A change in the visual analog scale score represents a relative change in the magnitude of pain sensation. Use of the VAS in comparative analgesic trials can now meaningfully quantify differences in potency and efficacy.

To the Editor:-I have spent some time considering the conclusions reached by Aubrun et al. 1 in their study investigating the relationship between the pain visual analogue scale (VAS) and morphine requirements in the postanesthesia care unit. They demonstrated that the Hill equation could be used to derive a sigmoid relationship between the pain VAS and morphine dosage, and concluded that this relationship is not linear. In their discussion, they referred to an earlier study in which I and colleagues had found a linear relationship between pain intensity and the pain VAS, 2 and they seemed to suggest that their study contradicted our earlier results. The results of the Aubrun et al. study could have been readily predicted from their methodology, in that the Hill equation is a standard approach used to generate a sigmoid curve using population pharmacokinetics. This approach generates a typical dose-response curve. Such a curve could be derived for the dose of a hypotensive drug and blood pressure, or the dose of a sedative drug and hypnotic state. Yet, the effect of interest-be it blood pressure, hypnotic state, or pain intensity-can still be linear phenomena. The results of the study data analyzed by the authors do not preclude the possibility that the pain VAS has ratio scale properties and is linear. In fact, if the pain VAS is linear then it would have a sigmoid relationship with morphine dosage. I have one other concern regarding the conclusions drawn by the authors. In their Methods, they arbitrarily defined "severe pain" as when the patient received a total dose of intravenous morphine greater than 0.15 mg/kg. Because they found an association between an initial VAS score of 70 or greater (VAS 70) and morphine dosage greater than 0.15 mg/kg, they concluded that VAS 70 was indicative of severe pain. But VAS 70 itself may not signify severe pain. Opioid dose-response has substantial interindividual variability, with a five-to tenfold range in requirements. Patients may fail to respond to a specific dose not because they have greater pain intensity but because they have had a lesser response to that dose.