FEA
TUR
ES
/R
U
bR
iq
UE
S
Validation of the MENQOL for use
with women who have been treated
for gynecologic or breast cancer
by Catherine Dolye, Lauran Adams, Alison McAndrew, Stephanie Burlein-Hall, Tracey DasGupta, Jennie Blake, and Margaret Fitch
ABstrAct
Purpose: To conduct a psychometric evalua-tion of the MENQOL, a condievalua-tion-specific, self-report instrument to assess menopausal symptoms in women with gynecologic and breast cancers.
Methods: Identify face and content validity of the MENQOL with experts, and reliabil-ity and construct validreliabil-ity with a group of women diagnosed with cancer who are suf-fering from treatment-induced menopause.
Results: Eighty-two women with treat-ment-induced menopause completed the MENQOL, EORTC-C30, and the SVQ. The MENQOL was shown to have good face and content validity, and acceptable reliability (homogeneity and test-retest) and validity (concurrent and construct). Additionally, 85.5% of the women reported experiencing hot flashes. However, the most bothersome symptoms were weight gain and fatigue (feeling worn out).
Implications: The MENQOL can be used to assess treatment-induced menopausal symptoms in women diagnosed with breast or gynecologic cancer.
iNtrODuctiON
C
ancer and its treatment have more than a physical impact. There are social, emotional, psychological and spiritual consequences, as well (Fitch, 2000). For women diagnosed with gyne-cologic or breast cancer, treatment-in-duced menopause may be a side effect they experience that has varying degrees of influence on quality of life (Davis, Zinkand & Fitch, 2000). It is important from a clinical perspective to be able to measure the impact of treatment-in-duced menopause with a standardized instrument that is reliable and valid for the population of concern. Standardized measurement allows for easy monitor-ing and documentation over time, as well as identification of individuals expe-riencing side effects and quality-of-life issues. In this study, we completed a psy-chometric evaluation of the Menopause-Specific Quality of Life Questionnaire (MENQOL) in women experiencing can-cer treatment-induced menopause.BAcKGrOuND
With natural menopause, the ova-ries can continue to produce testoster-one and androsteneditestoster-one, which are converted to estrogen peripherally, for many years after menopause (Society of Obstetricians and Gynecologists of Cancer, 2006). Treatment-induced menopause, whether from surgery, radi-ation, or chemotherapy, is unlike natu-ral menopause in that ovarian sources of androgen and estrogen are removed pre-maturely and simultaneously. Hormone levels fall dramatically, typically result-ing in more sudden and severe symp-toms (North American Menopause Society [NAMS], 2007).
Pre-menopausal women receiving treatment for gynecologic cancers (ovar-ian, endometrial, cervical, vulvar) may experience ovarian failure, as a result of
surgery (bilateral oophorectomy) or pel-vic radiation. Women of any age remain permanently amenorrheic at radiation doses of 500–600 cGy to the ovaries. Pre-menopausal women treated for breast cancer may develop ovarian premature menopause, as a result of chemother-apy, ovarian oblation, or estrogen-block-ing drugs (MacDonald, 2007). The effect of chemotherapy on the ovaries is a result of age, cumulative dose, and type of agent (Feldman, 1989). Between 63% and 83% of women menstruating before being treated experience menopause following treatment (Davis, Zinkand, & Fitch, 2000).
Women experiencing treatment-in-duced menopause may have more frequent and/or severe vasomotor symptoms, less energy, insomnia, vag-inal atrophy, hypoactive sexual desire, depression, mood swings, migraine headaches, vaginal dryness, cardiac symptoms, and earlier onset of osteo-porosis. Complaints of cognitive dys-function are not unusual in women experiencing treatment-induced meno-pause (NAMS, 2007). Each of these changes could have an influence on dimensions of quality of life including relationships, sexual satisfaction, self-es-teem, emotional well-being, and capac-ity to return to work (Davis et al., 2000; Howell, Fitch & Deane, 2003; Cebeci, Yangin, & Tekeli, 2010; Abbott-Anderson & Kwekkeboom, 2012).
It is important in clinical practice to be able to easily identify the women who are feeling a significant impact on their quality of life because of treatment-in-duced menopause. Having a reliable, valid measurement tool is imperative for consistent and correct assessment, as well as monitoring over time. The tool must also be easily administered and scored if it is to be used in a busy ambulatory setting, with relatively little response burden for the women.
ABOut tHe AutHOrs
Catherine Dolye, RN, BScN, Primary Nurse* Lauran Adams, RN, BScN, Primary Nurse* Alison McAndrew, BSc, RAP, Research Coordinator*
Stephanie Burlein-Hall, RN, BScN, Med, Advanced Practice Nurse*
Tracey DasGupta, RN, MN, CON(C), Advanced Practice Nurse*
Jennie Blake, MD, MSc, FRCSC, Consulting Gynecologist*
Margaret Fitch, RN, MScN, PhD,*,** Nurse
Researcher
*At the time of the study, all were staff members at the Odette Cancer Centre, Sunnybrook Health Sciences Centre
FEA
TUR
ES
/R
U
bR
iq
UE
S
The Menopause-Specific Quality of Life (MENQOL) was developed as a tool to measure the impact of menopause in a population of women who had under-gone menopause as a natural life event (Hilditch et al, 1996). Women with a diagnosis of gynecologic or breast can-cer may undergo treatment that will, in all likelihood, render them menopausal. It has been suggested that side effects and symptoms are more intense when menopausal change happens as a treat-ment-induced alteration. Additionally, cancer treatment may have other side effects and can have an impact on qual-ity of life. This could result in elevated scores on a measurement tool that had originally been developed and validated for a general population. In particular, somatic items could contribute to the elevation of scores. Thus, it is import-ant to validate the tool in a cancer pop-ulation prior to its use in routine clinical practice.
PurPOse
This study was undertaken to vali-date the psychometric properties of the condition specific MENQOL for use in a population of women treated for gyne-cologic or breast cancer who had expe-rienced treatment-induced menopause. Once validated for this population, the instrument can be used as a foundation for providing person-centred care and tailoring the care for women suffering from this potentially debilitating side effect.
MetHODs
A psychometric evaluation of the MENQOL was designed to assess its reliability (test-retest and internal con-sistency) and validity (face, content and construct) in women treated for gyne-cologic or breast cancer who had expe-rienced treatment-induced menopause. Eligibility criteria included the follow-ing: a confirmed diagnosis of gyneco-logic or breast cancer, minimum of two months post-cancer treatment, experienced menopause as a result of the cancer treatment, 18 years of age or older, and able to read and under-stand English. If women had had hor-mone replacement therapy within the three months prior to accrual they were
excluded from participation. Ethics approval for this study was granted by the Sunnybrook Health Sciences Centre Research Ethics Committee.
As a first step, face and content validity were evaluated. Face validity was assessed by purposefully select-ing 10 women who met the eligibil-ity criteria and asking them to review the MENQOL instrument and answer the following question: “How well do you think this tool measures a wom-an’s quality of life after she has experi-enced treatment-induced menopause because of her cancer treatment?” Their response was recorded on a five-point response Likert item. The response options ranged from “0” (not at all) to “4” (very well). Any additional com-ments the women wanted to offer about the tool and the instruments were cap-tured in an open comment item.
Content validity was assessed by pur-posefully selecting a cross-section of 10 experts in cancer care, quality of life, gynecologic or breast cancer treatment, or treatment-induced menopause and asking them to review the MENQOL. Specifically, they were asked to indi-cate if there were important omissions or inappropriate items when consid-ering the gynecologic or breast cancer populations and treatment-induced menopause. If a particular item was recommended for addition or deletion, participants were asked to record the reasons for their suggestions.
Reliability and construct validity were assessed in a sample of 82 women attending a large ambulatory compre-hensive cancer clinic. A cross-sectional, convenience sample of women who met the eligibility criteria was accrued and asked to complete measurement instru-ments on two occasions. At Time 1, women completed the following instru-ments: Demographic Items (age, time since diagnosis, education, marital sta-tus, treatment), MENQOL (menopause quality of life), EORTC-C30 (quality of life), JENSEN – SVQ (sexual function – vaginal changes questionnaire), and visual analogue scale for hot flushes. Two weeks later, at Time 2, the women completed the MENQOL and the visual analogue scale for hot flushes. The sam-ple size was determined on the basis of
requiring at least five participants per questionnaire or scale item (Nunnally, 2006).
Women were accrued in an outpa-tient cancer clinic. Initially, a nurse approached the women and determined if they met the criteria for the study and were interested in speaking with the research co-ordinator. The co-or-dinator followed up with any women who expressed interest in the study, described what participation involved, and sought their informed consent. Women completed the questionnaires in the clinic during the appointment time. Those who were not scheduled to return to the clinic in the two-week time interval were given an envelope to take home containing the MENQOL, visual analogue scale, and a prepaid return envelope in which to mail back their completed second survey.
Instruments
The Menopause-specific Quality of life Questionnaire (MENQOL) was designed as a condition-specific qual-ity-of-life measure (Hilditch, Lewis, Peter, van Maris, Ross, Franssen et al, 1996). It is a self-administered instru-ment designed from the experiences of women with menopause. It was devel-oped with women two to seven years post-menopause with a uterus and not currently on hormone replacement therapy. It has 30 items covering four domains (i.e., subscales of vasomotor, physical, psychosocial, sexual) and a global quality-of-life question. Each item asked the women whether or not they are experiencing a specific issue (e.g., hot flushes, difficulty sleeping, vaginal dryness) and, if so, how bothered are they by it (rated on a Likert-type scale of 0–6). Face content and construct validity have been reported in addition to reli-ability and responsiveness. Test-retest coefficients for subscales ranged from 0.55 to 0.81 and the scale discriminates between women according to their qual-ity of life.
FEA
TUR
ES
/R
U
bR
iq
UE
S
was field tested in a cross-cultural sam-ple of lung cancer patient in 13 coun-tries to confirm the hypothesized scale structure, to establish reliability, and to evaluate validity (Aaronson, Ahmedzai, Bergman, Bullinger, Cull, Duez, et al., 1993). The QLQ-C30 version 3.0 incor-porates five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain, and nausea/vomiting), a global health sta-tus/QLQ scale, and a number of single items assessing additional symptoms commonly reported by patients (dys-pnea, loss of appetite, insomnia, con-stipation and diarrhea) and perceived financial impact of the disease.The Sexual Function-Vaginal Changes Questionnaire (SVQ) was developed to investigate sexual and vaginal problems in gynecologic can-cer patients (Jensen, Klee, Thranov, & Groenvold, 2004). The instrument con-sists of 20 core items measuring sexual interest, lubrication, organism, dyspa-reunia, vaginal dimensions, intimacy, sexual problems of the partner, sexual activity, sexual satisfaction, and body image. The SVQ was validated in two cancer patient groups (n=75 gynecologic cancer patients; n=257 cervical cancer patients). Ten items are applicable to all patients and create three subscales (inti-macy, sexual interest, global sexual satis-faction). For sexually active respondents, two additional subscales are incorpo-rated to measure the level of problems with vaginal changes and sexual func-tioning. The internal consistency of the subscales is reported to range from 0.76 to 0.83 (Cronbach’s alpha) and construct validation has been reported.
The visual analogue scale for hot flushes and the demographic item ques-tions were designed for the purposes of this study. The visual analogue scale is a 10 cm line on which the respondent can indicate the intensity of hot flashes in the past 24 hours. The demographic items include age, time since diagnosis, education, marital status, and treatment received.
Analysis
The analysis for the face validity assessment consisted of collating and summarizing the responses on the Likert-type and open-ended comment
items. The analysis for the content valid-ity assessment consisted of collating and summarizing the responses regarding each item and any written responses.
The internal consistency of the MENQOL was determined by calcu-lating Cronbach’s alpha for the respec-tive subscales using the data from Time 1. Test-retest reliability was calculated by applying Pearson Product Moment Correlation Formula to the MENQOL subscale data from Time 1 and Time 2. The linear analogue data were used to determine if significant changes occurred in the intensity of hot flushes between Time 1 and Time 2 (paired t-test).
Construct validity was assessed using data from Time 1. The results from the MENQOL subscales were compared to the respective results from the sub-scales of the EORTC-C30 and the SVQ using the Pearson Product Moment Correlation Formula. It was anticipated that the performance on the subscales for the MENQOL would be similar to the performance on the same subscales for the other two instruments and the calculated correlation coefficients would be significant (p< 0.05).
results
Sample
A total of 82 women participated in this study. The mean age was 47 years with a range of 29 to 58. Seventy-three percent were married and 73% had completed college or university educa-tion. On average, the women were 33 months post diagnosis. Three-quarters (75.6%) had been diagnosed with breast cancer and the remainder with gyneco-logic cancer. Sixty percent had received chemotherapy, 37% had received radi-ation therapy, and 17% had undergone
surgery. All had experienced meno-pause, as a result of their cancer treat-ment. The mean rating on the hot flushes linear analog was 2.09 (stan-dard deviation = 0.95) at Time 1 and 1.98 (standard deviation = 0.87) at Time 2. No statistical difference was observed.
Face Validity
The ten women who reviewed the MENQOL indicated all items were appropriate in light of their experiences with treatment-induced menopause. None of the women suggested making any changes in the instrument. They rated all items as highly relevant to their experiences with treatment-induced menopause.
Content Validity
The 10 healthcare professionals who reviewed the MENQOL included the disciplines of gynecology, gyneco-logic oncology, general practice (fam-ily medicine), and oncology nursing. Each individual indicated the items on the instrument were appropriate and relevant to the situation based on their clinical experience. No one suggested removing or adding specific items.
Reliability
Cronbach’s alphas (internal consis-tency) for the four MENQOL subscales ranged from 0.88 (sexual) to 0.94 (vaso-motor). The test-retest coefficients for the four subscales were calculated using Pearson Product Moment Correlation Formula. The coefficients range from 0.79 (sexual) to 0.89 (vasomotor) and all were significant at p< 0.0001 (See Table 1). The visual analogue item mea-suring hot flushes at Time 1 and Time 2 showed no significant differences (t=1.35, p=0.182) for the two-week time interval.
Table 1: Mean scores and reliability results for MENQOL subscales MENQOL Subscale
(bother) Mean (Standard Deviation) Cronbach’s alpha coefficient*Test-retest
Vasomotor 4.76 (2.2) 0.94 0.89
Sexual 4.20 (2.5) 0.88 0.79
Physical 3.84 (1.5) 0.91 0.84
Psychosocial 3.82 (1.8) 0.89 0.81
FEA
TUR
ES
/R
U
bR
iq
UE
S
Mean Subscale Scores
The mean bother scores on the MENQOL subscales ranged from 3.82 to 4.76 (highest score is 6). The highest mean bother scores were reported for the vasomotor (4.8) and sexual (4.2) sub-scales (see Table 1). The mean function-ing scores for the EORTC-C30 subscales ranged from 62.1 (emotional) to 86.3 (physical) (see Table 2) (highest scores equal 100). High scores on the EORTC instrument reflect a better level of func-tioning. The mean scores for the SVQ subscales ranged from 40.1 (sexual sat-isfaction) to 67.7 (sexual interest) (See Table 2). Higher scores on these sub-scales reflect fewer difficulties.
Construct Validity
Moderate to high correlations between the physical and psychosocial subscales of the MENQOL and each of the EORTC-Q30 subscales and global health scores were statistically signifi-cant and showed negative relationships, as predicted (see Table 3). Correlations were also significant between the EORTC-Q30 emotional and social sub-scales and the vasomotor and sexual MENQOL subscales. The cognitive EORTC-Q30 subscale and the MENQOL sexual subscale were also correlated.
Moderate positive correlations were observed between the physical and psy-chosocial MENQOL subscales and the Intimacy SVQ subscale (see Table 4). The MENQOL sexual subscale was pos-itively correlated with the sexual interest subscale and global sexual satisfaction, but was negatively correlated with vag-inal changes scores on the SVQ. The physical subscale score on the MENQOL was also positively correlated with global sexual satisfaction and negatively with vaginal changes.
Frequently Identified Issues for Women
The most frequently identified issues experienced by women and endorsed on the MENQOL are listed in Table 5. Experiencing hot flashes (85.5%) and feeling tired or worn out (82.1%) were endorsed most frequently. However, issues listed as most bothersome were slightly different. Table 6 lists the top five issues rated as most bothersome for the women in this study. Physical issues (weight gain, fatigue, difficulty sleeping)
Table 2: Mean scores for EORTC-Q30 function subscales and SVQ subscales EORTC-Q30 function subscales: mean
(standard deviation) SVQ Subscales Mean (standard deviation)
Physical 86.3 (16.0) Intimacy 44.6 (17.6) Role 83.9 (26.1) Sexual interest 67.7 (28.8) Social 70.7 (31.9) Sexual satisfaction 40.1 (11.4) Cognitive 68.1 (27.7) Vaginal changes 50.18 (22.6) Emotional 62.1 (28.8) Sexual function 42.5 (16.4)
Table 3: Correlation between MENQOL and EORTC-Q30 subscales EORTC-C30
Subscale MENQOL Vasomotor MENQOL Sexual MENQOL Physical Psycho-socialMENQOL
Physical -0.21 -0.15 -0.56***** -0.42***** Emotional -0.28* -0.27* -0.61***** -0.84***** Cognitive -0.17 -0.24* -0.62***** -0.64***** Social -0.29** -0.34**** -0.59***** -0.70*****
Role -0.22 -0.27 -0.42***** -0.31*
Global health -0.14 -0.20 -0.51***** -0.54*** *p<0.05; **p<0.01; ***p<0.005 ****p<0.002; *****p<0.0001
Table 4: Correlation between MENQOL and SVQ subscales
SVQ Subscales MENQOL
Vasomotor MENQOL Sexual MENQOL Physical PsychosocialMENQOL
Intimacy+ -0.03 0.06 0.38**** 0.44***** Sexual interest+ 0.11 0.33*** 0.20 0.20 Global sexual satisfaction+ 0.14 0.30** 0.29* 0.16 Vaginal changes- -0.11 -0.66***** -0.35* -0.12 Sexual function+ 0.21 0.33 0.21 0.27 *p<0.05; **p<0.01; ***p<0.005 ****p<0.001; *****p<0.0001
Table 5: Most frequently identified issues (endorsed on MENQOL)
Item %
Hot flashes or flushes 85.5
Feeling worn out or tired 82.1
Difficulty sleeping 81.8
Sweating 77.2
FEA
TUR
ES
/R
U
bR
iq
UE
S
and sexual changes (interest in sex and intimacy) were the issues that resulted in the highest bothersome ratings.
DiscussiON
This work was undertaken to vali-date the psychometric properties of the condition-specific MENQOL for use in a population of women treated for gyne-cologic or breast cancer who had expe-rienced treatment-induced menopause. The MENQOL had been designed for use in a general population of women experiencing menopause and needed to be tested in a population of women experiencing menopause as a result of their cancer treatment.
The face and content validity testing did not result in any changes to the tool prior to its application with women in the cancer population. The women who participated had all experienced treat-ment induced menopause. Their clini-cal picture had not changed between the Time 1 and Time 2 completions of the instruments.
The reliability of the MENQOL was acceptable in this cancer population and reflects what has been reported in other psychometric evaluations of this instrument (Hilditch et al., 1996;
Radtke, Terhorst & Cohen, 2011). Both homogeneity and test-retest coeffi-cients reached high levels. Additionally, the construct validity reflected antic-ipated results. There were acceptable correlations between the subscales on the EORTC-30 and MENQOL physi-cal and psychosocial subsphysi-cales, but not the sexual and vasomotor. This makes sense in light of the identified need to develop the SVQ as a scale to augment the EORTC-30 for use in relation to sex-uality issues. Comparison of the SVQ and the MENQOL showed good agree-ment only between SVQ Intimacy and the MENQOL Physical and Psychosocial subscales. Other subscales are not over-lapping. In particular, the MENQOL is unique in measuring the vasomotor items.
Clearly this group of women who had treatment-induced menopause were experiencing a range of symp-toms and changes in their sexual lives, which is similar to reports by other investigators (Davis et al., 2000; Cebeci, Yangin, & Tekeli, 2010; Abbott-Anderson & Kwekkeboom, 2012). More than three-quarters reported experiencing hot flashes and fatigue. It is of interest, however, that the most frequently expe-rienced issues were not the same as
the issues that were most bothersome. Exactly what makes an issue bothersome bears further investigation. One could imagine that certain clusters of symp-toms could be especially bothersome, as well. For example, the combination of night sweats, difficulty sleeping, and feeling tired could be seen as a cluster of symptoms that might have an increased effect over experiencing a single one.
iMPlicAtiONs
The primary implication from these study results is that the MENQOL can be considered reliable and valid for use in the cancer population of women expe-riencing treatment-induced menopause. The instrument can be used in daily practice, as a way of providing a base-line assessment when a women experi-ences treatment induced menopause. It could also provide a way of monitoring changes over time.
The other implication from this work is the need for women diagnosed with cancer and their care providers to be aware of the myriad of symptoms that could be experienced with treatment-in-duced menopause. Oncology nurses are in an ideal position to help women be prepared for what they might expect and to manage with the subsequent issues. Ongoing communication with patients will about the topic will help to identify any issues early.
Future research ought to focus on finding effective interventions for the symptom clusters women experience with treatment-induced menopause. Ideally it would be preferable to find ways to prevent the issues; but if that is not possible, then early identification and intervention would improve the patient experience.
reFereNces
Aaronson, N.K., Ahmedzai, S., Bergman, B., Bullinger, M., Cull, A., Duez, N.J. et al. (1993). The European Organization for Research and Treatment of Cancer QLQ-C30: A quality of life instrument for use in international clinical trials in oncology.
Journal of the National Cancer Institute, 85,
365–376.
Abbott-Anderson, K., & Kwekkeboom, K.L. (2012). A systematic review of sexual
concerns reported by gynecological cancer survivors. Gynecologic Oncology, 124(3), 477–489. http://doi.org/10.1016/j. ygyno.2011.11.030
Cebeci, F., Yangin, H.B., & Tekeli, A. (2010). Determination of changes in the sexual lives of young women receiving breast cancer treatment: A qualitative study.
Sexuality and Disability, 28(4), 255–264.
http://doi.org/10.1007/s11195-010-9169-1
Davis, C., Zinkand, J., & Fitch, M.I. (2000). Cancer treatment induced menopause: Meaning for breast and gynecological cancer survivors. Canadian Oncology
Nursing Journal, 10(1), 14–21.
Feldman, J.E. (1989). Ovarian failure and cancer treatment: incidence and interventions for the premenopausal woman. Oncology Nursing Forum, 16,
651–657.
Table 6: Items identified as most bothersome for women (endorsed on MENQOL)
Item Mean* Standard Deviation
Weight gain 4.44 1.937
Change in your sexual desire 4.32 1.942
Avoiding intimacy 4.11 2.162
Feeling tired or worn out 4.04 1.527
Difficulty sleeping 4.02 1.944
FEA
TUR
ES
/R
U
bR
iq
UE
S
Fitch, M. (2000). Supportive care for cancer patients. Hospital Quarterly, 3(4), 39–44. Hilditch, J.R., Lewis, J., Peter, A. van
Maris, B., Ross, A., Franssen, E. et al. (1996). A menopause-specific quality of life questionnaire: Development and psychometric properties. Maturitas, 24,
161–175.
Howell, D., Fitch, M.I., & Deane, K.A. (2003). Women’s experiences with recurrent ovarian cancer. Cancer Nursing, 26(1), 10–17.
Jensen, P.T., Klee, M.C., Thranov, I., & Groenvold, M. (2004). Validation of a questionnaire for self-assessment of sexual function and vaginal changes after gynecologic cancer. Psycho-oncology, 13,
577–592.
MacDonald, P. (2007). Update on treatment for breast cancer. Practice Nurse, 33(3), 39–41.
North American Menopause Society (NAMS). Menopause practice: A clinican’s
guide (3rd ed.). Cleveland, OH: Author.
Nunnally, J.C. (2006). Psychometric Theory.
New York: McGraw-Hill.
Radtke, J.V., Terhorst, L., & Cohen, S.M. (2011). The Menopause Specific Quality of Life (MENQOL) Questionnaire: Psychometric evaluation among breast cancer survivors. Menopause, 18(3), 289– 295. doi:10.1097/qme.Ob013e3181ef975a Society of Obstretrics and Gynecologists of
Canada. (2006). Canadian consensus conference on menopause update.
Journal of Obstetrics and Gynecoloy
