Validation of the Vulvovaginal Symptom Questionnaire in Indonesian Women with Postmenopausal Syndrome using Confirmatory Factor Analysis

Study design

This study employs a quantitative cross-sectional design and will be conducted at community health centers (Puskesmas) in Jakarta, Indonesia.

Data collection

Research was conducted involving 95 postmenopausal women to evaluate the clarity, relevance, and comprehensiveness of the Vulvovaginal Symptom Questionnaire (VSQ). Data were collected through direct primary data collection, where participants completed the questionnaire in person. The process took approximately 30 minutes per participant.

Participants

The target population includes postmenopausal women living in Jakarta. Convenience sampling will be used to recruit 95 participants who meet the following Inclusion criteria: Postmenopausal women who can read, write, and speak Indonesian. Exclusion criteria: were women with a history of mental health disorders. The descriptive statistics of the respondents’ demographic characteristics are summarized in Table 1. Regarding employment status, the majority of the respondents were not employed, comprising 61.96% (n = 57), while 36.96% (n = 34) were employed. In terms of education level, a significant portion of respondents, 55.43% (n = 51), had completed secondary education, followed by 40.22% (n = 37) who had attained a tertiary level of education. Only a small percentage, 2.17% (n = 2), reported having primary level education. For pregnancy sequence, the largest group of respondents was in their first pregnancy, making up 36.96% (n = 34), followed by those in their second pregnancy at 32.61% (n = 30). A smaller proportion of respondents were experiencing their third pregnancy (20.65%, n = 19), while 8.70% (n = 8) were in their fourth pregnancy. These data reflect the demographic diversity among the sample population, highlighting variations in employment, education, and pregnancy experience.

Instruments

The Vulvovaginal Symptom Questionnaire (VSQ) translated into Indonesian using the forward and back-translation method, as outlined by Sousa and Rojjanasrirat (2011). The translation process consisted of several steps. First, two independent translators will perform forward translation, translating the VSQ from English to Indonesian. Following this, the two translations will be reviewed for clarity and linguistic accuracy, and any discrepancies or ambiguities will be reconciled. The reconciled Indonesian version will then undergo back translation, where two independent translators will translate the Indonesian version back into English. Afterward, the original English version and the back-translated version will be compared to ensure consistency and equivalence in meaning. Finally, a panel of experts, including healthcare professionals and researchers, will review the final Indonesian version to ensure that it is culturally and contextually appropriate for use with Indonesian women. This process will ensure that the VSQ is both linguistically accurate and culturally relevant for the target population.

Development of Indonesian version of VSQ

The Vulvovaginal Symptom Questionnaire (VSQ) is a tool designed to assess the impact of vulvovaginal symptoms on various aspects of a woman’s health, including physical, emotional, social, functional, and sexual well-being. This questionnaire consists of 20 items, with responses recorded as “Yes” or “No,” indicating the presence or absence of specific symptoms. The items are divided into five main dimensions. The Physical/Somatic Dimension includes items 1 to 7, which focus on symptoms such as itching, burning, pain, dryness, discharge, and odor in the genital area. The Emotional/Psychological Dimension comprises items 8 to 10, addressing feelings of worry, despair, and embarrassment associated with these symptoms. The Social/Social Interaction Dimension (items 11 to 13) examines how vulvovaginal symptoms affect social interactions, including difficulties in meeting others and expressing affection. The Functional Dimension is represented by item 14, which explores the impact of these symptoms on daily activities. Finally, the Intimacy/Sexuality Dimension (items 15 to 20) looks into the effect of vulvovaginal symptoms on sexual relationships, including intimacy, pain during intercourse, dryness, and bleeding. The total possible score on the VSQ is 20, with the cut-off point set at 12, indicating a high level of symptom severity. The VSQ has internal consistency reliability of α = 0.90 for the full scale, and α = 0.85–0.90 for each dimension, which is excellent and suitable for use in research.

Data analysis

Data analysis was conducted using the Statistical Package for the Social Sciences (SPSS) version 26 (Morgan et al., 2019) and MPlus version 8.0 (Muthén & Muthén, 2017). Descriptive statistics, including frequency, mean, and standard deviation (SD), were used to summarize the characteristics of the participants.

In the development phase of the Vulvovaginal Symptom Questionnaire (VSQ), each of the 20 items were evaluated using confirmatory factor analysis (CFA) to assess the dimensionality of the questionnaire. CFA is a frequently used analytical method for testing construct validity in the construction of theory-based instruments in psychology, education, and other social sciences (Umar & Nisa, 2020). By using relevant error variance as a parameter within the structural equation modeling framework, the CFA model accounts for the differences between observed scores and true scores (Li, 2016).

The method was commonly used for estimating parameters in CFA models is maximum likelihood (ML). In this case, the use of maximum likelihood was assumed that the observed indicators follow a multivariate normal distribution and are continuous. This assumption is inappropriate for ordinal variables. In such cases, the researcher used the MLMV estimator, which adjusts the distribution to align with the mean and variance of the chi-square distribution with the relevant degrees of freedom. Based on the Type I error ratio from the omnibus model, preliminary findings suggest that the MLMV estimation is more effective than the MLM method. The MLMV test statistic works well in most situations involving extreme non-normality and small sample sizes (Maydeu-Olivares & Shi, 2017). The detailed factor loading coefficients for each VSQ item, along with statistical values and their significance, are presented in Table 2.

Goodness-of-fit indices were used to assess model fit. These indices include the Chi-square test (χ2) and its p-value (≥0.05 = acceptable fit; ≥0.10 = good fit), which is sensitive to sample size. Other fit indices considered in the study were the root mean square error of approximation (RMSEA) ≤ 0.06, which is considered acceptable. The comparative fit index (CFI) > 0.95 indicates a good fit, while 0.90 is considered traditional. Additionally, the Tucker-Lewis index (TLI) > 0.95 indicates a well-fitting model (Hu & Bentler, 1999). Models with RMSEA ≥ 1.0 are considered poor fits. At least three adequacy indices with values within the acceptable ranges were considered in analyzing the goodness of fit of the data.

Confirmatory Factor Analysis (CFA)

The researcher conducted a construct validity test, and the following results were obtained: Degree of freedom (df ) = P-value = 0.0146, Comparative Fit Index (CFI) = 0.386, Tucker-Lewis Index (TLI) = 0.267, Standardized Root Mean Square Residual (SRMR) = 0.113, and Root Mean Square Error of Approximation (RMSEA) = 0.054, with a 95% Confidence Interval of 0.025 – 0.077. The validity test of the VSQ showed that the model fit the data. This is evident from the RMSEA value of 0.054, with the confidence interval ranging from 0.025 to 0.077. Although the values for CFI and TLI are ≤ 0.90, the RMSEA value can serve as an alternative indicator of model fit (Umar & Nisa, 2020).

After confirming that the model fits the data, the next step is to report the obtained parameter values and their significance tests. These are typically presented in the form of a table showing the factor loading coefficients for each item, along with their standard error, z or t values, and probabilities (significance). A t or z value (absolute) greater than 1.96 is usually considered significant at the 5% level, and the factor loading values should be positive.

The table above shows that there are six items that are invalid, with negative factor loading coefficients and T-values <1.96 (Umar & Nisa, 2020). These six items include item 1, item 2, item 3, item 4, and item 5 in the physical/somatic dimension, as well as item 16 in the intimacy/sexuality dimension. The researcher suggests that these six items be revised.