Trial to Evaluate the Impact of Various Behavioural Interventions to Increase Mammography Uptake Among Singaporean Women

Overview

Breast cancer is the leading cancer among women in Singapore, yet mammography screening rates remain critically low at 34.7%, despite the availability of a national program since 2002. This shortfall significantly limits the potential of screening to reduce mortality at a population level. While various behavioral interventions have been tested internationally to increase uptake, their applicability in the local context remains uncertain. Existing studies in Singapore are either outdated, small in scale, or not sufficiently targeted. In particular, two subgroups of women warrant closer examination due to the scarcity of data: (1) never-screeners: those who have never undergone a mammogram and (3) recent screeners: those who had a mammogram more than two years ago but have not scheduled their next one. These cohorts exhibit distinct behavioral and psychological barriers that prevent adherence to regular, guideline-recommended screening. E.g., key barriers for never- screeners include low perceived susceptibility to breast cancer, lack of awareness about the benefits of early detection, fear or anxiety about pain or receiving a cancer diagnosis, cultural or personal modesty concerns or lack of a physician's recommendation. In contrast, the repeat recent screeners often have a misconception that a single normal result is sufficient, negative experiences during prior screenings (e.g., false positives, pain), absence of reminder systems and competing life priorities and lack of time. Among these, engaging never-screeners is particularly challenging, as they may not perceive screening as personally relevant or necessary, and many have successfully avoided the healthcare system for years. Yet, failure to engage these groups; poses a significant public health concern, as it undermines the effectiveness of population-level cancer control strategies. The PROMPT study is designed to fill this critical gap by systematically evaluating multiple low-cost, scalable outreach strategies in two parallel randomized trials. Its findings will provide robust, local evidence to inform policy and improve engagement of these hard-to-reach populations.

The study comprises two parallel randomized controlled trials targeting two distinct populations: (1) Women who have never undergone a mammogram ("Never Screeners"): Eligibility criteria- Singaporean women, aged 50-69, who are eligible for free biennial mammograms but have not gone for mammogram yet, with no history of prior breast cancer, will be identified using the NUH electronic medical health records system (EPIC). The selection will be restricted to NUHS patients who are registered users of the NUHS App. To further select/narrow down the eligible participants, app-activity/ last login date will be used (Active app user is defined as having used the app at least once within 12 months of the study start date. If there are not enough active NUHS app user, the last login interval will be increased to 18 to 24 months until sufficient eligible participants are obtained. If the last login needs to be extended beyond 24 months, IRB approval will be sought). The required 3,000 patients for the study will be randomly selected from these eligible patients.(2) Women who have undergone mammogram in the past 2 years but not scheduled a repeat screening yet ("Recent Screeners"): Eligibility criteria- Singaporean women, aged 50-69, who are eligible for free biennial mammograms and have gone for mammogram in the past 18-24 months but have not scheduled an appointment for repeat screening yet, with no history of prior breast cancer, will be identified using the NUH electronic medical health records system (EPIC). Selection will be restricted to NUHS patients who are registered users of the NUHS App (as defined above). The required 5,910 patients for the study will be randomly selected from these eligible patients. The study will enroll a total of 3000 never-screeners (600 participants in each arm) and 5910 recent screeners (1182 participants in each arm) across 5 intervention groups. Sample size calculations were performed to ensure 90% power at a two-sided significance level of 5% (α = 0.05) to detect a clinically meaningful difference in the primary outcome, defined as an absolute increase of 7% in mammogram uptake. The study employs a ≥"pick-the-winner" design, wherein each intervention arm (Arms 2-5) is compared individually against the control arm (Arm 1). An intervention will be considered successful if it demonstrates at least a 7% absolute improvement in uptake relative to the control. Statistical analysis plan: Pairwise comparisons between each intervention arm (Arms 2-5) and the control arm (Arm 1) will be performed using logistic regression to estimate the odds ratios (ORs), absolute differences in proportions, and 95% confidence intervals (CIs) for mammogram uptake. Primary endpoint: All data on mammogram completion will be obtained from existing sources such as electronic medical records and the national screening registry. The primary outcome is mammogram uptake within four months of the first intervention (date of sending the mailer/ push notification, whichever is later). Uptake data might be also extracted at the 6 and 12 month time-point to see if there was any additional uptake. These extended follow-up points will allow for the assessment of any delayed uptake, which may help to understand the longer-term impact of the interventions. No new clinical data will be generated from participants. Screening behavior will be assessed through electronic medical record (EMR) review, specifically from the National University Health System's (NUHS) mammography booking and screening databases (the national screening registry). Data extracted will include mammogram booking and attendance status, date of screening (if any), and relevant demographic details. All data on mammogram completion will be obtained from existing sources such as electronic medical records. Findings will provide evidence-based strategies to improve breast cancer screening participation, potentially reducing breast cancer mortality and informing future public health policy.