Study design and population

The present case–control study was conducted between January 2022 and March 2022 among 18–49-year-old women living in the suburban areas on Bandar Abbas in Hormozgan Province. This city is located in the 27.19 latitude and 56.28 longitude and the height of 9 m above the sea level. Bandar Abbas has a population of 352,173 making it the largest city in Hormozgan.

The coastal city of Bandar Abbas has commercial affairs with the countries of the Persian Gulf and is considered a destination for job seekers in Iran. Their arrival has significantly increased the population and led to the emergence of slums, which has also caused social and health problems [26].

In Iran, the breast and CCS is currently done as a health program in the middle-aged population of 30 to 59 years of age in national healthcare centres [27]. Women’s participation in screening programs is voluntary. Of note is that the screening test is free of charge in comprehensive healthcare centres. However, screening services are not free in midwifery centres and gynecology clinics and women need to pay for that.

In health care centers, the health information about households, including the status of women's CCS, is recorded in the “Sib” integrated national health system, also, Pap test sampling is performed by the midwives.

In this study, women in the case group had performed the CCS regularly in the past 3–5 years, while women in the control group had not performed the CCS in the past 3–5 years. To ensure the absence of CCS in the control group, the women were asked about when last they had screening done and their status was checked on “Sib” to verify.

Inclusion criteria 1.

Women aged between 18 and 49 years

2.

Residing in areas covered by health service centers in the suburbs: Payambar Azam, Chahestaniha, Islamabad, Soro, Katbi and Tawheed streets

3.

Conducting CCS regularly within the past 3–5 years in the case group and not performing CCS during the past 3–5 years in the control group

4.

Providing an informed consent to participate in the study

5.

Being able to read and write

Exclusion criteria 1.

Incomplete questionnaires

2.

Having family history of CC (family history of CC increases the knowledge of CC and encourages the performance of Pap test [28])

Sampling and sample size

A multivariable logistic regression analysis was run with 20 variables. As a general rule in logistic regression models, 10 subjects are needed for each independent variable within the model. Therefore, for each research group, the minimum sample size needed was estimated at 200. To increase the statistical efficiency, the sample size for the control group was twofold (i.e., 400).

The sampling method comprised of four steps. Firstly, six comprehensive health service centers in Payambar Azam, Chahestaniha, Islamabad, Soro, Katbi and Tawheed were selected. This health service centers covered the suburban areas. Secondly, the researchers went to all six suburban centers to prepare a list of the women receiving services. Thirdly, based on the information recorded in Sib in each center, the list of women who had been screened for CC within the past three years were extracted and then they were separated from other women who had not performed the screening regularly in the past three years. Lastly, through simple randomization, 33 women were selected as the case group from the list of screened women in each center and 66 women were selected as the control group from the second list. In Tawheed health center, which covered more women, 35 participants were selected as the case and 70 as the control (Fig. 1).

Fig. 1Instrument

The data collection instrument was a tripartite researcher-constructed questionnaire based on a comprehensive review of the related literature [22, 29]. The data were collected as self-reports.

Part 1: Demographic information and reproductive information: Demographic information included age, marital status, educational level, occupation, socio-economic status, and smoking status. Reproductive information included the age of marriage, parity, history of sexually transmitted diseases, use of oral contraceptives and personal sexual health.

Part 2: The knowledge questions were tested with three choices, “yes”, “no”, and “I don't know”. They were scored 1, 0 and 0, respectively. There were 15 questions including those related to CC with 10 questions (#1–10), such as “A family history of CC can increase the probability of this disease”, and questions related to the Pap test, with 5 questions (#11–15), such as “Performing a pap test is necessary only for women who experience abnormal bleeding”. Items 2, 8, 10, 12 and 14 were reversely scored. In these items, a NO answer was considered correct and a YES answer as incorrect. Thus, one who answered NO received the score 1 and he who answered YES or DON’T KNOW received 0. The classification of knowledge questions was done based on the cut-off point by dividing 15 total scores into three categories, as follows: 0–5 low, 6–10 medium, and 11–15 high.

Part 3: Questions related to access to screening services: These include ten questions to be rated on a 5-point Likert scale (very often, often, less frequently, rarely, never) to be scored between 1 and 5. An example item is "To what extent does the distance from the laboratory to deliver the sample (Pap test sample) affect your decision to screen for CC?" The classification of access questions based on the cut-off point based on scores was as follows: 10–23 referred to low, 24–37 referred to medium, and 38–50 referred to high.

Data quality management

The validity of the present researcher-constructed questionnaire was tested using content validity. The first draft was provided to a panel of experts (5 health education and promotion specialists and 5 obstetrician-gynaecologists) to check the readability, simplicity, relevance and significance of the content. Their comments were used to revise the questionnaire.

Before the main data collection phase, the questionnaire was piloted with 30 participants older than 18 years of age, and their feedback was used to improve the readability and organization of the content. These respondents were selected from six different areas. Their comments were used to revise the initial draft and then these respondents were excluded from the main data collection phase. Its reliability was approved, based on the Cronbach’s alpha test result. The 2-week test–retest with 30 participants was used to check the reliability of the instrument.

Next, the scores of the first and second tests were compared. If the correlation coefficient of the first and second tests in each part was higher than 0.7, the construct was reliable. To calculate the agreement between test and retest, the ICC index was estimated. Thus, to estimate the agreement between the mean test scores and the mean retest scores of knowledge, a CC of 0.94 was estimated. For the access questionnaire, Cronbach's alpha was estimated at 0.876 with an ICC of 0.843. Therefore, the reliability was approved. As the researcher was physically present when the data were collected, there were no defective questionnaires.

Data collection

After obtaining ethical approval and the required permissions to conduct the study from hormozgan university of medical sciences, the researchers went to the selected health centers and collected the data. A researcher fluent in the local language together with a health care provider of each area visited the participants (case and control groups) that met the inclusion criteria. Having completed the consent form, the participants were asked to complete the questionnaires. The participants in face-to-face meetings with the researcher completed structured questionnaires which took approximately ten minutes to complete. The women could also complete the questionnaire surveys at home at their convenience and return them later to the researcher.

Ethical considerations

In order to collect the data, the researcher visited the comprehensive health service centers with an official introduction letter from the University describing the study and data collection needs. Moreover, the researcher introduced himself to all participants and explained the purpose of study in a clear and concise manner that was easy to understand. This was followed by obtaining written consent. Participation in this study was completely voluntary and anonymous. The confidentiality of information was preserved. This study was approved by the Ethics Committee of Hormozgan University of Medical Sciences (#IR.HUMS.REC.1401.228).

Data management and analysis

The case and control interval variables included the knowledge of CC, knowledge of pap test and access. Descriptive statistics (mean and standard deviation) were used for non-interval variables (age, marital status, age of marriage, education, occupation, socio-economic status, parity, history of sexually transmitted diseases, use of oral contraceptives, personal sexual hygiene and smoking status). These were described as frequency and percentage. Independent-samples T-test was run to compare the mean scores of the model constructs between the two groups (one with and the other without the screening program). To relate the screening behavior with demographic variables, fertility variables, knowledge and access, univariable logistic regression was first performed, and then the variables with a p-value > 0.025 in the univariable logistic regression analysis were used in multivariable logistic regression analysis. Since age and occupation were not significant in univariate regression, they did not enter the multivariate logistic regression. The data were statistically analyzed in STATA 14.2 and a p-value < 0.05 was considered as statistically significant.