Assessing women’s stated preferences for breast cancer screening: a systematic review and a meta-analysis

Abstract

Background

Breast cancer is the most common diagnosed disease, and the second leading cause of death among women. Thus, due to its importance, the current research is aimed at identifying the preferences of individuals for improving breast cancer screening programs and the related policies.

Method

A systematic search was applied on databases including - PubMed, Scopus, the Web of Science, Embase, Cochrane, SID- up to October 2022. The including articles were original or review papers that assessed individuals’ willingness to pay. Also, articles including the effective variables or attributes for breast cancer screening program were included. Meta-analysis was applied to calculate Willingness to Pay (WTP) as a mean for breast cancer screening followed by vote-counting for identifying the variables and attributes correlated with screening.

Results

A total of 721 articles were identified during the first phase. After the screening process, thirteen papers were chosen, out of which, nine assessed mammography as a breast cancer screening program. The results of random effect meta-analysis on the including studies indicated that the rate of willingness to pay for screening was 0.28% of GDP per capita (95%CI: 0.14–0.43), which was found to be statistically significant. The result of stratified meta- analysis indicated that the rate of willingness to pay for screening was 0.22% of GDP per capita (95%CI: 0.07–0.37), which was found to be statistically significant. Generally, income was the basic factor for receiving screening services, and cost was an effective attribute for participating in screening programs.

Conclusions

To increase women’s participation in breast cancer screening programs; it is essential to provide legitimate information and eliminate the barriers to women’s non-participation. Offering rapid tests at low costs in healthcare centers (both in terms of travel and screening time) delivered by female staff can lead to an increase in women’s willingness to participate in breast cancer screening programs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-024-11847-7.

Introduction

Breast cancer is the most common diagnosed disease, and the second leading cause of death among women. Approximately, 14.9 million new cases of breast cancer were identified in 2012, which was expected to reach 22 million during the next two decades. About 1.67 new cases of breast cancer are identified each year and breast cancer accounts for 25% of all the cancers known among women worldwide [1]. Many oncologists believe that screening programs and early diagnosis are the main steps in reducing the cancer mortality as a result of early interventions [2–4].

Without considering the possibility of false positives (showing a noncancerous dense tissue in breast) and false negatives (wrong diagnosis of noncancerous tumors), screening for diagnosing the cancer in early stages has the potential to treat cancer successfully and increase patient’s life span [5].

Different methods for breast cancer screening has been indicated by the American Cancer Society in the last decades, including breast self-examination (BSE), clinical breast examination(CBE), mammography, ultrasonography, and magnetic resonance imaging [6]. In the same vein, the American Cancer Society suggests monthly and continuous breast self-examination for women over 35 years [7].

Among various methods, mammography, as a non-invasive method for breast cancer diagnosis, assessment, and screening has been widely studied [8]. Screening location and method, as well as women’s preferences for the screening test, are expected to affect women’s participation in the screening programs [9, 10]. In order to have a regular screening program with maximum attendance, it is essential to identify women’s preferences in conducting screening programs efficiently. Generally, individuals’ preferences for each policy or outcome can be estimated from two general approaches: the stated preference and the revealed preference methods [11]. However; between the two methods; the revealed preference method is not able to answer all related research questions raised in health economics.

Over the past decade, the possibility to examine individual’s preferences has been raised via initiative methods, mainly WPT approach [12]. The revealed preference methods are the only methods that can measure the whole economic value of a good or service, and therefore, make the valuing of goods and hypothetical interventions possible [13]. The key to the success of the stated preference survey is asking the questions in a versatile way [12] (Supplement 1).

In line with other studies, the current study is aimed at determining women’s preferences for breast cancer screening using stated preferences and identifying the effective factors in women’s participation in breast cancer screening programs. The results of this research can help policy makers in determining the cost of breast cancer screening services, implementing screening programs, and making informed policies to motivate women to take part in breast cancer screening programs.

Method

A systematic review based on the standard methods was conducted and reports were presented according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [14].

Search strategy and databases

The search for the articles was done in the valid databases including PubMed, Scopus, Embase, Web of Science, Cochrane, and SID. Searching was aimed at recognizing the studies that estimated women’s willingness to pay for the screening, and finding out the effective factors on willingness to pay and finally extracting the effective attributes for improving the programs that can increase more women participation.

The main keywords for this search strategy were “willingness to pay”, “breast cancer”, “conjoint analysis”, and “stated preference”. The initial search was conducted by (S.R) and reviewed by (S.N). All the papers published in both English and Persian (author’s native language) went under a complete review. The search was performed in October 2022 and a complete PubMed search strategy is available as Supporting Information (Supplement 2).

Eligibility criteria (Inclusion/Exclusion)

The original or review papers that assessed individuals’ willingness to pay and other effective variables or other attributes related to breast cancer diagnostic tests were included in the current study. Supplement 3 presents the inclusion and exclusion criteria.

Quality appraisal of studies

The quality of papers was assessed using the STROB tool [15] through evaluations conducted by two independent authors, ( S.R & M.Y), and one independent third reviewer in case of possible disagreements (R.Kh). Supplement − 4 details the number of articles that met each STROBE recommendation. The recommendations that were fully met were those related to the reporting of the reasons and rationale of the investigation (recommendation 2), the study design (recommendation 4) and the provision of summary measures (recommendation 15). On the other hand, the recommendation with the lowest scores were those related to the description of the sensitivity analysis (recommendation 12e; 1/13 8%).

Data extraction

In order to maintain the reliability in test results, the data extraction and examination were done separately by the first and second authors. The extracted items included; author’s name, study year, country, sample size, study design, data collection method, despondence rate, method, screening procedures and results.

Data analysis and synthesis

The analysis of the included studies consisted of two phases: determining the willingness to pay (WTP) and identifying the effective factors and attributes influencing screening preferences. These phases are explained separately below.

Phase 1: determination of willingness to pay

• Estimating Effect Size: In this meta-analysis, the effect size represents the amount women are willing to pay for screening procedures. WTP values were initially reported in US dollars in all included studies. To ensure comparability across countries, these values were converted to a percentage of each country’s GDP per capita. Expressing WTP as a percentage of GDP per capita allows for a more accurate reflection of the economic context of each country.

• Estimating Standard Error of the Effect Size: In each study, the weight effect size was determined by the standard error (SE) associated with the effect size. For studies that reported a standard deviation (SD) or confidence interval (CI), the SE was calculated using the appropriate formula.

• Random Effects Meta-Analysis: A random effects meta-analysis was conducted on all included studies to estimate the mean WTP.

• Stratified Meta-Analysis: To explore potential differences based on economic context, the studies were stratified into subgroups based on the level of development for each country, distinguishing between those conducted in developed and developing countries.

All analyses were conducted using Stata version 17.

Phase 2. Determination of the effective factors and attributes on screening

The effective factors on individual’s preferences were extracted which was found significant only in one study [16–22]. Moreover, effective attributes were extracted on individual’s preferences for rising and improving screening programs.

Results

The data from six valid databases were 721 papers, which were reduced to 338 after removing duplicates. Also, a number of 322 papers were excluded based on their titles and abstracts. Three abstracts did not have a full text, and the full text of three papers did not meet the required criteria due to the limitation of the studies in confining the screening program to women over 70 years old. Also, the articles that calculated willingness to pay for comparing diagnostic test of cancer among the patients were excluded from the study. Finally, a number of 13 papers (11 studies) were selected for further analysis (two papers published through one study) (Fig. 1).

Fig. 1.

PRISMA flow diagram of screenong article for breast cancer

Quality appraisal of studies

Approximately all the included studies contained the existing items in STROB checklist. however, after the quality assessment by a shift in focus on the papers’ methodology, no article was excluded. The results of the quality assessment have been reported in Supplement − 4.

Characteristics of included studies

From 13 existing studies, nine studies were assessing mammography as a breast cancer screening program. Considering the place of the study, three studies were done in the United States [17–19]. Also, countries including Australia [20, 23], Denmark [24, 25], and Japan [16, 26] had two related studies. Finally, Belarus [27] and Malawi [28] with one study and Iran [21, 22] with two studies were included in the list.

The studies were conducted in two different designs:

• I.In six studies, willingness to pay for screening mammography was estimated using Contingent Valuation Method (CVM) and the effective factors were extracted [16–22].
• II.In the studies that were conducted using by Discrete Choice Method (DCM) [23–28], individuals’ preferences for rising and improving screening programs were extracted. The Attributes of these studies are presented in Table 1.

Table 1.

Characteristics of the included articles (studies)

Citation	Country	Data gathering technique	Sampling method	Method of study	Sample size	Number of scenario		Response rate (%)	Screening procedures	result
										WTP as a % of GDP per capita	Attributes (no. of levels)
Yasunaga et al., 2007 [16]	Japan	computer-assisted questionnaire survey	Randomly selected of internet user	CVM * (DBD)	1200	-		33.1%	mammography	0.0475	-
Khaliq et al., 2014 [17]	USA	questionnaire based interview	Randomly	CVM (DBD)	193	-		72%	mammography	0.153	-
Wagner et al.,2000 [18]	USA	interview	Stratified random sample	DCM** (BG. O)	52		90%		mammography	0.32	-
Wagner et al. 2001 [19]	USA	random digit dialing (RDD) technique,	Stratified random sample	CVM (OP, BG)	1465				mammography	0.374	-
						-		91%
Clarke 2000 [20]	Australian	Mail + telephone survey	Randomly	CVM (DBD)	458	-		90.6%	mammography	0.461	-
Karimabadi et al., 2022 [22]	Iran	questionnaire based interview	Randomly	CVM	384			100%	mammography	0.632	-
Ghaderi et al., 2011 [21]	Iran	questionnaire based interview	Stratified random sample	CVM (DBD)	1800	-		81.27%	mammography	0.461	-
Kohler et al., 2015 [28]	Malawi	reviewed Face-to-face	Randomly	DCE***	20	12		100%	clinical Breast Exams	-	Cost (1) Invitation (2) Time (3) Staff or provider (4) Privacy (3) Accuracy (3) Psychosocial (3) Convenience (6) Facility infrastructure (3) Detection strategy (4) Frequency (2) Results notification(2)
Gerard et al., 2003 [23]	Australian	postal questionnaire	Stratified random sample	DCE	87	4		48.33%	mammography	-	Invitation (4) Time (10) Staff or provider (2) Privacy (2) Accuracy(4)
Tsunematsul et al., 2013 [26]	Japan	postal questionnaire	Stratified random sample	DCE	3200	7		40%	Breast cancer screening	-	Cost (2) Time (2) Staff or provider (2) Place (hospitals or clinics or car screening) (2)
Gyrd-Hansen, 2000 [24]	Denmark	questionnaire based interview	Stratified random sample	DCE	207	4		81.17%	mammography	-	Cost (2) Risk (2)
Gyrd-Hansen and Søgaard, 2001 [25]
Mandrik et al., 2019 [27]	Belarus	in-depth interviews	Stratified random sample	DCE	428	3		89%	Breast cancer screening:	-	Cost (3) Invitation (2) Information (1) Time (4) Test sensitivity (4) Screening modality (3) Staff or provider (1)

*CVM Contingent Valuation Method

**DBD Double-Bounded Dichotomous Choice

***Discrete Choice Model

****Discrete Choice Experiment

Determining the amount of willingness to pay and the effective factors

From the seven papers that studied individuals’ willingness to pay in the monetary item, six papers were found appropriate for meta-analysis and went through meta-analysis process. Also, one study [20] was removed from the analysis as it did not provide enough information for calculation of SE.

The results of random effect meta-analysis on whole studies indicate that the rate of willingness to pay for screening is 0.28% of GDP per capita (95%CI: 0.14–0.43), which is found to be statistically significant (Table 2; Fig. 2).

Table 2.

The results of meta-analysis of the included studies (WTP^* as a percentage of GDP^**)

	method	Pooled estimate	95% confidence interval	P-value	N of studies [ref]
WTP for mammography (as a %GDP per capita)	Random model	0.28	0.14–0.43	0.000	6 [16–19, 21, 22]

*Willingness to Pay

**Gross Domestic Product

Test for heterogeneity: Q = 2576.20 on 5 degrees of freedom (p-value < 0.001)

Fig. 2.

The reported willingness to pay for mammography in different studies as a percentage of gross domestic product (GDP) per capita

he studies were categorized into subgroups based on the level of development of the countries, distinguishing between those conducted in developed and developing nations. Since both studies from developing countries focused on a single country (Iran), a stratified meta-analysis was not conducted. The result of stratified meta- analysis indicated that the rate of willingness to pay for screening was 0.22% of GDP per capita (95%CI: 0.07–0.37), which was found to be statistically significant (Table 3; Fig. 3).

Table 3.

The results of random effect meta-analysis of the included studies (WTP* as a percentage of GDP)

	method	Pooled estimate	95% confidence interval	P-value	N of papers [ref]
WTP for mammography (as a %GDP** per capita)	Random model	0.22	0.07–0.37	0.000	4 [16–19]

*Willingness to Pay

**Gross Domestic Product

Test for heterogeneity: Q = 2576.20 on 5 degrees of freedom (p-value < 0.001)

Fig. 3.

The reported willingness to pay for mammography in different studies as a % of Gross Domestic Product (GDP) per capita

Determination of the effective factors

The effective factors extracted from the studies [16–22] were categorized into four groups including: demographic information (age, education, race, marital status), socioeconomic information (income, employment status), health history or status (family history of cancer, mammography required after a clinical breast examination, three or more comorbidities, health status, family health costs, history of receiving mammography, perceived risk of breast cancer), and knowledge about mammography and breast cancer (level of information, health concerns) as explained in Table 4.

Table 4.

The effective factors on WTP for mammography

Citation	Demographic				Socio- economic			health history or status							Understanding and knowledge about mammography and breast cancer
					Income	Employment status
	Age	Education	race	married	Income	Retired or homemaker	Unemployed	Family history of cancer	Need mammography after clinical breast	3 or more comorbidities	Health status	Family health costs	History of receiving mammography	Perceived Risk of breast cancer	level of information	Health concerns
Yasunaga et al., 2007 [16]	↑				↑			↑↑					↑↑		↑↑	↑↑
Khaliq et al., 2014 [17]	↑	↓	↓		↑			↑		↑↑
Wagner et al.,2000 [18]	↑	↑	↑↑		↑↑				↑↑				↓↓	↑↑
Wagner et al. 2001 [19]	↓↓	↑			↑↑	↓	↓↓								↑↑
Clarke 2000 [20]	↓	↑		↑↑
Ghaderi et al., 2011 [21]		↑↑			↑↑	↑↑					↑↑	↑↑	↑↑
Karimabadi et al., 2022 [22]	↓↓	↑↑			↑↑	↑		↑↑							↑↑

↑↑ The effect of variable is positive and significant

↓↓ The effect of variable is negative and significant

↓ The effect of variable is negative and non-significant

↑ The effect of variable is positive and non-significant

Demographic variables including age, education, race, and marital status were found significant in one study suggesting that married people with higher education and lower age had more willingness to pay. Moreover, race as a cultural feature was effective factor on individuals’ willingness to pay.

Among the analyzed socioeconomic variables, income and employment were found to have a positive effect on individuals’ willingness to pay, with income exerting the most significant influence reflected in multiple studies [18–20, 22]. Furthermore, all variables related to health history and status exhibited a positive and statistically significant effect on willingness to pay. However, an exception was observed with the variable concerning the history of receiving mammography. Contrary to the other health-related variables, a history of receiving mammography was associated with a negative and statistically significant effect on willingness to pay [19].

Generally, understanding and knowledge about mammography and breast cancer had a positive and significant effect on willingness to pay [16, 19, 22].

Determination of effective attributes

DCM was used in 6 studies [23–28], in which the detailed information about policies of breast cancer screening have been presented. In order to select attributes, a review of literature and qualitative interviews were applied. Effective attributes on screening behavior by women included time, cost, service provider, accuracy, and risk which are illustrated in Table 5. Based on the findings, reducing the time and cost of traveling to the screening location (between 20 and 50 min) could lead to an increase in individuals’ willingness to participate in screening programs [19, 23, 26–28]. Furthermore, cost and out of pocket payment had a negative effect on individuals’ behaviors on screening [23–25, 27, 28]. Also, increasing the accuracy of screening tests and reducing false positive risk had a positive effect on screening behavior of individuals. Thus, more accurate screening tests could lead to earlier diagnosis of disease and a reduction in the risk of false positive diagnosis [23–26].

Table 5.

Effective attributes on screening behaviors of women

Citation	Time		Cost	Provider		Accuracy	Risk
	Travel time	Screen time	Out of pocket expense	Male	Female	100%	Risk reduction over life time	Risk of false positive
Kohler et al., 2015 [28]	✓		✓		✓
Gerard et al., 2003 [23]	✓	✓				✓
Tsunematsul et al., 2013 [26]		✓	✓		✓
Gyrd-Hansen, 2000 [24]			✓				✓	✓
Gyrd-Hansen and Søgaard, 2001 [25]
Mandrik et al., 2019 [27]	✓	✓	✓			✓		✓

Discussion

The current study represents the first comprehensive examination of stated preferences in breast cancer screening. Among the 13 articles reviewed, seven focused on assessing willingness to pay for mammography, while six utilized conjoint analyses to determine preferences for improving screening programs. The study had two primary objectives: To calculate willingness to pay for screening mammography as a percentage of GDP, and to identify the factors and attributes influencing participation in screening programs.

The findings from the analysis of willingness to pay, conducted through a random-effects meta-analysis, revealed that women’s willingness to pay for mammography screening is equivalent to 0.28% of GDP per capita. Stratified by the developmental level of the countries, the number falls to 0.22% of GDP per capita.

In examining the factors influencing willingness to pay, two categories namely, demographic and socioeconomic information were found to differ among the included studies as the result of variations in differences in the norms and conditions of nations. Among the socioeconomic factors, income was the only variable that had a positive effect on willingness to pay in all the included studies. Considering the variables of “health history or status " all had a positive effect on willingness to pay, except for history of receiving mammography in one study [18], which revealed a negative, significant effect. Unlike the expectations, peoples who had more mammography were not showing willingness to pay due to their knowledge and awareness of the issue. Variables related to the group encompassing understanding and knowledge about mammography and cancer were found to have a positive effect on willingness to pay. The analysis indicated that as awareness and information increase, the individual’s willingness to pay also rises [19, 22] among the variables, the stage of change (stage of information) is a trans-theoretical model which is highly sensitive to minor changes in behavior. This structure expresses that such a behavior occurs and brews in certain stages, and women who go through the stages of change are ready to participate in different screening programs, though, with different levels of readiness based on different races [19]. Hence, there are significant preference dissimilarity in women’s preferences for a breast screening programs. For example, race and social norms, in addition to awareness and information, affect individuals’ willingness to pay. Moreover, providing complete information and increasing the accuracy of screening tests have a positive effect on willingness to participate in the screening programs.

Nevertheless, accuracy, time, cost, and staff or service providers were also found effective and common attributes in all DCM studies. According to the findings, as the resukt of certain social and cultural norms in some nations, the presence of male staff was found as a barrier to women’s participation in screening program [23, 26–28]. The cost was also an important attribute, as the high cost of transportation to reach a public hospital, as well as the high cost of screening tests prevented people from participating in screening programs [24–28]. Time to reach screening locations and screening time was another important attribute to reduce participation in programs. Consequently, shorter time resulted in more participation [23, 26–28]. Test accuracy was found to be the most important attribute in women interest to take part in screening programs. Furthermore, test accuracy resulted in more willingness among individuals to participate in screening programs [23–25, 27].

Overall, women’s health seeking behaviors and challenges within health infrastructure highlights the potential social norms and structural factors that may affect the demand for diagnostic cancer services [28]. Instances of these attitudes include statements such as 1. “I am healthy and do not see any symptom of illness in myself, so I do not see any reason to participate in the screening program” [28]. 2. Many at risk women (over 40 years old) expressed dissatisfaction with the overcrowding of screening programs as they were mostly busy with employment commitments [26]. 3. Providing screening services and breast examination by male staff was identified another barrier to women’s participation in these programs [26, 28].

In order to increase participation in breast cancer screening programs, it is essential to provide accurate information on breast cancer screening and try to reduce participation barriers in cancer screening.

Providing accurate information about the purpose and meaning of breast cancer screening is important, as it provides the opportunity to participate in cancer screening program before feeling any symptoms [23]. Moreover, identifying women’s preferences regarding (in order that) the delivery of diagnostic services will help increase interest and subsequent participation in screening programs ultimately helping the early cancer diagnosis.

Women’s preferences for developing and increasing the attraction in screening programs varied in different countries according to norms and socioeconomic status.

In a country such as Malawi, which is a less developed country with a predominantly rural population, clinics and hospitals are typically constructed near to the place of residence and offer free services. Moreover, the presence of female doctors, along with other factors, can resolve the participation obstacles in screening programs among Malawi women [28]. In developed countries like Denmark and Australia, these infrastructural problems have been solved. Also, through increasing test accuracy and reducing the rate of false positive diagnostics, more women have been attracted to these programs [23, 25].

In Japan, as an instance of a developed country, due to the Japanese societal norms, emanation of breasts by male doctors has prevented women from participating in screening programs [26].

It should be noted that while some attributes may be important individually, their priority may shift when compared to other factors. Therefore, this study was limited in a sense that it could not estimate the value and importance of extracted attributes. In fact, the importance or lack thereof for any attribute should be assessed within the context of the entire set of included attributes and their range of levels. This means that they should not be considered in isolation but rather in comparison with one another.

While our study primarily reflects preferences and WTP for current breast cancer screening methods, it is important to consider the roles of emerging screening technologies, such as personalized risk-based screening and advanced imaging techniques, in influence these preferences. While personalization in screening programs could potentially increase WTP due to a higher perceived benefit and accuracy, the more recent and potentially costlier tests might have the opposite effect.

Limitations and future research perspectives

One of the limitations of our study was the restricted number of databases (six) applied for the current study, as a result of which, some evidence might have been excluded. The second limitation was the restriction of articles to those written in English and Persian. There is no doubt that there exists a rich literature on breast cancer screening in other languages.

Future research should investigate how developments in screening tests impact women’s stated preferences and WTP to ensure that screening programs remain both effective and accessible.

Conclusion

Generally, among the investigated variables in the included studies, income was the only factor that positively affected willingness to pay. Also, test and transportation cost were identified as effective attributes in DCE studies. Therefore, the results indicated that cost is an effective variable on the amount of individuals’ willingness to pay for mammography and participation in screening programs. To increase participation in breast cancer screening programs, it is essential to provide accurate information about breast cancer and reduce the obstacles according to the structure, culture and societal norms of each country and its women’s preferences. Factors such as cost, gender, service provider, travel time to a public hospital, short distance, receiving service in short time, test accuracy, and improving screening environment in terms of overcrowding, are among factors that are considered effective in increasing interest in screening programs. Ultimately, it would be easy to participate in a screening program if a comprehensive information be provided to people. The administration of the tests at a low cost in healthcare centers and also in short times (both in terms of travel time and screening) by female staff can increase the rate of participation in breast cancer screening programs.

Authors’ contributions

SR conceptualized and extracted the data, and wrote the manuscript. SN contributed to the design, interpretation, revising and preparing final version of the manuscript and provided critical feedbacks on the manuscript. MY contributed to the conception and interpretation of the study. RK contributed to the extraction of the data, interpretation, revising and preparing final version of the manuscript. All authors have read and approved the submitted and revised final version of the manuscript.

Funding

Tabriz University of Medical Sciences and Shams cancer charity, Grant No: 37305-202-01 97), provided financial support to this study.

Data availability

All data generated or analyzed during this study are included in this article and its supplementary information files are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study is part of a larger study conducted for a MSc thesis that was approved by Medical Ethics Committee at Tabriz University of Medical Sciences (Ethics approval No: IR.TBZMED.REC.1398.898).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.