Visual Analog Scale Scores for Labor Pain

To the Editor: I read with interest the article by Ludington and Dexter [1] describing the use of a "total pain score" for analysis of analgesic modalities in labor. The concept of a total pain score for labor should not be interpreted by readers as a valid or even a useful measure of analgesia. Readers should be aware that the total pain score has never been used in contemporary obstetric analgesia trials, has never been validated for this purpose, and may represent a remarkable oversimplification of the labor process. This concept should not have been proposed as a valid technique in a medical intelligence article, but rather presented as a hypothesis that needs further study and verification. To ascribe one single total pain score for an entire labor would present extreme difficulties to researchers seeking to assess subtle but important differences among drugs and/or techniques. Obtaining total pain scores by multiplying retrospective visual analog scale (VAS) measurements by duration of labor has no basis in either physiology or any commonly used current research paradigm. Can you realistically compare squeezing your head tightly in a vice for 10 min to squeezing it lightly for 8 h and assume that the experience is the same because the time-weighted average VAS scores were similar? Moreover, the authors claim that VAS scores are true ratios. They claim that a score of 6 is twice as severe as a score of 3, and that the magnitude of difference between a score of 2 and 3 is the same as between 7 and 8, for example. Although this may be true in a purely mathematical sense only (and the authors should have acknowledged this limitation), it is hardly true in reality; in fact, it even defies common sense. Certainly no one would claim that a baby with an Apgar score of 6 is exactly twice as healthy as one with a score of 3. These are all ordinal, not interval, data. Parametric statistics may be used with robustness to analyze ordinal data, as shown by Dexter and Chestnut [2]. However, the underlying numbers must still be recognized as ordered, not interval. Mathematical analysis notwithstanding, interpreting VAS data during labor as pure linear continuous data simply makes no physiological sense and represents a gross over-simplification of the complex factors associated with measurement of labor pain. To ascribe one single total pain score to an entire labor is equally inappropriate.