Factors Associated With the People Preferences for the Breast Cancer Genetic Tests: a Systematic Review

Background: Despite the genetic test’s benets in early detection of disease, many people do not have the test performed. This study aimed to systematically review the associated factors with a preference for the gene tests regarding breast cancer. Methods: It was a systematic review study. Data were collected from Scopus, PubMed, Web of Science (WOS), ProQuest, and Embase databases. All full-text English language studies up to 2020 that addressed the preference for the gene screening tests for early detection of breast cancer were included. Data extracted and assessed for quality by two independent reviewers. Disagreements were resolved by a consensus meeting with a third reviewer. The vote counting was determined for identifying the associated factors with the preference. Results: From 1269 initially retrieved citations, 25 studies were included. According to vote counting analysis, age was identied as strongly negative factor while being married, income, family history of breast cancer, personal history or presence of breast cancer, BRCA-related knowledge were identied as strongly positive factors with preference of genetic tests for breast cancer. About 96% of articles were in high quality. Conclusion: There was low evidence in associated factors with a preference on gene tests, especially for psychological factors. So, further research is needed to help policymakers to develop early detection strategies and increase people's participation that leads to the success of the strategies and avoiding the high cost of treatment as well.

Although the search question was in case of associated factor, the words such as 'factor*', 'predict*', 'associat*', 'relat*' and 'determin* were excluded from the search strategy because of more outputs in the absence of them as well as available to qualitative studies.
The strategy terms involved some keywords or medical subject headings (MeSHs) that identi ed by a research librarian. Speci c search strategies are presented in Appendix 1.
The protocol for this review was registered on PROSPERO in July 2020 (ID: CRD42020190811).

Inclusion and exclusion criteria
Studies were selected for inclusion if the full-text was written in the English language up to 2020 and addressed the preference for the gene screening tests for early detection of breast cancer.
The exclusion criteria were the studies (1) in the form of abstracts and conference papers, (2) addressed the breast cancer screening methods other than genetic testing such as mammography, breast self-assessment, (3) focused on the preference of people for treatment intervention for breast cancer, (4) focused on genome sequencing or Gene-Panel, (5) focused on the preference of genetic tests for other types of cancers, and (6) longitude studies that addressed the effect of an intervention on the preference of genetic tests.
We assumed that all researchers' studies were published as articles, so the gray literature was excluded from the present study.

Study selectionandData extraction
All citations were imported into an electronic database, Endnote version 8, with duplicates subsequently removed. The retrieved articles after deleting duplicated were collected into a separate Excel sheet and disagreements between researchers were examined by Cohen's kappa coe cient. A 0.6 to 0.8 for the kappa coe cient is substantial for the studies, although more than 0.8 for the coe cient is almost perfect based on accepted standards (14).
The selection of studies was done in two stages. First of all, titles and abstracts of retrieved articles were screened against the inclusion and exclusion criteria.
After that, the full text of eligible studies was examined.
Two reviewers (Z.M., A.A.) independently applied inclusion and exclusion criteria to screen titles and abstracts of the remaining articles.
The full text of eligible articles was screened thoroughly by two independent reviewers (Z.M. and N.M. or M.L.). Decisions regarding the inclusion of studies, where it was not clear as to whether or not the material was inclusion criteria, were resolved by a consensus meeting with a third reviewer (A.GH.). Finally, reference lists and citations of eligible articles were checked manually for additional relevant studies.
One author (Z.M.) performed data extraction independently and, then, these were checked by a second author (N.M). A self-assessment form was designed for extracting data. Information retrieved from the researches included the rst author, year, title, study design, sample size, methodology, the location, and associated factors with the preference for genetic testing.

Quality assessment
Two reviewers (Z.M. and N.M.) assessed the quality of the studies independently, and the nal quality assessment was based on consensus. For quality assessment of the selected studies using the criteria developed by Johannesen and LoGiudice. It contains 8 items that cover inclusion and exclusion criteria, a good response rate (i.e., ≥80%), minimized the selection bias, well-de ned outcome, valid and reliable instrument for measuring the outcomes, valid and reliable instrument for measuring risk factors, and adjusting the ndings for confounding risk factors. It has been used in previous study (15)(16)(17).
All criteria were scored. Items that covered the criteria were given a perfect score (=1), and those that did not cover the criteria, a score of zero was assigned. In case the necessary information was unclear, a negative score was provided, too. Therefore, the sum and percentage of quality assessment of articles was calculated from 8 points. The studies that awarded 50% or more than 50% of points, were of high quality (15,16).

Data analysis
Meta-analysis were not performed because of a number of reasons: 1) inclusion and exclusion criteria for sample size were not the same (i.e. some studies considered patients and some healthy as well as rst degree relatives of breast cancer patients), 2) a number of studies were performed in a quantitative and some qualitative, and 3) the method of the selected studies in terms of measuring related factors were different. So, the results synthesized by the vote counting as used in previous studies. The factors were categorized by their signs as well as signi cance. Factors categorized to positive, negative, or nonsigni cant effects on preference of genetic tests. Finally, gave a vote to each of factors and the votes were counted. Factors with the same sign as well as signi cance and factors that obtained at least three votes was considered as an affecting factor on preference of genetic tests for breast cancer (18).

Results
The literature search identi ed 1269 articles; 193 from PubMed, 428 from Embase, 354 from Scopus, 68 from ProQuest, 214 from WOS, and 12 from other sources. After removing 547 duplicates, 722 titles and abstracts were screened for eligibility, of which, 667 articles were excluded because of irrelevant title and abstracts, and 30 articles because of focusing on attitude of genetic tests, Gene-Panel, and measuring the effect of intervention on preference were excluded. Finally, 25 articles were included.
Inclusion and exclusion criteria to screen titles and abstracts of the identi ed articles were independently applied by two reviewers (Z.M., A.A.). There was disagreement in the rst steps of the study selection but the remaining stages gained authors agreement to the terms of including the articles. The Cohen's kappa coe cient between researchers authors (Z.M. and N.M.), was 0.72 that is substantial based on accepted standards. The study selection is shown in Fig. 1 based on the PRISMA diagram.

Study Characteristics
The characteristic of the selected studies was summarized in Table 1. Publication year of the selected articles ranged from 1999 (27) to 2019 (19,20) and the studies at 2002 were frequent (10,24,37,38).
A number of ten studies were performed in hospital or clinic or medical university setting (10,19,20,24,26,28,29,34,37,42), a study were considered medical university and community as setting (41), and remind were performed in the community.
Three studies were used interview (25,27,31). Although all of the articles were designed a self-administered questionnaire, for some dimensions of the questionnaire used identi ed scales, especially for psychological factors. Anxiety/depression was assessed by the Hospital Anxiety and Depression scale (HADS) (20), Health/Illness Orientation Scale (HIOS) (32), and the Health Belief Model (HBM) were measured for measuring a number of physiological factors (26).
Family functioning was measured by the Cohesion subscale of the FACES, the Family Crisis Oriented Personal Evaluation Scale (F-COPES). The stressful encounters of everyday living were measured by The Ways of Coping Questionnaire. Social support was measured by the Medical Outcomes Study (MOS), and mental health was measured by the SF-36 questioner (35). Cultural beliefs and values were assessed in terms of communalism (was measured by Communalism Scale), temporal orientation (was measured by the Temporal Orientation Scale (TOS)), and religious coping style (was measured by Religious Coping Style Scale (RCSS)). Cancer fatalism was measured by the Cancer Fatalism Inventory (CFI) (36). The intrusive thoughts subscale of the IES was used to assess breast cancer-speci c distress (10). The Family Environment Scale (FES), the Center for Epidemiologic Studies Depression (CES-D) Scale, the Spiritual Well-Being Scale (SWBS), the Rosenberg Self-Esteem Scale, and the Life Orientation Test (LOT) were measured for dimensions of family relationships, depressive symptomatology, personal spiritual meaning and satisfaction, global self-esteem, and dispositional optimism, respectively (39). Health locus of control (HLOC) was measured by internal, powerful others, and chance (37).
Although a study was used structural equation modeling (33) and a study was used content analysis (31), the remains were used regression as well as statistical analysis to identify the relationship between variables of interest.
Factors associated with the interest in BRCA gene screening test Presented in Table 2 are the selected factors of the studies organized into ve categories; namely, sociodemographic, health related factors, knowledge as well as awareness, psychological variables, and cues to action.

Sociodemographic factors
A number of eleven factors were identi ed in the selected articles that their association between preferences of genetic tests were addressed. Age were assessed in 19 articles that one (39) study declared a positive relationship and ve (21,24,28,30,31) of the selected articles were displayed a negative relation between age and preferences for genetic tests. Based on vote counting there is strong negative relation between age and preferences for genetic tests.
Marital status (20, 21, 27, 31, 34, 39) and income (19,21,29,30) were the factors that displayed a clear relationship with preferences for genetic tests. Based on vote counting there is strong positive relation between marital status as well as income and preferences for genetic tests.
Between gender as well as religion and preferences did not nd any positive and negative relation while a number of ve (20,28,31,39) and two (26,41) articles displayed any relation between them, respectively.

Health related factors
There were seven factors regarding health related factors identi ed in the included studies. Family history of breast cancer (21,27,28,30,38) and personal history of breast disease or presence of BC (30,31,41,42) were identi ed as positively strong association with preference on genetic tests. Based on vote counting there is inconclusive result for the relation of family history of other cancer (24,28), the number of previous breast biopsies (21), presence of rstdegree relatives affected with cancer (39), presence of family members testing positive (22,23), and prior hormonal treatment member(s) testing positive (29) with the preference because of any evidence of positive or negative effects of the variables.

Knowledge/Awareness
Knowledge (10,20,37) and awareness (29,30) about BRCA gene tests as well as breast cancer and its risk factors (27,37) were assessed by a number of selected studies. Although there were not enough evidence for association between knowledge as well as awareness about breast cancer and its risk factors with preference of genetic tests, the evidence were con rmed the strongly positive relation between BRCA-related knowledge and preference for genetic tests for breast cancer.
Women's positive attitude regarding the genetic tests stated as predictor of genetic tests although the signi cance did not assesse (33).

Reasons for refuse the genetic tests
Financial hardship (28) as well as the high cost of genetic tests (19,28,29), implications of a positive BRCA tests result on insurance (29), lack of expert consensus in recommending genetic testing (29), not satisfying with the treatment before the symptoms appear (19), lack of knowledge about the heritability of cancer (19), unavailability of decision makers (19), fear of the result as well as nding tumors (19), disrupting the life's projects, concern about the effect on your family, feel labeled or singled out, not remaining con dential of the results, do not trust modern medicine (34) were stated as the reasons for refusing for opting genetic tests.

Discussion
Genetic susceptibility testing is the newest in the medical science that has the potential to transform the practice of preventive care. Although its bene ts, many people do not have the test performed. The present systematic review study was assessed to identify the associated factors to preference of peoples for genetic tests.
The selected studies vary in terms of sample sizes, the associated factors with, and tools to measure the factors for preference of genetic tests. In sum, 80%, 40%, and 56% of selected articles were assessed socio demographic, health related factors, and physiological variables, while 12% and 8% of studies addressed knowledge as well as awareness and cues to action variables for preference of genetic tests, respectively.
Although the vote counting was used, the ndings about the associated factors with preference especially for knowledge and physiological variables should be interpreted with caution because of low evidence as well as different tools for measuring the factors.
Overall, based on the results age was identi ed as strongly negative factor while being married, income, family history of breast cancer, personal history or presence of breast cancer, BRCA-related knowledge were identi ed as strongly positive factors with preference of genetic tests for breast cancer.
The associations between sociodemographic characteristics and preference for genetic tests were complex, although base on vote counting age, marital status, and income were the important associated factors. Based on vote counting there was strong negative relation between age and preferences for genetic tests (21,24,28,30,31), although a study (39) declared a positive relationship. Based on the results there was strong positive relation between marital status (20,21,27,31,34,39) as well as income (19,21,29,30) and preferences for genetic tests. Insurance status is a variable that even low income persons have an opportunity for getting the medical services with lower price that it was con rmed in two studies (20) (42). The reaction of peoples for price changing, leads us to the conclusion that predictive tests such as genetic tests are elastic and by lower price the demand increase. It is con rmed in the Sun et al study, too.
They were assessed the effect of costs for high risk Korean females on preference the tests after the implementation of national health insurance coverage for the tests and based on the results refusal rate were decreased (28).
Gender as well as religion and preferences did not have any positive and negative relation while a number of ve (20,28,31,39) and two (41) (26) articles displayed any relation between them, respectively.
Although the number of the selected articles con rmed the positive relation between preference for genetic tests and ethnicity, especially Ashkenazi Jewish (42) (25), education level (24,29,30,37), employment status (24), live in urban areas (19) and, having children (31), there were evidence in contrast. So, the result for the stated variables was more unclear and they are inconclusive based on vote counting.
It was expected that risk factors for getting breast cancer such as family history (people with a greater probability of being a carrier of the BRCA1/2 gene) had got a positive association with preference for genetic tests that studies con rmed it for family history (21,27,28,30,38), although a study was in contrast (24). Gene mutation is frequent in some ethnicity such as Ashkenazi descent and it was expected that they have more interest for genetic tests but the studies did not nd the same results. Women with a family history of breast cancer may have learned more about gene tests and be more anxious about undergoing testing (24).
Personal history of breast disease or presence of BC (30,31,41,42) were identi ed as positively strong association with preference on genetic tests in the number of studies. Based on vote counting there was inconclusive result for the relation of family history of other cancer (24) (28), the number of previous breast biopsies (21), presence of rst-degree relatives affected with cancer (39), presence of family members testing positive (22) (23), and prior hormonal treatment member(s) testing positive (29) with the preference because of any evidence of positive or negative effects of the variables.
Although lack of evidence, there were the strongly positive relation between BRCA-related knowledge (10,20,37) and preference for genetic tests for breast cancer. Aging had negative relation with preference among knowledgeable women about genetic testing. Knowledgeable 59 and 60 years old women expressing the greatest and lowest interest in genetic testing, respectively (38). Awareness (29) (30) about BRCA gene tests as well as breast cancer and its risk factors (37) (27) did not associated with preference for genetic tests. Women's positive attitude regarding the genetic tests as well as its results stated as a predictor of genetic tests although the signi cance did not assesse (33).
BRCA testing is a sensitive issue and can lead to a range of opinions and emotions. So, Psychological variables have an important role for preference for the genetic tests and more than half of selected articles were addressed them. Breast cancer risk perception (20-22, 27, 30, 40, 42) and perception of the bene ts of BRCA testing (20,41,42) were identi ed as positively strong association with preference on genetic tests based on the results.
There were negative relation between having a locus of control of others (21) for genetic tests while it was in contrast for having a locus of health providers control for genetic tests (37). being a high monitor (21), greater family cohesion (39), and access to testing (40) were the variables that lead to preference for genetic tests while ways of coping (seeking social support) (35)  It is interesting that although women displayed the high rate of interest for genetic tests to identify children's risk, to plan for cancer screening tests or preventive interventions, to help for advance research, to be reassured (34) (38), to make decisions about having (more) children (34), curiosity, to warn the family, to achieve peace of mind, reduced worry and to warn the family (38), it did not lead to opt the gene tests (29).

Limitations And Strengths
Although in our knowledge the present study was the rst one that assessed the preference for a new medical technology, genetic tests, for early detection of disease and did not have time-limited and included the qualitative and quantitative studies as well, considering English language studies was our limitation.
Low evidence to decision making about cues to action factors that associated with the preference, using different de nition for measuring the variables were the other limitation of the present study that need to further studies.

Conclusion
The burden of breast cancer can reduce by early detection interventions such as genetic tests as a new technology in medical science. But before the implication of early detection strategies, the participant's rate of people should be estimated. Identifying the associated factors with a preference for gene tests can help policymakers to remove barriers and had high coverage of people in such plans. Based on the result of the present study a number of factors were associated with preference but there was low evidence, especially for psychological factors. So, further research is needed to help policymakers to develop early detection strategies and increase people's participation that leads to the success of the strategies and avoiding the high cost of treatment as well. Availability of data and materials All data generated or analyzed during this study are included in this published article.

Competing interests
The authors declare that there is no con ict of interest regarding the publication of this article Funding This study was a part of a Ph.D. thesis that is supported by Iran University of Medical Sciences (grant No: IUMS/SHMIS-98-4-37-16709).
The funder has only sponsored. The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.