Design and Psychometric Evaluation of Mammography Behavior Predicting Scale

Breast cancer disease is an important public health trouble in the world. The research shows there are 2.3 million women with breast cancer and 685 000 deaths worldwide in 2020. The results indicated that 7.8 million women were with diagnosed breast cancer in the past 5 years and, as a result, breast cancer is the world’s most prevalent cancer. Research shows disability-adjusted life years (DALYs) due to breast cancer in women is more than in other cancer. Breast cancer mortality did not change much from the 1930s to the 1970s (1). The previous studies showed that the incidence of breast cancer has increased alarmingly among Iranian women (2, 3). They also encounter fatality because of the advanced stage of cancer (4).

Mammography screening can lead to early detection and management of breast cancer (5) and it also plays an important preventive role in decreasing breast cancer, especially among women over 40 years old (6). Apart from the quick rate of increase, it has also been seen to start at an earlier age as per statistics, which suggest that the mean age of Iranian women with breast cancer is 49.6 years (95%CI 49.5 - 49.6) (2) compared to 55 to 60 years among women in the United States (7) and 65 to 69 years in Australian women (8). Iranian women have been consistently reported as having low partnership in mammography practice and, unfortunately, their disease has spread due to delayed referrals, thus putting them at risk of dying from advanced cancer diagnosed (4). Because of this, breast cancer is considered a health priority in Iran (8). The health action process approach (HAPA) is a universal and psychological theory of health behavior change, developed by Ralf Schwarzer for assessing health promotion behaviors such as breast cancer (9). There is no valid Persian instrument for evaluating influential factors.

This paper aimed at reporting the psychometric properties of a questionnaire on mammography behavior based on HAPA.

In this cross-sectional study in the first phase, a questionnaire including 41 items was developed based on literature about the mentioned constructs of HAPA and preventing breast cancer behavior. In the second phase, healthcare professionals discussed the initial instrument. This initial item was piloted and, then, reduced via item analysis and scaling methods (10-14).

The face validity was determined by 70 women, who had similar characteristics to the study sample and qualified for participation based on the inclusion/exclusion criteria. The inclusion criteria were age 40 years and older, being literate, and having a mental and disabling disorder. The age of participants should be older than 40 years because basic mammography is recommended at the age of 40 years. The exclusion criteria were women with a positive history of breast cancer in their family or friends, as well as if someone suffered from any defect or disease that interfered with mammography. In the qualitative approach, the ‘ambiguity’, ‘relevancy’, and ‘difficulty’ of the tool were assessed by 70 women older than 40 years. At this stage, 4 items improved. In the qualitative method, the content validity was assessed by an expert panel and 22 members, including 8 health educators, 7 specialize in reproductive health and gynecology, 5 psychologists, and 2 oncologists and content validity ratio (CVR) and the content validity index (CVI) were evaluated. According to Lawsh as cited in Wilson et al. (15), CVR with a score < 0.418 was deleted and a CVI value of 0.78 or above was considered satisfactory for each statement (16).

The subjects were selected by multi-stage cluster sampling. Firstly, among the 10 health networks under the auspices of the Shahid Beheshti University of Medical Sciences, 3 health networks were selected. In the next phase, 5 health centers were randomly selected from each network. Then, 80 people were selected from each center in equal proportions.

The ideal sample size based on exploratory factor analysis (EFA) was estimated at 4 to 10 participants for each item (45 × 9 ~ 400) (17). Figure 1 presents the procedure of sampling. The primary questionnaire included 20 demographic variables and the questionnaire had 45 items relevant to 8 HAPA constructs. The construct validity of the questionnaire was examined by EFA.

Figure 1.

Sampling flow chart of the study population

Principal component analysis was performed by varimax rotation to extract the essential factors and factor loadings 0.5 ≥ were properly considered. For determining the range of statements eigenvalues above one and scree plots were applied. For proper sample size, the Kaiser-Meyer-Olkin measure (KMO = 0.73) and Bartlett’s test of sphericity (P < 0.001) were used. The internal consistency was calculated by infraclass correlation coefficient (ICC) and Cronbach’s alpha coefficient values. The appropriate amount for ICC was considered (ICC ≥ 0. 4). Data analyses was done, using the Statistical Package for Social Science (SPSS).

In this study, all the women (100%) were asked to fill in and return the HAPA questionnaire on mammography behavior. The mean age of the women was 45.6 ± 5.45 years and the level of education, 37.2% of women were less than high school (Table 1).

This study was conducted to design the Mammography Behavior Predicting Scale (MBPS) and evaluate its psychometric property questionnaire on mammography behavior based on HAPA.

Our findings underlined the favorable psychometric properties of this 41-item questionnaire and showed its enforceability as a tool to assess elements affecting mammography behavior. Mammography Behavior Predicting Scale was a valid and reliable measure.

The results indicated a significant correlation between the motivational stage constructs (action self-efficacy and coping planning) and mammography intention.

These findings were consistent with the concepts of HAPA (9), which show high correlations between behavioral intentions and self-efficacy. To better understand these findings, research is needed on a larger sample of women. In the study by Schwarzer et al. (10), the best direct predictor of mammography intention and AP was ASE. In the current study, planning was a key factor and the best direct predictor of mammography behavior. Consistent with our research, Schwarzer and Renner (11) showed there were correlations between RP and behavioral intention. Interestingly, there was a correlation between ASE and OE with mammography intentions. Regarding the relationship between the power of ASE and OE with behavioral intention, the power of the relationship between ASE with behavioral intention compared to OE was strongly related to the situation in other studies (22). According to the HAPA scale, OE and ASE have a high effect on the prediction of behavioral intention, while other studies have reported risk perception is more of a “distant introduction” that informs about the intentions (9). In the social cognitive theory (SCT) of Bandura, SE has a more powerful effect than OE on behavioral intentions (23). Self-efficacy has a greater effect on behavioral intention than outcome expectancies. This study had several limitations. First, generalization is limited by the fact that this sample was taken from Iranian women. Moreover, using self-report instrument mammography behavior could have been over-reported or under-reported. Further studies with adequate verification of the information reported in their design are recommended.