Perspective on the paper by Bellù and Milesi et al (see pages 241 and 272).

Pain, stress, distress and discomfort. All these words relate to the subjective experience of intensive care for adults and children. Babies, whether term or preterm, undergo experiences that these words might well describe, but unfortunately they cannot tell us about them. This does not make studies of pain in babies undergoing intensive care any less important, it just makes them more difficult. If there has been one important development in humanising neonatal care in the last two decades, it has been the recognition that pain matters; this realisation has resulted in a great deal of research effort to create a scientific underpinning for improved clinical practice. In this issue, Bellù et al have undertaken a meta-analysis of the use of opioids in neonates receiving mechanical ventilation,1 and Milesi et al report on the validation of a new neonatal pain scale that does not rely on facial expression.2 Where do these two papers take us?

There are two main difficulties with clinical studies of pain in babies. The first is trying to find a valid surrogate measure of pain intensity, since asking babies to rate their pain is not possible. The second is to deal with the very different kinds of pain (or distress, or discomfort) that babies have to contend with during their intensive care experience.

Let's take the validity first. This is simple: since no rating of pain intensity can ever be measured against self-report, no scale of pain in the neonate can be validated, or calibrated, against the gold standard of subjective experience. …