Pain is a subjective feeling; its assessment is therefore difficult, and no "gold standard" method exists for humans. Major improvements have, however, been made in the last decade by widespread acceptation of the concept of pain evaluation and widespread use on surgical wards. Evaluation by the patient himself is the rule (unless communication is impaired), as assessment of pain by nurses or doctors systematically leads to underestimation (which also occurs with observational scales). Theoretically, pain should be evaluated in its multiple dimensions such as intensity, location, emotional consequences and semiologic correlates. Scales which have been developed to evaluate these dimensions are, however, too complex for widespread and repetitive use in surgical patients. The Mac Gill Pain Questionnaire is therefore only used in the surgical setting for research purposes. Moreover, its scientific accuracy, although often accepted, is poor and in our opinion cannot be accepted as a reference method. Only methods assessing pain intensity can be used in the clinical setting because of their simplicity. The verbal rating scale (VRS), the numerical rating scale (NRS) and the visual analogue scale (VAS) are preferred by an increasing number of groups. Although scientific validation is difficult, VAS seems the most accurate and reproducible scale. Postoperative pain should be assessed several time!; a day in every patient, at rest and in dynamic conditions (cough, movement) and should focus on present pain rather than on pain in the previous hours. Assessment of pain is essential before quality-assurance programmes can be implemented. (C) 1998 Elsevier, Paris.
