Breast Cancer Screening Knowledge and Sentiments in Singaporean Women: Mixed Methods Study Using Topic Modeling,...

Mammography screening uptake in Singapore remains below 40% despite campaigns and subsidies. Natural language processing (NLP) can extract nuanced attitudes from free text that fixed response options miss, revealing latent factors influencing breast cancer (BC) screening behavior.

This study characterized women's attitudes toward mammography using mixed methods data, examined associations between BC awareness and screening willingness, and identified barriers and facilitators through NLP of free-text responses.

We conducted a cross-sectional study within the Breast Screening Tailored for Her multicenter cohort in Singapore (October 2021-December 2023). In total, 4169 women aged 35-59 years (median 48, IQR 43-54) were recruited via convenience sampling (3 hospitals and 2 polyclinics). Participants completed online structured questionnaires on demographics and screening history, then a BC education quiz with feedback. Participants answering >80% correctly were classified as "BC-aware." Posteducation, participants reported screening willingness (motivated or neutral) with optional free-text explanations. Logistic regression models (adjusted for study site, age, ethnicity, marital status, housing, and education) examined the associations with willingness. For 3819 English-language respondents, biterm topic modeling identified themes and sentiment analysis quantified emotional tone. Statistical significance: α=.05.

Overall, 79% (3287/4169) were BC-aware, and 94% (3908/4169) reported increased motivation posteducation. BC-aware women had higher screening motivation than BC-unaware women (adjusted odds ratio [aOR] 2.88, 95% CI 2.19-3.80; P<.001). Motivation was higher among those with larger public housing (OR 1.81, 95% CI 1.30-2.50; P<.001) and private housing vs 1-3 room units (OR 2.69, 95% CI 1.75-4.13; P<.001), married vs not separated, divorced, or widowed (OR 2.38 [inverse of 0.42], 95% CI 1.75-3.13; P<.001), and prior screening attendance (OR 3.49, 95% CI 2.71-4.50; P<.001). Women who disagreed that mammography was expensive had higher motivation (aOR 1.94, 95% CI 1.50-2.50; P<.001). Among 3819 English respondents, 94% (3579/3819) were motivated and 6% (240/3819) neutral. Free-text responses came from 34% (1220/3579) of motivated and 64% (153/240) of neutral participants. Biterm topic modeling revealed motivated participants emphasized early detection benefits, health awareness, BC risk, and logistics; neutral participants focused on mammography pain experiences and cost barriers. Mean sentiment was 0.207 (range: -1.00 to 1.65), with motivated participants displaying more positive sentiments than neutral participants (linear regression, P<.001). Identical words carried different emotional tones across subgroups: "health" had positive sentiment among motivated participants (mean difference, Welch t tests P<.05) but negative sentiment among neutral participants. Word frequency analysis showed motivated participants used positive-sentiment words ("better," "cure," and "prevention"). Neutral participants emphasized negative words ("painful" and "uncomfortable").

Integrating quantitative surveys with NLP revealed that the same screening concepts are emotionally framed differently by motivated vs neutral women, a finding missed by knowledge- or intent-focused approaches alone. In practice, these findings support the need for emotionally tailored BC education and prevention strategies.