THE IMPACT OF PERSONALISED CANCER RISK DISCLOSURE ON URBAN WOMEN'S PREFERENCES FOR BREAST CANCER SCREENING: A RANDOMISED TRIAL USING BEST-WORST DISCRETE CHOICE EXPERIMENTS
Author(s)
ZHENGNAN MENG, B.S.1, Ping Zhou, PhD1, Junwei Wang, MPH2, Hengzhe Zhao, PhD3, Dongni Hong, MSM4, Li Du, MSc5, Yuanyuan Gu, PhD6.
1Fudan University, Shanghai, China, 2Guangdong Provincial Hospital of Chinese Medicine, Guangzhou, China, 3Macquarie University Centre for Health Economy (MUCHE), Sydney, Australia, 4Eye and ENT Hospital of Fudan University, Shanghai, China, 5Shanghai Center for Women and Children's Health, Shanghai, China, 6Macquarie University Centre for the Health Economy, Sydney, Australia.
1Fudan University, Shanghai, China, 2Guangdong Provincial Hospital of Chinese Medicine, Guangzhou, China, 3Macquarie University Centre for Health Economy (MUCHE), Sydney, Australia, 4Eye and ENT Hospital of Fudan University, Shanghai, China, 5Shanghai Center for Women and Children's Health, Shanghai, China, 6Macquarie University Centre for the Health Economy, Sydney, Australia.
OBJECTIVES: This study aimed to examine how personalised breast cancer risk disclosure influences urban women’s screening participation and attribute preferences, and to explore whether screening knowledge moderates these effects.
METHODS: A total of 923 urban women were recruited using multi-stage stratified sampling. Participants were randomly assigned to four groups defined by objective risk status and disclosure (disclosed high-risk group, undisclosed high-risk group, disclosed non-high-risk group and undisclosed non-high-risk group). Screening preferences were elicited using a best-worst discrete choice experiment (BW-DCE) with seven attributes, including out-of-pocket cost, screening setting, pain level, physician gender, screening interval, false-negative rate, false-positive rate and an opt-out option to capture screening particpation decisions. A heteroscedastic conditional logit model was employed to estimate preferences and account for best-worst scale heterogeneity; interaction terms were used to examine the moderating role of screening knowledge.
RESULTS: All groups demonstrated a statistically significant preference for participating in breast cancer screening, with substantial heterogeneity: high-risk disclosure strengthened participation willingness, while non-high-risk disclosure weakened it. Risk disclosure also reshaped attribute preferences, high-risk disclosure increased sensitivity to diagnostic accuracy, while non-high-risk disclosure shifted attention towards out-of-pocket cost and service-related attributes. Screening knowledge significantly moderated these effects, women with lower knowledge showed stronger responses to risk information and reduced cost sensitivity, suggesting simplified decision-making, whereas higher-knowledge women maintained stable multi-attribute trade-offs and were less influenced by disclosed risk information.
CONCLUSIONS: Personalised breast cancer risk disclosure not only affects screening participation willingness, but also reshapes women’s preferences for screening attributes, with the effects varying according to screening knowledge levels. These findings support the development of risk communication strategies integrated with risk status and screening-related knowledge, enhancing participation efficiency and optimizing resource allocation in breast cancer screening programmes.
METHODS: A total of 923 urban women were recruited using multi-stage stratified sampling. Participants were randomly assigned to four groups defined by objective risk status and disclosure (disclosed high-risk group, undisclosed high-risk group, disclosed non-high-risk group and undisclosed non-high-risk group). Screening preferences were elicited using a best-worst discrete choice experiment (BW-DCE) with seven attributes, including out-of-pocket cost, screening setting, pain level, physician gender, screening interval, false-negative rate, false-positive rate and an opt-out option to capture screening particpation decisions. A heteroscedastic conditional logit model was employed to estimate preferences and account for best-worst scale heterogeneity; interaction terms were used to examine the moderating role of screening knowledge.
RESULTS: All groups demonstrated a statistically significant preference for participating in breast cancer screening, with substantial heterogeneity: high-risk disclosure strengthened participation willingness, while non-high-risk disclosure weakened it. Risk disclosure also reshaped attribute preferences, high-risk disclosure increased sensitivity to diagnostic accuracy, while non-high-risk disclosure shifted attention towards out-of-pocket cost and service-related attributes. Screening knowledge significantly moderated these effects, women with lower knowledge showed stronger responses to risk information and reduced cost sensitivity, suggesting simplified decision-making, whereas higher-knowledge women maintained stable multi-attribute trade-offs and were less influenced by disclosed risk information.
CONCLUSIONS: Personalised breast cancer risk disclosure not only affects screening participation willingness, but also reshapes women’s preferences for screening attributes, with the effects varying according to screening knowledge levels. These findings support the development of risk communication strategies integrated with risk status and screening-related knowledge, enhancing participation efficiency and optimizing resource allocation in breast cancer screening programmes.
Conference/Value in Health Info
2026-09, ISPOR Asia Pacific 2026, Bangkok, Thailand
Value in Health, Volume 55, Issue S1
Code
PCR18
Topic
Patient-Centered Research
Disease
SDC: Oncology, SDC: Reproductive & Sexual Health, STA: Personalized & Precision Medicine