate failed to improve the accuracy of the behavioural scale as shown by the results of a discriminant analysis.

Despite the multidimensional approach and the corresponding multivariate analyses, a unidimensional scale consisting of behavioural items was found to be a valid indicator of an postoperative analgesic demand. Due to the lack of diagnostic properties and difficulties to obtain sound values even under research conditions, physiological measurements like blood pressure, respiratory rate and heart rate are not suitable for the assessment of a postoperative analgesic demand in infants, neither for clinical nor for research purposes.