Comparing visual-analog and numeric scales for assessing menstrual pain

Comparing visual-analog and numeric scales for assessing menstrual pain – Statistical Data Included

Cristina Larroy

The precise measurement of pain intensity represents one of the most frequent challenges that healthcare professionals have to face. Traditionally, numeric scales (almost always using a total of 6-11 points as ratings) have been used. Patients are asked to quantify their pain by providing a simple general rating, usually from 0 to 5 or 0 to 10, with 0 equal to no pain and 5 or 10 representing the worst pain the patient can imagine. Verbal scales (generally with from 5-7 categories) have also frequently been used (for instance, the famous McGill Pain Questionnaire, which consists basically of lists of adjectives), although they are not used as often as the numeric scales.

In the last 10 years, the visual-analog scale (VAS) has been used more and more frequently. In this method, which is useful for evaluating variations in pain intensity, the patient is instructed to indicate the intensity of his or her pain by marking a 100-mm line with 2 extremes: no pain and worst imaginable pain. The VAS is based on the theory that pain intensity is continuous, without jumps or intervals, and therefore the categorical scales (such as those of the verbal or numeric types) cannot adequately reflect the changes produced in this variable. The increase in the use of the VAS in the last few years has been significant, although from the beginning different authors such as Jensen and Karoly, (1) Jensen and McFarland, (2) and White (3) have pointed out potential problems with its use.

In fact, use of the VAS appears to be mandatory in many pain clinics to the detriment of numeric and verbal scales. (4) The latter, despite their disadvantages, seem more adequate than the visual-analog type when one is working with certain populations (eg, people whose abstract thinking skills and capacity to establish analogies are not highly developed (5)). Furthermore, the numeric and verbal scales are more time and cost efficient (ie, for data transcription in epidemiologic studies that analyze hundreds of participants). Finally, both these categorical scales have a high, often significant, correlation with other types of pain measurement and have demonstrated excellent reliability and validity. (5) In this study, I sought to confirm the hypothesis that the evaluation of primary dysmenorrhea can be accomplished with either numeric scales or visual-analog scales and that the results will be equivalent no matter which scale is used.


This study is part of a larger study that deals with prevalence and symptomatology of dysmenorrhea in the state of Madrid. A sample of 2,000 women completed a questionnaire about their menstrual pain. Participants accepted in the sample were required to be free of oral contraceptives, intrauterine contraceptive devices (IUDs), and gynecologic illnesses.

I obtained my sample through advertisements in different schools of the Complutense University of Madrid, in clinics and hospitals of the city’s sanitary area VII, in public and private secondary schools, and in women’s associations in the Madrid community. All of the participants gave their permission to use the data in this study. I rejected incorrectly completed questionnaires. Demographic characteristics of the participants are summarized in Table 1. The final sample consisted of 1,387 women who answered a questionnaire that asked them to evaluate the average intensity of their menstrual pain in the last menstrual period by VAS and by a numeric scale (Table 2).


To compare the measurements from both scales, I multiplied the data gathered from the numeric scale (range = 0-10) by 10 so that the new range was from 0 to 100, just as was the VAS of 100 mm. For the global sample, the average pain intensity during the last menstrual period was 45.79 points on the numeric scale (with a standard deviation of 29.89) and 42.84 points on the VAS (SD = 29.90). The Pearson correlation coefficient of the measurements in both scales was .957 (p .0001). I believe this result was related to the large number of participants (19.3% of the sample) who scored either 0 or 100 on both scales (in other words, they chose only the extreme scores). After those scores were eliminated, the new Spearman correlation coefficient was .933 (p < .0001). The difference in the average pain intensity rated by each of the scales was 2.95 points, with SD of 8.81 (t = 12.339; df = 1357; p < .001), and range from 0 to 64 points.

Because the average difference was statistically significant, I made the following analyses. I determined that 21.4% of the women’s scores did not differ, although 19.3% of the sample scored in the extremes on both scales. The percentage of women whose scores differed in absolute value from 0 to 5 points was 62.9%. There were no significant differences between these 2 groups regarding age or educational/professional levels. The percentage of women whose differences ranged from 5 to 10 points was 22.9%, whereas 7.6% of the women had rating differences of 15 to 20 points, and 4.1% had rating differences of 20 to 25 points. Only 2.5% of the sample showed rating differences greater than 25 points.

I also examined the scoring tendencies of these participants, using the difference in nonabsolute values between the scores on both scales. The results showed that 52.6% of the sample scored pain intensity greater on the numeric scale than the mark on the VAS (M = 8.39; SD = 5.16; range 1-64; 12% of the sample had scoring differences greater than 10 points), whereas 26.7% of the women scored pain intensity with a higher mark on the VAS than on the numeric scale (M = 5.47; SD = 5.16; range 1-50, and only 2.2% of the sample had rating differences greater than 10 points). The participants, therefore, tended to quantify their pain intensities at higher levels when they used the numeric scale than they did when they used the VAS.


The VAS and the numeric scale have often been used to evaluate pain intensity, and both types of scales I used in this study have previously been shown to be useful in the evaluation of menstrual pain. Each scale has advantages and disadvantages. For example, the VAS is very precise, but some people have great difficulty understanding the principles behind the analog scale. The numeric scale, on the other hand, is convenient, but the marks are not continuous.

Despite the inherent differences between the 2 scales, the results of this study indicate that they correlate in a significant and seemingly very important manner because I took the extreme evaluations, which could potentially have biased the data, into consideration and eliminated them in further analyses. Although the average difference in pain intensity measured in each scale is significant, I found that a relatively small percentage of women (14.2%) had differences on both scales that were greater than 10 points. The significance of the finding may be attributable to the large difference (as great as 64 points) in marks from a very small percentage of participants (0.3%). This discrepancy may be the consequence of some participants’ having misunderstood the VAS because, in fact, the women had more questions and problems when trying to complete this scale.

The tendency for the women to rate a higher level of pain on the numeric scale may be partially explained by the non-continuous type of measurement it uses. Participants are forced to round their estimations up (in a majority of the cases) or down. Why some people overestimate and some do not, as well as the relationship between these tendencies and other variables, requires further study that is beyond the scope of my current investigation. A number of studies have reported that patients significantly overestimate their pain when asked to recall previous levels of pain, and the results of this study support this tendency.

In summary, the results of this study affirm that both the VAS and the numeric scale are useful in evaluating menstrual pain. However, because of comprehension difficulties some participants using the VAS reported and because of the lack of important differences between measurements taken from both scales, the numeric scale appears to be more adequate and convenient to use, especially in epidemiologic studies requiring analysis of data from a large number of participants.


For further information, please send correspondence to Dr Cristina Larroy, Profesor Titular, Dpto de Personalidad, Evaluacion y Tratamiento Psicologico I (Psicologia Clinica), Facultad de Psicologia, Universidad Complutense de Madrid, 28223 Madrid, Spain (e-mail:


Demographic Characteristics of Participants in a Study

Comparing Numeric and Visual-Analog Pain Scales

Age range (y)

Variable 13-20 21-30 31-40 41-52

Percentage 50 25.4 16.3 8.4

Age (y)

M 16.83 24.63 35.17 43.69

SD 1.94 2.97 2.85 2.5


M 0.01 0.17 1.09 2.11

SD 0.13 0.54 1.02 1.26


None 0.4 0 0.4 2.6

Primary school 23.4 12.9 18.8 38.8

Secondary school 58.3 23.1 38.1 29.3

University 17.9 64 42.6 29.3


Comparison of Pain Ratings on

Visual-Analog (VAS) and Numeric Scales

VAS (mm)

0 100

No pain Worst imaginable pain

Numeric scale

0 1 2 3 4 5 6 7 8 9 10

No pain Worst imaginable pain

Note. Participants mark the VAS at a chosen distance from 0 mm

to 100 mm to indicate severity of pain; those using the numeric

scale indicate a specific number to show pain.


(1.) Jensen M, Karoly P. Self-report scales and procedures for assessing pain in adults. In: Turk DC, Melzack R, eds. Handbook of Pain Assessment. New York: Guilford; 1992:135-151.

(2.) Jensen M, McFarland C. Increasing the reliability and validity of pain intensity measurement in chronic pain patients. Pain. 1993;55:1:195-203.

(3.) White P. Pain and measurement. In: Warfield CA, ed. Principles and Practice of Pain Management. New York: McGraw-Hill; 1993:27-41.

(4.) DeLoach L, Higgins M, Caplan A, Stiff J. The visual analog scale in the immediate postoperative period: Intrasubject variability and correlation with a numeric scale. Anesth Analg. 1998;86:102-106.

(5.) Jensen M, Karoly P, Braver S. The measurement of clinical pain intensity: A comparison of six methods. Pain. 1986;27: 117-126.

Dr Larroy is profesor titular with the Dpto de Personalidad, Evaluacion y Tratamiento Psicologico I (Psicologia Clinica), Facultad de Psicologia, Universidad Complutense de Madrid, Spain.

COPYRIGHT 2002 Heldref Publications

COPYRIGHT 2002 Gale Group