Study design
This study is part of a larger joint research project between Switzerland and Iran that aimed at developing a PSQLS in both countries following similar methodologies. In this article, we report the findings based on the project phases that were conducted in Iran.
We employed a two-phase mixed-method approach to develop a scale to assess the postpartum sexual quality of life. In the first phase, we conducted a qualitative study to generate items and scale development and in the second phase, we assessed the psychometric properties of the scale. The qualitative results from the first phase of the study are being published elsewhere; in this article, we present the results of the psychometric assessment of the PSQLS. The joint research project was approved by National Institute for Medical Research Development (NIMAD) in Iran and Zurich committee in Switzerland. All participants completed a written informed consent that described the study phases, the processes used to ensure their confidentiality, and that participation was voluntary.
Participants
Participants were recruited through healthcare centers affiliated with Tehran University of Medical Sciences. The healthcare centers provided us with a list of names of postpartum women and their contact information. One of the research assistants on the team contacted the women and screened them for eligibility. The inclusion criteria for eligibility were: primiparous mothers, older than 18 years old, being married, having Iranian nationality, experienced a healthy pregnancy, and low risk vaginal birth (without complications and large episiotomies) or cesarean section, during the last three months. The latter because studies have shown that sexual dysfunction is most prevalent during the first three months after delivery. Mothers who had a negative experience such as the death of a close relative or friend during the last 6 month, reported a history of psychological diseases or sexual dysfunctions before pregnancy, and felt uncomfortable talking about sex were excluded from the study to minimize the effect of confounding variables on sexual quality of life. Specific procedures and analytic approaches for each phase of the study are described below.
Phase 1: Qualitative phase and item generation
The qualitative phase was conducted from December 2018 to March 2019. We employed both individual interviews and focus group discussions for data collection. Our final sample consisted of two focus group discussions (six participants) and 17 individual interviews.
Data collection was conducted in a private space at one of the healthcare centers (for the focus group discussions), or at the home of the participant (for individual interviews based on participants’ preference). All participants completed a demographic survey (age, level of education, economic status, occupational status, and mode of delivery) at the beginning of the focus group discussions or individual interviews. We used the same semi-structured interview guide with open-ended questions for both focus group discussions and individual interviews. Data collection was conducted in Farsi. The interviewers, two research assistants on the team, received training in qualitative interviewing and the use of optional probes that were included for situations when additional information was needed. The interviews were audio-recorded, lasted 14–65 min and were transcribed verbatim.
Examples of interview questions included: can you describe the quality of your sexual life after giving birth? What changes have you experienced after delivery? What are some important factors that you think affected these changes? Can you describe the quality of the non-sexual relationship with your husband/partner? We also asked additional questions regarding participants’ beliefs and feeling regarding themselves, their sexual dominant beliefs, and challenges and problems that may affect their sexual quality of life after having given birth.
Maximum variation sampling was conducted to reach greater transferability and data saturation. We therefore selected participants from different age groups, various educational levels, and different socioeconomic statuses. After two focus group discussions and 17 individual interviews, data saturation was achieved.
We used Graneheim and Lundman’s approach of qualitative content analysis for analyzing the data (19). MAX-QDA 10 software was used for data analysis. A coding scheme was developed from the first five interviews and then used to code the remaining individual interviews and focus group discussions. Following the completion of coding and extraction of the final themes, Likert-type survey items were generated for each of the themes and sought to capture both covert and overt manifestations of postpartum sexual quality of life. The study team reviewed and modified survey items to enhance clarity and minimize redundancy. We viewed these items as representing the maximum known scope of participants’ thoughts and feelings regarding their postpartum sexual quality of life in the targeted domains. After finalizing qualitative analysis and extracting the items, we conducted a literature review to evaluate whether important items may have been missed. This resulted in the addition of seven items; however, none of these items were retained in the final version of the scale. Finally, an item pool with 67 items was generated and used for psychometric evaluation.
Phase 2: Psychometric evaluation of the PSQLS
The pre-final draft of the PSQLS contained 67 items and each item was rated on a five-point response scale (completely agree to completely disagree). Eligibility criteria, recruitment strategies, and setting were the same as those described in the first phase. Since sample size equal to or greater than 200 is considered as enough for exploratory factor analysis (20), we approached 300 mothers for this phase. Two hundred and eighty-two eligible mothers approved to participate and they completed the self-administered survey at a private space at one of the healthcare centers.
Analytic Approach
Several statistical methods including validity and reliability were used to analyze the data:
Validity
In this phase, face, content, and construct validity were assessed as follow:
Face Validity.
Qualitative and quantitative methods were used to evaluate face validity. Ten eligible participants were asked to evaluate the questionnaire and assess its difficulty level for the qualitative face validity. For the quantitative face validity, the impact score (Frequency × Importance) was calculated to indicate the percentage of participants (n = 10) who identified the item as important or quite important. Items were considered appropriate if they had an impact score equal to or greater than 1.5 (which corresponds to a mean frequency of 50 % and mean importance of 3 on the 5-point Likert scale) (21).
Content Validity.
We applied qualitative and quantitative content validity methods. An expert panel consisting of 12 investigators specialized in sexuality and psychometry assessed the content validity of the questionnaire. Qualitatively, they evaluated wording, grammar, item allocation, and scaling of the questionnaire. Quantitatively, the content validity index (CVI) and the content validity ratio (CVR), were calculated. Clarity, simplicity, and relevancy of each item were assessed by CVI evaluation (22). To assess CVI, we used a Likert-type, ordinal scale with four possible responses. The responses contained a rating from 1 = not relevant, not simple, and not clear to 4 = very relevant, very simple, and very clear. CVI was calculated as the proportion of items that received a rating of 3 or 4 by the experts (23). The essentiality of items was tested by calculating CVR where the experts rated each item as essential, useful but not essential, or not essential (24).
Construct Validity.
Exploratory factor analysis (EFA) was performed (n = 282) to determine the underlying constructs of the questionnaire. A principal component analysis (PCA) with varimax rotation was applied and the factor loading equal to or greater than 0.4 was considered acceptable (25).
Reliability.
Cronbach’s alpha coefficient was calculated to assess the internal consistency of the questionnaire. Values equal to or greater than 0.70 were considered satisfactory (26). To assess the questionnaire’s stability, we conducted a test-retest reliability to estimate the intraclass correlation coefficient (ICC). Twenty-five participants completed the questionnaire twice with two-week intervals. ICC values of 0.40 or higher were considered satisfactory (r ≥ 0.81 - 1.0 as excellent, 0.61 - 0.80 very good, 0.41 - 0.60 good, 0.21 - 0.40 fair, and 0.0 - 0.20 poor) (27).