- Email: [email protected]
Validating a pencil-and-paper measure of perimenopausal menstrual blood loss
Women’s Health Issues 14 (2004) 242-247
VALIDATING A PENCIL-AND-PAPER MEASURE OF PERIMENOPAUSAL MENSTRUAL BLOOD LOSS Phyllis Kernoff Mansfield, PhDa*, Ann Voda, RN, PhDb, and Gary Allison, PhDc a
Pennsylvania State University, University Park, Pennsylvania b University of Utah, Salt Lake City, Utah c Department of Evolution, Ecology and Organismal Biology, Ohio State University, Columbus, Ohio Received 2 February 2004; received in revised form 28 June 2004; accepted 6 July 2004
In the study presented here, we describe our efforts to develop and validate a new measurement tool for perimenopausal menstrual blood loss. We validate this simple-to-use, subjective pencil-and-paper scale, the Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (MVJ), against an objective measure, the weight of used menstrual products. Thirty-one women from the Minneapolis-St. Paul, MN region saved all their used products over three menstrual cycles, storing them in airtight zip bags, and rated their menstrual fluid volume loss using the MVJ. The overall correlation between MVJ scores and log discharge rate was quite high (r ⴝ 0.683); all MVJ categories were statistically unique with the exception of categories “1” and “2.” The overall fit, then, was particularly good at the heavy bleeding end of the scale. When five women with poor performance were removed, the correlations ranged from 0.480 – 0.894. The MVJ is a promising tool for identifying women with excessive menstrual bleeding. It is a simple-to-use, pencil-and-paper scale that offers an inexpensive and practical method to clinicians who work with perimenopausal women, the group most vulnerable to unnecessary worry and/or interventions when menorrhagia is incorrectly diagnosed.
Introduction and Background
I
rregular and heavy menstrual bleeding are common gynecological complaints of perimenopausal women (Mansfield & Boyer, 1990; Mansfield & Voda, 1997; Roy & Mattox, 2003; Warren, 2002). Too often, a woman’s report of “heavy bleeding” is transformed into a clinical diagnosis of menorrhagia, which, by definition, is menstrual blood loss of ⱖ80 mL per cycle (Fraser et al., 1985; Hallberg & Nilsson, 1966; Hallberg et al., 1966; Janssen, Scholten, & Heintz, 1997; Roy & Mattox, 2003; Warren, 2002). A diagnosis of menorrhagia increases the possibility that invasive medical and/or surgical procedures will be performed (e.g., dilatation and curettage [D&C] and/or hysterectomy). Excessive menstrual bleeding is estimated to affect at least 10 million American women annually, half of whom are in the 40- to 50-year-old age group—the
* Address correspondence to: Phyllis Kernoff Mansfield, PhD, 102 Willard Building, Pennsylvania State University, University Park, PA 16802. E-mail: [email protected] Copyright © 2004 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc.
years of the menopausal transition (Warren, 2002). Heavy bleeding can cause disruptions to women’s personal and professional lives; in one study, 40% of women with heavy bleeding reported that they could not work outside the home during their menstrual period (Warren, 2002). Women with heavy bleeding feel they have lost bodily control and they worry about accidents due to the unpredictability or heaviness of their menstrual bleeding (Berken, 1986; Kittell et al., 1997; Kittell et al., 1998; Mansfield & Jorgensen, 1992; Patterson & Hale, 1985; Warren, 2002). Changes in the bleeding pattern are often perceived as a sign of illness, and, as such, become a vitally important emotional health issue for women (Berken, 1986; Mansfield & Jorgensen, 1992; Root, Voda, and Smith, 1988). Women’s subjective reports of heavy bleeding, when correlated with objective measures of blood loss of more than 80 mL, have been found to be inaccurate. A study done in the United Kingdom (Haynes et al., 1977) found that 53% of the women who complained of heavy menstrual bleeding did indeed have a loss that repeatedly measured more than 80 mL, but the 1049-3867/04 $-See front matter. doi:10.1016/j.whi.2004.07.005
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
remainder had variable amounts of blood loss from mild to moderate. Hallberg et al. (1966) reported that 45% of the women who reported heavy blood loss had a measured blood loss of less than 40 mL. Other studies also document wide variability in subjective perception of amount of blood lost (Fraser et al., 1984; Fraser et al., 1985; Fraser et al., 1981; Hallberg et al., 1966; Roy & Mattox, 2003; Warren, 2002). Fraser et al. (1984) found that many women who reported using a large number of menstrual products per cycle were actually changing products that were barely discolored. For example, one woman used 48 tampons yet the measured blood loss (using the hematin method) was only 9.6 mL; others have found similar results (Chimbira, Anderson, and Turnbull, 1980). The “current gold standard” (Wyatt et al., 2001) for accurately assessing menstrual blood loss volume is the alkaline hematin method described by Hallberg and Nilsson (1964) and modified by Newton and colleagues (1977). This method requires that women collect their used menstrual products, which are then processed to convert the blood to alkaline hematin and optical density is determined (Oehler & Rees, 2003). The method is both expensive and impractical for routine clinical use. As such, several nonlaboratory methods have been developed to assess the quantity of menstrual blood loss. One is the “pictorial assessment chart” (Higham, O’Brien, and Shaw, 1990). This technique asks women to score their blood loss based on both the number of products used and the degree of staining of each product. Its discriminatory power has recently been called into question (Wyatt et al., 2001), giving rise to an improved pictorial method that takes into account extraneous blood loss (e.g., blood on clothes, on the body), which is thought to be a most distressing event for women (Wyatt et al., 2001). Other visual assessment techniques have been reported in the last decade (Janssen, Scholten, and Heintz, 1995) and all are laden with questionable reliability and validity issues. In an effort to design a convenient and accurate method for measuring menstrual blood loss, Fraser and colleagues (2001) evaluated how closely weights of menstrual protection products approximated estimation of total menstrual fluid volume. Total fluid volume was calculated by subtracting the dry weight from the wet weight of each product. The investigators concluded that the weighing method was not only more convenient than the alkaline hematin method but more accurate than the pictorial assessment chart. Nonetheless, this method still requires the collection of menstrual products and access to clinical facilities. In this paper, we describe our efforts to validate a simple-to-use, subjective scale to estimate total menstrual fluid volume loss, the Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (MVJ), against an
243
objective measure, the weight of used menstrual products. The research we report is the degree of relationship between the ratings made by women on the MVJ and the measured weight of used products.
Methods Sample From a list of 180 participants of the TREMIN Research Program on Women’s Health (formerly Alan Treloar’s Menstrual and Reproductive History Program), 40 women were randomly chosen to participate in this special study and received invitations by phone. All of the eligible participants lived in the Minneapolis/St. Paul, Minnesota area, were between the ages of 35 and 55, were still menstruating, and were not using exogenous hormones. TREMIN is the oldest ongoing research program of women’s menstrual health in the world (see Mansfield & Bracken, 2003a; Mansfield & Bracken, 2000b for a history of TREMIN). Since 1934, participants, nearly all of them white and highly educated, and mostly married, have recorded the onset and closure of their menstrual bleeds on calendar cards. They have also completed annual health report forms. As such, these TREMIN participants are a unique group of committed record keepers. Procedure Collection of products. Participants were instructed to collect all used menstrual products (tampons and/or pads) for three menstrual cycles. They were provided with self-closing, zippered bags on which a self-stick label was affixed with the MVJ and the participants’ unique identification number. Participants were instructed to immediately place each used menstrual product into the bag, expelling the air, as they changed to a clean product, and in addition to rate the blood loss on the MVJ and to provide information about the day of menses, date and time of the product change, and the product brand and absorbency, all on the printed label affixed to the bag. Participants were provided with their choice of tampons and/or pads. The reason for this was based upon the knowledge that women have preferences regarding menstrual products. Allowing them to use their products of choice was assumed to increase the likelihood of cooperation. Participants were asked to store the bags in a cool place (the refrigerator), and as soon as bleeding had ended, to call the research assistant who would drive to the participants’ homes, pick up the used products, place them in a cooler, and transport them to the laboratory.
244
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Table 1. Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (1) Spotting, a drop or two of blood, not even requiring sanitary protection though you may prefer to use some. (2) Very light bleeding (you would need to change the least absorbent tampon or pad one or two times per day, though you may prefer to change more frequently). (3) Light bleeding (you would need to change a low or regular absorbency tampon or pad two or three times per day, though you may prefer to change more frequently). (4) Moderate bleeding (you would need to change a regular absorbency tampon or pad every 3 to 4 hours, though you may prefer to change more frequently). (5) Heavy bleeding (you would need to change a high absorbency tampon or pad every 3 to 4 hours, though you may prefer to change more frequently). (6) Very heavy bleeding or gushing (protection hardly works at all; you would need to change the highest absorbency tampon or pad every hour or two).
Instruments. To measure the total quantity of menstrual fluid discharge on used products, a total of 1,489 bags of used products from 31 participants were collected for weighing. To measure the true weights of these products, it was necessary first to have accurate measures of the original unused (dry) weights of these same brands and types of products. This was accomplished by measuring each unused brand and type three times to be sure we had an accurate weight for each. Products to be weighed were randomly selected from each category of brand/type, and to minimize error from handling (sweat), the research assistant wore gloves. The mean dry weight of each unused product was entered into the database and subtracted from the weight of the used product to obtain the weight of the menstrual discharge. This method is nearly identical to that employed by Fraser and colleagues (2001), and like them, we assume unity specific gravity for fluid volume captured in menstrual products and therefore report results in milliliters rather than grams. The instrument used to weigh the menstrual products was an Ohause mechanical scale, which had a sensitivity of 0.01 g. The scale was calibrated prior to weighing as well as during weighing of large numbers of products. The instrument was evaluated as stable; interweigher reliability between the research assistant and the project director was .98. The used products were weighed in triplicate and the mean of the three weights entered into the database. The MVJ was used to estimate the total quantity of menstrual fluid discharge. This Likert-type scale was developed by the investigators for use by participants who were keeping menstrual diaries. The 6-point scale ranges from 1 (spotting) to 6 (gushing) and is based on the participant’s perception of how often she would need to change a product (not how often she actually does). Previous studies (Chimbira et al., 1980; Fraser, et al., 1984) have shown that some women will change products that are barely discolored, and we wanted to avoid that problem. Table 1 presents the 6 scale points and the guidelines for selecting each rating.
Statistical analyses Our aim was to compare MVJ scores with the total fluid volume captured in each menstrual product. Because the products were worn for differing amounts of time, volumes were transformed to rates (dividing the volume by the amount of time the product was used). The individual rates obtained are an integration of the potentially variable discharge rates during the time the product was used. For example, the rate of flow for a tampon that captured a sudden gush in the last 2 minutes of its use would be calculated as the average rate of flow over the entire time of use, not just during the gush. The flow rate data were positively skewed and were transformed to log base 10, a common procedure to use for many rates and appropriate when error grows with the size of the dependent variable (Sokal & Rohlf, 1981). As is customary, a constant value of 1 was added to all rates before transformation to permit conversion of rates of zero to the log scale. Transformation was necessary to satisfy the normality assumption of analysis of variance, but in addition served to improve the fit of the MVJ score with rate. We used a multiple comparison procedure to test for the uniqueness of the six MVJ scores. To calculate the percentage of total menstrual blood loss in the measured fluid volume, we used the conversion factor given in Fraser et al. (2001, p. 808).
Results Thirty-one women completed the study. One participant dropped out because she had a hysterectomy. The mean age of the participants was 47.6 years (range 38 –52). Women selected 28 different brands of tampons and pads for use during the three menstrual cycles of this study. If we include different absorbencies of those brands selected by the participants, that figure rises to 87. The average number of products used by participants was 48.3 (SD ⫽ 24.4) Although our sample was selected randomly and was not a clinical sample, the range of total discharge
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Figure 1. Range of total fluid discharge/cycle within the study. There were 89 total cycles within the study.
volumes was large (from barely measurable to greater than 500 mL/cycle; Figure 1) and provides an adequate pool of data to validate the MVJ. Our range was similar to that reported by Fraser et al. (2001) in which they intentionally sampled women with normal and with heavy bleeding. The overall correlation between MVJ scores and log discharge rate is quite high (Figure 2). With all data combined, the correlation (r) is 0.683. Using multiple comparison tests (with Holm p-adjustment), all MVJ categories were statistically unique with the exception of categories 1 and 2. Thus, the overall fit was particularly good at the heavy end of the scale. Correlations within individual participants were strong. The majority of participants (26 of 31) showed positive and statistically significant correlations between scale ratings and actual menstrual fluid volume loss, ranging from 0.480 – 0.894. Figure 3 depicts the strong correlation for one such participant. Five participants showed positive but nonsignificant slopes (see Figure 4 for an example of a poor relationship for one of these individuals). These five women shared characteristics that explained their relatively poor performance. Two of them apparently did not understand the instructions, because both used the scale point “1” (spotting) to rate some of their highest bleed. The three other participants used relatively few products and scored them with only a few MVJ categories, thus reducing their correlative power. None of the cycles rated by these five women were abnormally heavy (maximum ⫽ 143.7 mL total fluid discharge or 70.23 mL blood loss).
245
Figure 2. Relationship between fluid discharge as log10(rate) in units of mL/hour for each menstrual product and the MVJ score selected by the participants for all products collected. Filled triangles indicate the mean of that MVJ score.
Discussion In this study, we tested the validity of the MVJ by comparing the scale ratings selected by participants to describe the magnitude of blood loss against an objective measure, the weight of used menstrual products. Work by Fraser et al. (2001) confirmed that the weight of total menstrual volume is an accurate method to use
Figure 3. Example of a strong relationship between fluid discharge rate and MVJ score for a single individual.
246
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Figure 4. Example of a poor relationship between fluid discharge rate and MVJ score for a single individual. Although there is a slight increasing trend, this relationship is not statistically significant.
to estimate the volume of blood lost. The findings of our study confirm that the MVJ is a valid way to estimate menstrual fluid volume loss. The overall fit is particularly good at the heavy end of the MVJ (heaviest bleeding, rating range 4 – 6). In other words, the MVJ is particularly appropriate for estimating episodes of moderate to very heavy bleeding. The need for a practical and inexpensive way to accurately assess the volume of women’s menstrual discharge is particularly great for women in the perimenopause, when heavy bleeding is a common problem. Estimates are that from one-fifth to two-thirds of the approximately 600,000 hysterectomies and 40% of the 555,000 D&Cs performed in the United States every year are to remedy heavy bleeding (Oehler & Rees, 2003; Voda, 1997; Warren, 2002; Wyatt et al., 2001). Hysterectomy is not without risk and is accompanied by complications in 24 – 43% of cases (Oehler & Rees, 2003). As such, Fraser and colleagues (1984) and others (Shaw, 1973) have considered the accurate assessment of volume of menstrual blood loss to be the most important single parameter of the menstrual history, as it is on this basis that the decision to perform these procedures is usually made. In response to the reality that women’s self-reports of heavy bleeding, or menorrhagia, are frequently inaccurate, researchers have attempted to design instruments and tools to measure menstrual discharge, with only fair success. The most accurate method, measuring blood hematin, is expensive, invasive, and time-consuming. Simpler methods have been found to lack validity.
The MVJ is a promising tool for identifying women with excessive menstrual bleeding. It is particularly sensitive when estimating moderate to heavy bleeding. It is a simple to use pencil-and-paper scale that offers a practical and very inexpensive method for clinicians to use with midlife/perimenopausal patients, the group most vulnerable to unnecessary worry and/or interventions when menorrhagia is incorrectly diagnosed. The MVJ does not measure menstrual blood loss per se. It is a tool for estimating blood loss from total menstrual fluid volume. Menstrual fluid includes mucus and endometrial tissue along with blood (Heath, Skeaff, & Gibson, 1998). Using the equation of Fraser et al. (2001) for calculating the percent of blood in total menstrual fluid volume, we were able to determine that for menstrual cycles when a woman reported at least one score of 5 or 6, the mean blood loss was 67.2 mL. Taking just those menstrual cycles when 6 was reported, the mean blood loss was 107.7 mL. Given our pool of participants, the report of at least one 6 within a cycle has a 50% probability (using the median) of being greater than or equal to 113.6 mL of total blood loss, a notable volume. While a few reports of 6 occurred in cycles with normal volumes in our study, a participant’s use of 6 may indicate a bleeding volume of some concern. We recommend careful observation and perhaps further testing in such cases. The MVJ did not work well for five participants. Three of them, none heavy bleeders, used relatively few products and scored them with few scale points, thus reducing their correlative power. Ratings on the lower end of the scale, or none at all, are more likely to occur with light bleeders. The other two women obviously did not understand the instructions for how to use the scale. When the investigators met with the participants following the termination of the study, a few women admitted that they did not give much attention to the scale— or to observing their used products— because they felt they knew their cycles well enough! Perhaps the fact that our participants were experienced menstrual record keepers turned out to be a detriment. In any case, it is important to provide clear instructions to women who will use the scale. We collected data over three menstrual cycles and recommend a similar strategy to researchers and clinicians. There is ongoing debate about the extent of menstrual variability from cycle to cycle (see Fraser et al., 2001; Hallberg & Nilsson, 1964). Fraser et al. make a convincing argument that women’s complaints of very heavy bleeding are more likely to be based on the occurrence of one very heavy period and that in a short collection period, such a very heavy period might be missed.
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Acknowledgments This research was funded in part by Tambrands, Inc. and Public Health Service Research Grant No. RR-64 from the Division of Research Resources, U.S. Public Health Service.
REFERENCES Berken, C. (1986). In behalf of women over 40: understanding the importance of the menopause. Social Work, Sept/Oct, 378 –384. Chimbira, T., Anderson, A. B., & Turnbull, A. (1980). Relation between measured menstrual blood loss and patients’ subjective assessment of loss, duration of bleeding, number of sanitary towels used, uterine weight and endometrial surface area. British Journal of Obstetrics and Gynaecology, 87, 603– 609. Fraser, I. S., McCarron, G., & Markham, R. (1984). A preliminary study of factors influencing perception of menstrual blood loss volume. American Journal of Obstetrics and Gynecology, 149, 788 – 793. Fraser, I. S., McCarron, G., Markham, R., & Resta, T. (1985). Blood and total fluid content of menstrual discharge. Obstetrics and Gynecology, 65(2), 194 –198. Fraser, I. S., Pearse, C., Sherman, R. P., Elliot, P. M., McIlveen, J., & Markham, R. (1981). Efficacy of mefanamic acid in patients with a complaint of menorrhagia. Obstetrics and Gynecology, 58(4), 543–551. Fraser, I. S., Warner, P., & Marantos, P. A. (2001). Estimating menstrual blood loss in women with normal and excessive menstrual fluid volume. Obstetrics and Gynecology, 98, 806 – 814. Hallberg, L., Hogdahl, A., Nilsson, L., & Rybo, G. (1966). Menstrual blood loss: a population study. Acta Obstetricia et Gynecologica Scandinavica, 45, 320 –351. Hallberg, L., & Nilsson, L. (1964). Constancy of individual menstrual blood loss. Acta Obstetricia et Gynecologica Scandinavica, 43, 352–359. Hallberg, L., & Nilsson, L. (1966). Determination of menstrual blood loss. Scandinavian Journal of Clinical Laboratory Investigation, 16, 244 –248. Haynes, P. T., Hodgson, H., Anderson, A. B. M., & Turnbull, A. C. (1977). Measurements of menstrual blood loss in patients complaining of menorrhagia. British Journal of Obstetrics and Gynaecology, 84, 763–768. Heath, A.-L. M., Skeaff, C. M., & Gibson, R. S. (1998). Validation of a questionnaire method for estimating extent of menstrual blood loss in young adult women. Journal of Trace Elements in Medical Biology, 12, 231–235. Higham, J. M., O’Brien, P. M. S., & Shaw, R. W. (1990). Assessment of menstrual blood loss using a pictorial chart. British Journal of Obstetrics and Gynaecology, 97, 734 –739. Janssen, C. A. H., Scholten, P. C., & Heintz, A. P. (1995). A simple visual assessment technique to discriminate between menorrhagia and normal blood loss. Obstetrics and Gynecology, 85(6), 977–982. Janssen, C. A. H., Scholten, P. C., & Heintz, A. P. (1997). Menorrhagia—a search for epidemiological risk markers. Maturitas, 28, 19 –25. Kittell, L., Mansfield, P. K., Morse, J., & Voda, A. (1997). Experiencing changes in menstrual bleeding during the menopausal transition. Menopause, 4(3), 173–183. Kittell, L., Mansfield, P. K., & Voda, A. (1998). Keeping up appearances: the basic social process of the menopausal transition. Qualitative Health Research, 8(5), 618 – 633. Mansfield, P. K., & Boyer, B. (1990). The experiences, concerns and health care needs of women in the menopausal transition. Annals of the New York Academy of Sciences, 592, 448 – 449.
247
Mansfield, P. K., & Bracken, S. (2003a). The TREMIN Program: sixty-eight years of research on menstruation and women’s health. Women’s Studies Quarterly, 31(1/2), 25– 41. Mansfield, P. K., & Bracken, S. (2003b). TREMIN: a history of the world’s oldest ongoing study of menstruation and women’s health. Lemont, PA: East Rim Publishers. Mansfield, P. K., & Jorgensen, C. (1992). Menstrual pattern changes in middle-aged women. In A. Dan & L. Lewis (Eds.), Menstrual health in women’s lives (pp. 213–225). Chicago: University of Illinois Press. Mansfield, P. K., & Voda, A. (1997). Woman-centered information on menopause for health care providers. Health Care for Women International, 18, 55–72. Newton, J., Barnard, G., & Collins, W. (1977). A rapid method for measuring menstrual blood loss using automatic extraction. Contraception, 16, 269 –282. Oehler, M. K., & Rees, M. C. P. (2003). Menorrhagia: an update. Acta Obstetricia et Gynecologica Scandinavica, 82(5), 405– 422. Patterson, E. T., & Hale, E. S. (1985). Making sure: integrating menstrual care practices into activities of daily living. Advances in Nursing Science, 4, 18 –31. Root, J. L., Voda, A., & Smith, K. (1988). Women’s perceptions of life events which may have the potential to alter menstrual cycles. Paper presented at the Third International Congress on Women’s Health, Tampa, FL. Roy, K. H., & Mattox, J. H. (2003). Endometrial ablation for perimenopausal menorrhagia. Menopause Management, Sep/Oct, 13–17. Shaw, S. T., Jr. (1973). Quantification of menstrual blood loss. Postgraduate Medicine, 53(4), 204 –208. Sokal, R. R., & Rohlf, F. J. (1981). Biometry: the principle and practice of statistics in biological research. New York: W.H. Freeman and Company. Voda, A. M. (1997). Menopause, me and you: the sound of women pausing. Binghamton, NY: Harrington Park Press. Warren, M. (2002). The need for proper diagnosis and treatment of menorrhagia. Menopause Management, Nov/Dec, 10 –15. Wyatt, K., Dimmock, P. W., Walker, T., & O’Brien, P. M. S. (2001). Determination of total menstrual blood loss. Fertility and Sterility, 76(1), 125–131.
Author Descriptions Phyllis Kernoff Mansfield, PhD, is a professor of Women’s Studies and Health Education, and senior scientist at the Population Research Institute at Penn State. Her research interests include women’s health, the reproductive concerns of midlife women, sexuality, and the psychology of health. She is the Director of the TREMIN Research Program on Women’s Health and co-director of the Midlife Women’s Health Survey and serves on the board of directors of the Society for Menstrual Cycle Research. Ann Voda was the second director of the TREMIN Research Program on Women’s Health from 1984 to 1998 at the University of Utah. She is a former president of the North American Menopause Society. Dr. Gary Allison focuses on computationally intensive analyses of large datasets in fields ranging from bioinformatics to ecology and conservation.
VALIDATING A PENCIL-AND-PAPER MEASURE OF PERIMENOPAUSAL MENSTRUAL BLOOD LOSS Phyllis Kernoff Mansfield, PhDa*, Ann Voda, RN, PhDb, and Gary Allison, PhDc a
Pennsylvania State University, University Park, Pennsylvania b University of Utah, Salt Lake City, Utah c Department of Evolution, Ecology and Organismal Biology, Ohio State University, Columbus, Ohio Received 2 February 2004; received in revised form 28 June 2004; accepted 6 July 2004
In the study presented here, we describe our efforts to develop and validate a new measurement tool for perimenopausal menstrual blood loss. We validate this simple-to-use, subjective pencil-and-paper scale, the Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (MVJ), against an objective measure, the weight of used menstrual products. Thirty-one women from the Minneapolis-St. Paul, MN region saved all their used products over three menstrual cycles, storing them in airtight zip bags, and rated their menstrual fluid volume loss using the MVJ. The overall correlation between MVJ scores and log discharge rate was quite high (r ⴝ 0.683); all MVJ categories were statistically unique with the exception of categories “1” and “2.” The overall fit, then, was particularly good at the heavy bleeding end of the scale. When five women with poor performance were removed, the correlations ranged from 0.480 – 0.894. The MVJ is a promising tool for identifying women with excessive menstrual bleeding. It is a simple-to-use, pencil-and-paper scale that offers an inexpensive and practical method to clinicians who work with perimenopausal women, the group most vulnerable to unnecessary worry and/or interventions when menorrhagia is incorrectly diagnosed.
Introduction and Background
I
rregular and heavy menstrual bleeding are common gynecological complaints of perimenopausal women (Mansfield & Boyer, 1990; Mansfield & Voda, 1997; Roy & Mattox, 2003; Warren, 2002). Too often, a woman’s report of “heavy bleeding” is transformed into a clinical diagnosis of menorrhagia, which, by definition, is menstrual blood loss of ⱖ80 mL per cycle (Fraser et al., 1985; Hallberg & Nilsson, 1966; Hallberg et al., 1966; Janssen, Scholten, & Heintz, 1997; Roy & Mattox, 2003; Warren, 2002). A diagnosis of menorrhagia increases the possibility that invasive medical and/or surgical procedures will be performed (e.g., dilatation and curettage [D&C] and/or hysterectomy). Excessive menstrual bleeding is estimated to affect at least 10 million American women annually, half of whom are in the 40- to 50-year-old age group—the
* Address correspondence to: Phyllis Kernoff Mansfield, PhD, 102 Willard Building, Pennsylvania State University, University Park, PA 16802. E-mail: [email protected] Copyright © 2004 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc.
years of the menopausal transition (Warren, 2002). Heavy bleeding can cause disruptions to women’s personal and professional lives; in one study, 40% of women with heavy bleeding reported that they could not work outside the home during their menstrual period (Warren, 2002). Women with heavy bleeding feel they have lost bodily control and they worry about accidents due to the unpredictability or heaviness of their menstrual bleeding (Berken, 1986; Kittell et al., 1997; Kittell et al., 1998; Mansfield & Jorgensen, 1992; Patterson & Hale, 1985; Warren, 2002). Changes in the bleeding pattern are often perceived as a sign of illness, and, as such, become a vitally important emotional health issue for women (Berken, 1986; Mansfield & Jorgensen, 1992; Root, Voda, and Smith, 1988). Women’s subjective reports of heavy bleeding, when correlated with objective measures of blood loss of more than 80 mL, have been found to be inaccurate. A study done in the United Kingdom (Haynes et al., 1977) found that 53% of the women who complained of heavy menstrual bleeding did indeed have a loss that repeatedly measured more than 80 mL, but the 1049-3867/04 $-See front matter. doi:10.1016/j.whi.2004.07.005
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
remainder had variable amounts of blood loss from mild to moderate. Hallberg et al. (1966) reported that 45% of the women who reported heavy blood loss had a measured blood loss of less than 40 mL. Other studies also document wide variability in subjective perception of amount of blood lost (Fraser et al., 1984; Fraser et al., 1985; Fraser et al., 1981; Hallberg et al., 1966; Roy & Mattox, 2003; Warren, 2002). Fraser et al. (1984) found that many women who reported using a large number of menstrual products per cycle were actually changing products that were barely discolored. For example, one woman used 48 tampons yet the measured blood loss (using the hematin method) was only 9.6 mL; others have found similar results (Chimbira, Anderson, and Turnbull, 1980). The “current gold standard” (Wyatt et al., 2001) for accurately assessing menstrual blood loss volume is the alkaline hematin method described by Hallberg and Nilsson (1964) and modified by Newton and colleagues (1977). This method requires that women collect their used menstrual products, which are then processed to convert the blood to alkaline hematin and optical density is determined (Oehler & Rees, 2003). The method is both expensive and impractical for routine clinical use. As such, several nonlaboratory methods have been developed to assess the quantity of menstrual blood loss. One is the “pictorial assessment chart” (Higham, O’Brien, and Shaw, 1990). This technique asks women to score their blood loss based on both the number of products used and the degree of staining of each product. Its discriminatory power has recently been called into question (Wyatt et al., 2001), giving rise to an improved pictorial method that takes into account extraneous blood loss (e.g., blood on clothes, on the body), which is thought to be a most distressing event for women (Wyatt et al., 2001). Other visual assessment techniques have been reported in the last decade (Janssen, Scholten, and Heintz, 1995) and all are laden with questionable reliability and validity issues. In an effort to design a convenient and accurate method for measuring menstrual blood loss, Fraser and colleagues (2001) evaluated how closely weights of menstrual protection products approximated estimation of total menstrual fluid volume. Total fluid volume was calculated by subtracting the dry weight from the wet weight of each product. The investigators concluded that the weighing method was not only more convenient than the alkaline hematin method but more accurate than the pictorial assessment chart. Nonetheless, this method still requires the collection of menstrual products and access to clinical facilities. In this paper, we describe our efforts to validate a simple-to-use, subjective scale to estimate total menstrual fluid volume loss, the Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (MVJ), against an
243
objective measure, the weight of used menstrual products. The research we report is the degree of relationship between the ratings made by women on the MVJ and the measured weight of used products.
Methods Sample From a list of 180 participants of the TREMIN Research Program on Women’s Health (formerly Alan Treloar’s Menstrual and Reproductive History Program), 40 women were randomly chosen to participate in this special study and received invitations by phone. All of the eligible participants lived in the Minneapolis/St. Paul, Minnesota area, were between the ages of 35 and 55, were still menstruating, and were not using exogenous hormones. TREMIN is the oldest ongoing research program of women’s menstrual health in the world (see Mansfield & Bracken, 2003a; Mansfield & Bracken, 2000b for a history of TREMIN). Since 1934, participants, nearly all of them white and highly educated, and mostly married, have recorded the onset and closure of their menstrual bleeds on calendar cards. They have also completed annual health report forms. As such, these TREMIN participants are a unique group of committed record keepers. Procedure Collection of products. Participants were instructed to collect all used menstrual products (tampons and/or pads) for three menstrual cycles. They were provided with self-closing, zippered bags on which a self-stick label was affixed with the MVJ and the participants’ unique identification number. Participants were instructed to immediately place each used menstrual product into the bag, expelling the air, as they changed to a clean product, and in addition to rate the blood loss on the MVJ and to provide information about the day of menses, date and time of the product change, and the product brand and absorbency, all on the printed label affixed to the bag. Participants were provided with their choice of tampons and/or pads. The reason for this was based upon the knowledge that women have preferences regarding menstrual products. Allowing them to use their products of choice was assumed to increase the likelihood of cooperation. Participants were asked to store the bags in a cool place (the refrigerator), and as soon as bleeding had ended, to call the research assistant who would drive to the participants’ homes, pick up the used products, place them in a cooler, and transport them to the laboratory.
244
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Table 1. Mansfield-Voda-Jorgensen Menstrual Bleeding Scale (1) Spotting, a drop or two of blood, not even requiring sanitary protection though you may prefer to use some. (2) Very light bleeding (you would need to change the least absorbent tampon or pad one or two times per day, though you may prefer to change more frequently). (3) Light bleeding (you would need to change a low or regular absorbency tampon or pad two or three times per day, though you may prefer to change more frequently). (4) Moderate bleeding (you would need to change a regular absorbency tampon or pad every 3 to 4 hours, though you may prefer to change more frequently). (5) Heavy bleeding (you would need to change a high absorbency tampon or pad every 3 to 4 hours, though you may prefer to change more frequently). (6) Very heavy bleeding or gushing (protection hardly works at all; you would need to change the highest absorbency tampon or pad every hour or two).
Instruments. To measure the total quantity of menstrual fluid discharge on used products, a total of 1,489 bags of used products from 31 participants were collected for weighing. To measure the true weights of these products, it was necessary first to have accurate measures of the original unused (dry) weights of these same brands and types of products. This was accomplished by measuring each unused brand and type three times to be sure we had an accurate weight for each. Products to be weighed were randomly selected from each category of brand/type, and to minimize error from handling (sweat), the research assistant wore gloves. The mean dry weight of each unused product was entered into the database and subtracted from the weight of the used product to obtain the weight of the menstrual discharge. This method is nearly identical to that employed by Fraser and colleagues (2001), and like them, we assume unity specific gravity for fluid volume captured in menstrual products and therefore report results in milliliters rather than grams. The instrument used to weigh the menstrual products was an Ohause mechanical scale, which had a sensitivity of 0.01 g. The scale was calibrated prior to weighing as well as during weighing of large numbers of products. The instrument was evaluated as stable; interweigher reliability between the research assistant and the project director was .98. The used products were weighed in triplicate and the mean of the three weights entered into the database. The MVJ was used to estimate the total quantity of menstrual fluid discharge. This Likert-type scale was developed by the investigators for use by participants who were keeping menstrual diaries. The 6-point scale ranges from 1 (spotting) to 6 (gushing) and is based on the participant’s perception of how often she would need to change a product (not how often she actually does). Previous studies (Chimbira et al., 1980; Fraser, et al., 1984) have shown that some women will change products that are barely discolored, and we wanted to avoid that problem. Table 1 presents the 6 scale points and the guidelines for selecting each rating.
Statistical analyses Our aim was to compare MVJ scores with the total fluid volume captured in each menstrual product. Because the products were worn for differing amounts of time, volumes were transformed to rates (dividing the volume by the amount of time the product was used). The individual rates obtained are an integration of the potentially variable discharge rates during the time the product was used. For example, the rate of flow for a tampon that captured a sudden gush in the last 2 minutes of its use would be calculated as the average rate of flow over the entire time of use, not just during the gush. The flow rate data were positively skewed and were transformed to log base 10, a common procedure to use for many rates and appropriate when error grows with the size of the dependent variable (Sokal & Rohlf, 1981). As is customary, a constant value of 1 was added to all rates before transformation to permit conversion of rates of zero to the log scale. Transformation was necessary to satisfy the normality assumption of analysis of variance, but in addition served to improve the fit of the MVJ score with rate. We used a multiple comparison procedure to test for the uniqueness of the six MVJ scores. To calculate the percentage of total menstrual blood loss in the measured fluid volume, we used the conversion factor given in Fraser et al. (2001, p. 808).
Results Thirty-one women completed the study. One participant dropped out because she had a hysterectomy. The mean age of the participants was 47.6 years (range 38 –52). Women selected 28 different brands of tampons and pads for use during the three menstrual cycles of this study. If we include different absorbencies of those brands selected by the participants, that figure rises to 87. The average number of products used by participants was 48.3 (SD ⫽ 24.4) Although our sample was selected randomly and was not a clinical sample, the range of total discharge
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Figure 1. Range of total fluid discharge/cycle within the study. There were 89 total cycles within the study.
volumes was large (from barely measurable to greater than 500 mL/cycle; Figure 1) and provides an adequate pool of data to validate the MVJ. Our range was similar to that reported by Fraser et al. (2001) in which they intentionally sampled women with normal and with heavy bleeding. The overall correlation between MVJ scores and log discharge rate is quite high (Figure 2). With all data combined, the correlation (r) is 0.683. Using multiple comparison tests (with Holm p-adjustment), all MVJ categories were statistically unique with the exception of categories 1 and 2. Thus, the overall fit was particularly good at the heavy end of the scale. Correlations within individual participants were strong. The majority of participants (26 of 31) showed positive and statistically significant correlations between scale ratings and actual menstrual fluid volume loss, ranging from 0.480 – 0.894. Figure 3 depicts the strong correlation for one such participant. Five participants showed positive but nonsignificant slopes (see Figure 4 for an example of a poor relationship for one of these individuals). These five women shared characteristics that explained their relatively poor performance. Two of them apparently did not understand the instructions, because both used the scale point “1” (spotting) to rate some of their highest bleed. The three other participants used relatively few products and scored them with only a few MVJ categories, thus reducing their correlative power. None of the cycles rated by these five women were abnormally heavy (maximum ⫽ 143.7 mL total fluid discharge or 70.23 mL blood loss).
245
Figure 2. Relationship between fluid discharge as log10(rate) in units of mL/hour for each menstrual product and the MVJ score selected by the participants for all products collected. Filled triangles indicate the mean of that MVJ score.
Discussion In this study, we tested the validity of the MVJ by comparing the scale ratings selected by participants to describe the magnitude of blood loss against an objective measure, the weight of used menstrual products. Work by Fraser et al. (2001) confirmed that the weight of total menstrual volume is an accurate method to use
Figure 3. Example of a strong relationship between fluid discharge rate and MVJ score for a single individual.
246
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Figure 4. Example of a poor relationship between fluid discharge rate and MVJ score for a single individual. Although there is a slight increasing trend, this relationship is not statistically significant.
to estimate the volume of blood lost. The findings of our study confirm that the MVJ is a valid way to estimate menstrual fluid volume loss. The overall fit is particularly good at the heavy end of the MVJ (heaviest bleeding, rating range 4 – 6). In other words, the MVJ is particularly appropriate for estimating episodes of moderate to very heavy bleeding. The need for a practical and inexpensive way to accurately assess the volume of women’s menstrual discharge is particularly great for women in the perimenopause, when heavy bleeding is a common problem. Estimates are that from one-fifth to two-thirds of the approximately 600,000 hysterectomies and 40% of the 555,000 D&Cs performed in the United States every year are to remedy heavy bleeding (Oehler & Rees, 2003; Voda, 1997; Warren, 2002; Wyatt et al., 2001). Hysterectomy is not without risk and is accompanied by complications in 24 – 43% of cases (Oehler & Rees, 2003). As such, Fraser and colleagues (1984) and others (Shaw, 1973) have considered the accurate assessment of volume of menstrual blood loss to be the most important single parameter of the menstrual history, as it is on this basis that the decision to perform these procedures is usually made. In response to the reality that women’s self-reports of heavy bleeding, or menorrhagia, are frequently inaccurate, researchers have attempted to design instruments and tools to measure menstrual discharge, with only fair success. The most accurate method, measuring blood hematin, is expensive, invasive, and time-consuming. Simpler methods have been found to lack validity.
The MVJ is a promising tool for identifying women with excessive menstrual bleeding. It is particularly sensitive when estimating moderate to heavy bleeding. It is a simple to use pencil-and-paper scale that offers a practical and very inexpensive method for clinicians to use with midlife/perimenopausal patients, the group most vulnerable to unnecessary worry and/or interventions when menorrhagia is incorrectly diagnosed. The MVJ does not measure menstrual blood loss per se. It is a tool for estimating blood loss from total menstrual fluid volume. Menstrual fluid includes mucus and endometrial tissue along with blood (Heath, Skeaff, & Gibson, 1998). Using the equation of Fraser et al. (2001) for calculating the percent of blood in total menstrual fluid volume, we were able to determine that for menstrual cycles when a woman reported at least one score of 5 or 6, the mean blood loss was 67.2 mL. Taking just those menstrual cycles when 6 was reported, the mean blood loss was 107.7 mL. Given our pool of participants, the report of at least one 6 within a cycle has a 50% probability (using the median) of being greater than or equal to 113.6 mL of total blood loss, a notable volume. While a few reports of 6 occurred in cycles with normal volumes in our study, a participant’s use of 6 may indicate a bleeding volume of some concern. We recommend careful observation and perhaps further testing in such cases. The MVJ did not work well for five participants. Three of them, none heavy bleeders, used relatively few products and scored them with few scale points, thus reducing their correlative power. Ratings on the lower end of the scale, or none at all, are more likely to occur with light bleeders. The other two women obviously did not understand the instructions for how to use the scale. When the investigators met with the participants following the termination of the study, a few women admitted that they did not give much attention to the scale— or to observing their used products— because they felt they knew their cycles well enough! Perhaps the fact that our participants were experienced menstrual record keepers turned out to be a detriment. In any case, it is important to provide clear instructions to women who will use the scale. We collected data over three menstrual cycles and recommend a similar strategy to researchers and clinicians. There is ongoing debate about the extent of menstrual variability from cycle to cycle (see Fraser et al., 2001; Hallberg & Nilsson, 1964). Fraser et al. make a convincing argument that women’s complaints of very heavy bleeding are more likely to be based on the occurrence of one very heavy period and that in a short collection period, such a very heavy period might be missed.
P. K. Mansfield et al. / Women’s Health Issues 14 (2004) 242-247
Acknowledgments This research was funded in part by Tambrands, Inc. and Public Health Service Research Grant No. RR-64 from the Division of Research Resources, U.S. Public Health Service.
REFERENCES Berken, C. (1986). In behalf of women over 40: understanding the importance of the menopause. Social Work, Sept/Oct, 378 –384. Chimbira, T., Anderson, A. B., & Turnbull, A. (1980). Relation between measured menstrual blood loss and patients’ subjective assessment of loss, duration of bleeding, number of sanitary towels used, uterine weight and endometrial surface area. British Journal of Obstetrics and Gynaecology, 87, 603– 609. Fraser, I. S., McCarron, G., & Markham, R. (1984). A preliminary study of factors influencing perception of menstrual blood loss volume. American Journal of Obstetrics and Gynecology, 149, 788 – 793. Fraser, I. S., McCarron, G., Markham, R., & Resta, T. (1985). Blood and total fluid content of menstrual discharge. Obstetrics and Gynecology, 65(2), 194 –198. Fraser, I. S., Pearse, C., Sherman, R. P., Elliot, P. M., McIlveen, J., & Markham, R. (1981). Efficacy of mefanamic acid in patients with a complaint of menorrhagia. Obstetrics and Gynecology, 58(4), 543–551. Fraser, I. S., Warner, P., & Marantos, P. A. (2001). Estimating menstrual blood loss in women with normal and excessive menstrual fluid volume. Obstetrics and Gynecology, 98, 806 – 814. Hallberg, L., Hogdahl, A., Nilsson, L., & Rybo, G. (1966). Menstrual blood loss: a population study. Acta Obstetricia et Gynecologica Scandinavica, 45, 320 –351. Hallberg, L., & Nilsson, L. (1964). Constancy of individual menstrual blood loss. Acta Obstetricia et Gynecologica Scandinavica, 43, 352–359. Hallberg, L., & Nilsson, L. (1966). Determination of menstrual blood loss. Scandinavian Journal of Clinical Laboratory Investigation, 16, 244 –248. Haynes, P. T., Hodgson, H., Anderson, A. B. M., & Turnbull, A. C. (1977). Measurements of menstrual blood loss in patients complaining of menorrhagia. British Journal of Obstetrics and Gynaecology, 84, 763–768. Heath, A.-L. M., Skeaff, C. M., & Gibson, R. S. (1998). Validation of a questionnaire method for estimating extent of menstrual blood loss in young adult women. Journal of Trace Elements in Medical Biology, 12, 231–235. Higham, J. M., O’Brien, P. M. S., & Shaw, R. W. (1990). Assessment of menstrual blood loss using a pictorial chart. British Journal of Obstetrics and Gynaecology, 97, 734 –739. Janssen, C. A. H., Scholten, P. C., & Heintz, A. P. (1995). A simple visual assessment technique to discriminate between menorrhagia and normal blood loss. Obstetrics and Gynecology, 85(6), 977–982. Janssen, C. A. H., Scholten, P. C., & Heintz, A. P. (1997). Menorrhagia—a search for epidemiological risk markers. Maturitas, 28, 19 –25. Kittell, L., Mansfield, P. K., Morse, J., & Voda, A. (1997). Experiencing changes in menstrual bleeding during the menopausal transition. Menopause, 4(3), 173–183. Kittell, L., Mansfield, P. K., & Voda, A. (1998). Keeping up appearances: the basic social process of the menopausal transition. Qualitative Health Research, 8(5), 618 – 633. Mansfield, P. K., & Boyer, B. (1990). The experiences, concerns and health care needs of women in the menopausal transition. Annals of the New York Academy of Sciences, 592, 448 – 449.
247
Mansfield, P. K., & Bracken, S. (2003a). The TREMIN Program: sixty-eight years of research on menstruation and women’s health. Women’s Studies Quarterly, 31(1/2), 25– 41. Mansfield, P. K., & Bracken, S. (2003b). TREMIN: a history of the world’s oldest ongoing study of menstruation and women’s health. Lemont, PA: East Rim Publishers. Mansfield, P. K., & Jorgensen, C. (1992). Menstrual pattern changes in middle-aged women. In A. Dan & L. Lewis (Eds.), Menstrual health in women’s lives (pp. 213–225). Chicago: University of Illinois Press. Mansfield, P. K., & Voda, A. (1997). Woman-centered information on menopause for health care providers. Health Care for Women International, 18, 55–72. Newton, J., Barnard, G., & Collins, W. (1977). A rapid method for measuring menstrual blood loss using automatic extraction. Contraception, 16, 269 –282. Oehler, M. K., & Rees, M. C. P. (2003). Menorrhagia: an update. Acta Obstetricia et Gynecologica Scandinavica, 82(5), 405– 422. Patterson, E. T., & Hale, E. S. (1985). Making sure: integrating menstrual care practices into activities of daily living. Advances in Nursing Science, 4, 18 –31. Root, J. L., Voda, A., & Smith, K. (1988). Women’s perceptions of life events which may have the potential to alter menstrual cycles. Paper presented at the Third International Congress on Women’s Health, Tampa, FL. Roy, K. H., & Mattox, J. H. (2003). Endometrial ablation for perimenopausal menorrhagia. Menopause Management, Sep/Oct, 13–17. Shaw, S. T., Jr. (1973). Quantification of menstrual blood loss. Postgraduate Medicine, 53(4), 204 –208. Sokal, R. R., & Rohlf, F. J. (1981). Biometry: the principle and practice of statistics in biological research. New York: W.H. Freeman and Company. Voda, A. M. (1997). Menopause, me and you: the sound of women pausing. Binghamton, NY: Harrington Park Press. Warren, M. (2002). The need for proper diagnosis and treatment of menorrhagia. Menopause Management, Nov/Dec, 10 –15. Wyatt, K., Dimmock, P. W., Walker, T., & O’Brien, P. M. S. (2001). Determination of total menstrual blood loss. Fertility and Sterility, 76(1), 125–131.
Author Descriptions Phyllis Kernoff Mansfield, PhD, is a professor of Women’s Studies and Health Education, and senior scientist at the Population Research Institute at Penn State. Her research interests include women’s health, the reproductive concerns of midlife women, sexuality, and the psychology of health. She is the Director of the TREMIN Research Program on Women’s Health and co-director of the Midlife Women’s Health Survey and serves on the board of directors of the Society for Menstrual Cycle Research. Ann Voda was the second director of the TREMIN Research Program on Women’s Health from 1984 to 1998 at the University of Utah. She is a former president of the North American Menopause Society. Dr. Gary Allison focuses on computationally intensive analyses of large datasets in fields ranging from bioinformatics to ecology and conservation.