Discussion
The VSQ is a questionnaire designed to measure the disease-specific quality-of-life impact of vulvovaginal symptoms on postmenopausal women. To date, there has been limited information on the presence of vulvovaginal symptoms in postmenopausal women and their impact on quality of life. As standardized methods for evaluating these symptoms are developed, our understanding of prevalence and life impact will progress.
In the absence of a gold standard for the quantification of vulvovaginal symptoms and life impact, we have tested the validity of the VSQ with other validated gynecologic questionnaires and physical examination findings. We based our instrument for measuring vulvovaginal symptoms on Skindex-16. Skindex has been adapted and used in the study of numerous skin conditions, including psoriasis, atopic dermatitis, and leg ulcers. Skindex has also been used to study the life impact of vulvodynia on women. The flexibility of Skindex in studying the life impact of many skin conditions made this the ideal instrument on which to base our questions. We demonstrated fair correlations between the VSQ and other questionnaires measuring other gynecologic diseases in older women, namely, the PFDI-20 and the PFIQ-7. We also demonstrated a fair correlation between the PISQ-12 (a validated instrument for studying sexual function in women with pelvic floor disorders) and the VSQ. However, we found a low correlation between the FSFI and the VSQ. This may be attributed to the different focus of the FSFI on six different aspects of sexual function (desire, arousal, lubrication, orgasm, satisfaction, and pain), whereas the VSQ focuses on pain, bleeding, and dryness with intercourse.
In addition, we assessed vulvar discomfort with cotton swab testing and found a 0.36 correlation between cotton swab testing scores and VSQ scores. Cotton swab testing scores are dependent on a woman's report of mild, moderate, or severe pain when each location on the vulva is touched. Pain can be present even when no visual abnormalities are observed, which is probably why we observed a fair correlation between the cotton swab testing and the VSQ. Consistent with the original Skindex, a woman's report of skin symptoms was not consistently related to a visual observation of physical examination findings. Although there were no women with significant visual physical examination findings without symptoms, many women reported symptoms without physical examination findings.
We diagnosed active Candida vulvovaginitis by (1) symptoms, (2) KOH preparation, and (3) yeast culture; however, we did not use polymerase chain reaction. Recently, polymerase chain reaction has been demonstrated to be a more sensitive method for diagnosing vulvovaginal candidiasis in women with a diagnosis of recurrent vulvovaginal candidiasis.
We found the test-retest reliability of the VSQ to be adequate 2 to 4 weeks after the women initially completed the VSQ. Because skin symptoms are dynamic and the VSQ asks about symptoms "in the past week," it is possible that women's vulvovaginal symptoms changed during the time interval from the initial completion of the VSQ to the second time they completed the questionnaire. Other validated measures of general skin symptoms (not specific for vulvovaginal symptoms) have reduced the time interval for measuring test-retest reliability to 72 hours. We did have lower κ and ICC values than reported on the psychometric testing of Skindex-16. This can probably be attributed to our choice of a longer time interval between the completion of the initial VSQ and the completion of the follow-up VSQ.
This research is limited by the population of women studied. Women were recruited at gynecology practices. Women seeking routine gynecologic care are probably more bothered by gynecologic conditions and probably have a higher prevalence of vulvovaginal symptoms than women who do not seek gynecologic care. We hypothesize that the true prevalence of vulvovaginal symptoms in all postmenopausal women is lower than that observed in this current study population. However, studying the VSQ in a population of women with a high prevalence of vulvovaginal symptoms allowed us to demonstrate the psychometric properties of the VSQ. In addition, we tested the VSQ for face validity among five women seek routine gynecologic care. Further testing of face validity may be necessary among different populations (eg, postmenopausal women not seeking medical care).
Future studies are needed to determine whether dichotomous responses are truly the best answer choices for evaluating these symptoms. In using a Likert scale consistent with the original phrasing of Skindex-16, women were asked about frequency ("How often have you been bothered?"), but not about severity. Future studies will determine if severity of bother ("How much?") might be a preferred question/response pair over the current dichotomous responses (yes/no). Authors have advocated analyzing the most bothersome symptom, in addition to recording the presence of vulvar symptoms, in the study of treatments for vulvovaginal atrophy. We believe that exploring the frequency and the severity of vulvovaginal symptom bother is important, and we expect future research to demonstrate that severity of bother is the most informative answer choice for the VSQ. In addition to asking about the most bothersome symptom, we concur with Simpson and Murphy that quality-of-life impact from vulvovaginal symptoms also needs to be explored, which is why the VSQ includes the emotional, life impact, and sexual impact scales. Finally, we demonstrated that postmenopausal women without symptoms did not have emotional or life impact; therefore, we propose that women who answer "no" to the first seven symptom questions do not need to complete the remaining three scales of the VSQ.