This cross-sectional, descriptive and analytical study was conducted on women from Kermanshah and Kurdistan provinces, Iran, from September 2020 to June 2021. Participants in this study were selected based on the multistage sampling methods and networked. First, a list of towns and villages in the provinces of Kermanshah and Kurdistan was prepared. Then, a quota was assigned to each province based on the number of cities, so three cities from Kurdistan province and four from Kermanshah province with higher population were selected. Finally, Sanandaj, Marivan and Saqez from Kurdistan province and Kermanshah, Islamabad Gharb, Sarpol Zahab and Javanroud from Kermanshah province were selected for sampling. Midwives working in urban health centers in each city who were willing to participate in this study were informed about the questionnaire and how to complete it. After the multi-stage online sampling, the link of the designed questionnaire was provided to the midwives online via WhatsApp. After informing them of the research objectives and procedures, the selected midwives then provided the link to the patients eligible to participate in the study. In cases where participants could not complete the online form, the midwife asked participants questions and the answers were entered into the online form. Sampling was carried out at this stage by the convenience sampling method.
Inclusion criteria were 18 to 70 years old, no history of ovarian cancer and willingness to participate in the study. The sampling process continued until the planned sample size was reached. Participants were asked to complete the questionnaire after expressing their informed consent. The questionnaire was created in Google Forms and the link was sent to participants via WhatsApp. The first section of the online questionnaire informed participants in detail about the objectives and how to complete them. Participants were also assured that their information would be kept confidential. The sample size was calculated using the formula n = (Z1−α/2+Zβ)2 p(1−p)/d2 assuming a type 1 error of 0.05, a power of 80% and a proportion of women with excellent knowledge of ovarian cancer (p) of 25% . The calculated sample size was 918 (d: 0.04); since approximately 30% of the questionnaires would be incomplete, the sample size was 1150. Cancer Awareness Measure (CAM) is a scale developed in the UK to help measure cancer awareness, identify risk factors related to poor awareness, and developing and evaluating interventions to promote cancer awareness. There are different versions of this questionnaire valid for different cancers applicable to face-to-face, online or telephone interviews and self-administration [10, 15]. This questionnaire included 35 items in “warning signs” (10 multiple-choice items and one open-ended item), “delay in seeing a doctor” (1 open-ended item), “age of ovarian cancer” (one multiple-choice item), “risk factors” (12 multiple-choice items and one open-ended item), “NHS screening programs” (8 items) and “confidence in diagnosing symptoms of ovarian cancer” (a multiple-choice item).
The questionnaire was independently translated from English to Persian (forward translation) by two experts initially, after obtaining permission from the Ovarian Cancer Awareness Measure (OCAM). The research team then combined the two translated versions to create a single copy. Two specialists separately translated the final form into English (backward translation) and the research team merged the two English translations into a single copy. . The tool was then reviewed by 10 experts (all with PhDs in reproductive health and health education and knowledgeable and experienced in instrument measurement and ovarian cancer) to provide their quality-correcting comments. of the grammar and vocabulary of the text, the arrangement of elements and the notation system. This version of the questionnaire was tested in a pilot study with ten women qualified for the study, and final modifications were made based on participant feedback. Questionnaire items were reviewed and modified for cultural adaptation to eliminate any erroneous items for Iranian culture. Open-ended questions were eliminated from the questionnaire due to the lack of proper facilities for face-to-face interviews. The open-ended item on care-seeking practices was converted to a multiple-choice item to facilitate responses. Response options included instantly, a week, two weeks, as soon as possible, a month, a few months later, and never. In this study, a modified version of OCAM was used. The questionnaire consisted of two sections; the first section included items on demographic information such as language, age, marital status, employment status, level of education, place of residence, and having a parent, friend or family member with cancer (taking into account the type of cancer). The second part contained items measuring “warning signs” (10 multiple-choice items), “delay in seeing a doctor” (one multiple-choice item), “age of ovarian cancer” ( one multiple-choice item), “risk factors” (12 multiple-choice items) and “confidence in the diagnosis of symptoms of ovarian cancer” (one multiple-choice item) as well as four items concerning national programs sensitization screening. Each correct answer received a score of 2 to determine awareness of symptoms, risk factors, typical age of ovarian cancer, and national screening programs. A score below 16, between 16 and 36, and above 36 was considered low, moderate, and adequate knowledge of ovarian cancer, respectively. The division is expressed by poor knowledge of people in the lower 33% of the population score, between 33 and 66% of the moderates and above 66% of the good score in the population studied.
Ten experts (reproductive health and health education specialists) were invited to assess the validity of the qualitative and quantitative content of the questionnaire. The questionnaire was then modified based on their feedback on grammar, vocabulary, necessity, syntax, collocation and scoring system. The Likert scale was used to assess the relevance, simplicity, clarity and necessity of each item, and the content validity ratio (CVR) was calculated quantitatively. The average CVR was 0.88, while the Content Validity Index (CVI) was more significant at 0.79. Internal correlation and test-retest reliability were used to assess the stability of the questionnaire. The most common method used to determine the internal correlation is Cronbach’s alpha, which ranges from 0 to 1. It should have an internal correlation more significant than 0.7 . Cronbach’s alpha for the entire questionnaire was 0.88, indicating the appropriate reliability of this tool for the Iranian population. In the retest technique, 20 qualified participants completed the questionnaire twice 2 weeks apart. The intraclass correlation coefficient (ICC) was used to assess the test-retest reliability of the instrument, which was calculated as 0.86. An ICC of 0.8 or greater indicates excellent stability . Quantitative variables in this study were defined by mean, standard deviation and interquartile range, while qualitative variables were presented by frequency and percentage. Percent sensitization was shown using bar charts and ordinal logistic regression was used to determine predictors of sensitization. Data were statistically analyzed in SPSS-19 at the 0.05 level of significance. Rank logistic regression was used in this study to measure the response variable, i.e. knowledge (good, moderate and poor). To predict higher odds of knowing age, education level, cancer history (reproductive tract, breast and bowel in family, friends and close relatives), dialect and marital status were added separately and then simultaneously to the model (ordinal logistic regression).