Yi-Roe Tan

Yi-Roe Tan

Auto-generated profile, based on publicly available information

Yi-Roe Tan

HealthAI - The Global Agency for Responsible AI in Health, Geneva, Switzerland

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 August 2023 in Journal of Global Antimicrobial Resistance
Verified
Journal Article
Open access
ZHZhilian Huang
Zhilian Huang
MGMithun Mohan George
Mithun Mohan George
YTYi-Roe Tan
Yi-Roe Tan
KNKarthiga Natarajan
Karthiga Natarajan
EDEmily Devasagayam
Emily Devasagayam
ACAngela Chow
Angela Chow
Artificial intelligence (AI)-driven clinical decision support systems (CDSSs) can augment antibiotic decision-making capabilities, but physicians' hesitancy in adopting them may undermine their utility. We conducted a cross-country comparison of physician perceptions on the barriers and facilitators in accepting an AI-enabled CDSS for antibiotic prescribing. We conducted in-depth interviews with physicians from the National Centre for Infectious Diseases (NCID), Singapore, and Christian Medical College Vellore (CMCV), India, between April and December 2022. Our semi-structured in-depth interview guides were anchored on Venkatesh's UTAUT model. We used clinical vignettes to illustrate the application of AI in clinical decision support for antibiotic prescribing and explore medico-legal concerns. Most NCID physicians felt that an AI-enabled CDSS could facilitate antibiotic prescribing, while most CMCV physicians were sceptical about the tool's utility. The hesitancy in adopting an AI-enabled CDSS stems from concerns about the lack of validated and successful examples, fear of losing autonomy and clinical skills, difficulty of use, and impediment in work efficiency. Physicians from both sites felt that a user-friendly interface, integration with workflow, transparency of output, a guiding medico-legal framework, and training and technical support would improve the uptake of an AI-enabled CDSS. In conclusion, the acceptance of AI-enabled CDSSs depends on the physician's confidence with the tool's recommendations, perceived ease of use, familiarity with AI, the organisation's digital culture and support, and the presence of medico-legal governance of AI. Progressive implementation and continuous feedback are essential to allay scepticism around the utility of AI-enabled CDSSs.

18 January 2025 in Oxford Open Digital Health
Verified
Journal Article
Open access
EBEvangelia Baka
Evangelia Baka
YTYi-Roe Tan
Yi-Roe Tan
BWBrian Li Han Wong
Brian Li Han Wong
ZXZhongyue Xing
Zhongyue Xing
PYPeiling Yap
Peiling Yap
Digital mental health (DMH) interventions leveraging digital technologies, such as mobile applications, web-based platforms, artificial intelligence and wearable devices, have emerged as a promising avenue for addressing the mental health needs of young people. This scoping review examines the landscape of primary and secondary preventive DMH interventions for young people aged 10–24 years. Six electronic databases were searched, leading to a final incorporation of 81 studies published between 2010 and 2022. Each of these studies corresponds to a unique DMH intervention. Our findings reveal that research activity in the area of promotive and preventive DMH interventions started gaining ground from 2019 onwards, with the majority of studies conducted in Australia and the USA. 70% of the total studies targeted the prevention of mental health conditions. Randomized controlled trials were the predominant study methodology, while mental well-being, depressive disorders, anxiety disorders, life skills and disorders specifically associated with stress were the most targeted mental health or well-being conditions. Finally, mobile applications and web interfaces were the most studied form of DMH interventions. Most of these applications have integrated advanced AI/ML algorithms to serve the purpose of personalization and real-time monitoring. However, there is a marked need for more emphasis on preventive and, especially, promotive mental health measures, as well as the active inclusion of low- and middle-income countries in future research.

15 January 2021 in Sexually Transmitted Infections
Verified
Journal Article
Open access
PDF file
YTYi-Roe Tan
Yi-Roe Tan
NKNashwinder Kaur
Nashwinder Kaur
AYAngeline Jiajun Ye
Angeline Jiajun Ye
YZYiwen Zhang
Yiwen Zhang
JLJerald Xuan Zheng Lim
Jerald Xuan Zheng Lim
PYPeiling Yap
Peiling Yap
Objectives Voluntary HIV testing rates are still low in several Asian countries including Singapore. HIV self-testing (HIVST) has the potential to increase testing, leading to earlier diagnosis and better prognosis. However, the views of at-risk individuals, especially heterosexual men (HSM), who are not coming forward for testing are still poorly understood. In this study, we examined the barriers and facilitators to and delivery preferences for HIVST in order to implement an effective intervention in Singapore. Methods From May 2017 to June 2018, 48 in-depth interviews were conducted with HSM aged 21–66 years and at risk of HIV infection. Participants were purposively sampled based on ethnicity, age and testing behaviour. Recruitment was done mainly at brothels and entertainment establishments in Singapore. Participants gave their views on HIV testing, factors affecting HIVST use and their preferred HIVST service delivery model. Results Most participants preferred HIVST over conventional testing for its convenience, privacy, anonymity and autonomy, but older men still preferred conventional testing. Low self-perceived risk, low awareness and self-efficacy for HIVST, and non-comprehensive test for other STIs were reported as barriers to HIVST. There were mixed opinions on kit preference. A blood-based kit was favoured for higher accuracy, while the oral-fluid-based kit was favoured for ease of use. Participants wanted a human touch for post-test counselling and linkage to care only if they self-tested positive. Traditional media, internet and social media, and venue-based outreach were potential advertising platforms mentioned. Conclusions A locally acceptable and feasible HIVST intervention must address the barriers and facilitators of using HIVST in order to improve HIV testing rates among this at-risk population who might otherwise delay or fail to present for testing.

19 March 2024 in BMJ Global Health
Verified
Journal Article
Open access
PDF file
YTYi-Roe Tan
Yi-Roe Tan
MNManh Duc Nguyen
Manh Duc Nguyen
CMCaroline Antonia Mubaira
Caroline Antonia Mubaira
DKDan Kajungu
Dan Kajungu
DKDinesh Kumar
Dinesh Kumar
PYPeiling Yap
Peiling Yap
Introduction Citizen science (CS) is an emerging approach in public health to harness the collective intelligence of individuals to augment traditional scientific efforts. However, citizens’ viewpoint, especially the hard-to-reach population, is lacking in current outbreak-related literature. We aim to understand the awareness, readiness and feasibility of outbreak-related CS, including digitally enabled CS, in low-income and middle-income countries. Methods This mixed-method study was conducted in nine countries between October 2022 and June 2023. Recruitment through civil society targeted the general population, marginalised/indigenous groups, youth and community health workers. Participants (aged ≥18 years) completed a quantitative survey, and a subset participated in focus group discussions (FGDs). Results 2912 participants completed the survey and 4 FGDs were conducted in each country. Incorporating participants’ perspectives, CS is defined as the practice of active public participation, collaboration and communication in all aspects of scientific research to increase public knowledge, create awareness, build trust and facilitate information flow between citizens, governments and scientists. In Bangladesh, Indonesia, the Philippines, Cameroon and Kenya, majority were unaware of outbreak-related CS. In India and Uganda, majority were aware but unengaged, while in Nepal and Zimbabwe, majority participated in CS before. Engagement approaches should consider different social and cultural contexts, while addressing incentivisation, attitudes and practicality factors. Overall, 76.0% expressed interest in digital CS but needed training to build skills and confidence. Digital CS was perceived as convenient, safer for outbreak-related activities and producing better quality and quantity of data. However, there were concerns over non-inclusion of certain groups, data security and unclear communication. Conclusion CS interventions need to be relatable and address context-specific factors influencing CS participation. Digital CS has the potential to facilitate collaboration, but capacity and access issues must be considered to ensure inclusive and sustainable engagement.

7 December 2020 in JMIR mHealth and uHealth
Verified
Journal Article
Open access
MLMay Oo Lwin
May Oo Lwin
JLJiahui Lu
Jiahui Lu
ASAnita Sheldenkar
Anita Sheldenkar
CPChitra Panchapakesan
Chitra Panchapakesan
YTYi-Roe Tan
Yi-Roe Tan
BABrenda Ang
Brenda Ang
Background: Existing studies have suggested that internet-based participatory surveillance systems are a valid sentinel for influenza-like illness (ILI) surveillance. However, there is limited scientific knowledge on the effectiveness of mobile-based ILI surveillance systems. Previous studies also adopted a passive surveillance approach and have not fully investigated the effectiveness of the systems and their determinants.Objective: The aim of this study was to assess the efficiency of a mobile-based surveillance system of ILI, termed FluMob, among health care workers using a targeted surveillance approach. Specifically, this study evaluated the effectiveness of the system for ILI surveillance pertaining to its participation engagement and surveillance power. In addition, we aimed to identify the factors that can moderate the effectiveness of the system.Methods: The FluMob system was launched in two large hospitals in Singapore from April 2016 to March 2018. A total of 690 clinical and nonclinical hospital staff participated in the study for 18 months and were prompted via app notifications to submit a survey listing 18 acute respiratory symptoms (eg, fever, cough, sore throat) on a weekly basis. There was a period of study disruption due to maintenance of the system and the end of the participation incentive between May and July of 2017.Results: On average, the individual submission rate was 41.4% (SD 24.3%), with a rate of 51.8% (SD 26.4%) before the study disruption and of 21.5% (SD 30.6%) after the disruption. Multivariable regression analysis showed that the adjusted individual submission rates were higher for participants who were older (<30 years, 31.4% vs 31-40 years, 40.2% [P<.001]; 41-50 years, 46.0% [P<.001]; >50 years, 39.9% [P=.01]), ethnic Chinese (Chinese, 44.4% vs non-Chinese, 34.7%; P<.001), and vaccinated against flu in the past year (vaccinated, 44.6% vs nonvaccinated, 34.4%; P<.001). In addition, the weekly ILI incidence was 1.07% on average. The Pearson correlation coefficient between ILI incidence estimated by FluMob and that reported by Singapore Ministry of Health was 0.04 (P=.75) with all data and was 0.38 (P=.006) including only data collected before the study disruption. Health care workers with higher risks of ILI and influenza such as women, non-Chinese, allied health staff, those who had children in their households, not vaccinated against influenza, and reported allergy demonstrated higher surveillance correlations.Conclusions: Mobile-based ILI surveillance systems among health care workers can be effective. However, proper operation of the mobile system without major disruptions is vital for the engagement of participants and the persistence of surveillance power. Moreover, the effectiveness of the mobile surveillance system can be moderated by participants’ characteristics, which highlights the importance of targeted disease surveillance that can reduce the cost of recruitment and engagement.

17 June 2025 in JMIR Preprints
Verified
Preprint
Open access
EBEvangelia Baka
Evangelia Baka
NKNiklas Krischer
Niklas Krischer
USUdani De Silva
Udani De Silva
YTYi-Roe Tan
Yi-Roe Tan
PYPeiling Yap
Peiling Yap
BWBrian Li Han Wong
Brian Li Han Wong
UNSTRUCTURED Artificial intelligence (AI) is rapidly transforming healthcare and health research, offering new opportunities for improving efficiency, accessibility, and equity. However, the ethical, societal, and regulatory challenges of AI development and deployment are particularly pronounced in low- and middle-income countries (LMICs). While existing literature often emphasises high-level ethical principles or technical frameworks, there is a notable gap in empirical, qualitative research that centers on human involvement and sociocultural dynamics throughout the AI lifecycle in LMIC contexts. This study addresses this gap by exploring the role of human involvement across the AI lifecycle, examining how cultural, societal, and governance factors influence AI perceptions and expectations in LMICs. Through 21 qualitative interviews with AI researchers and innovators across MENA, Africa, Latin America, and Asia, we identified five key themes: (1) the necessity of human oversight and the readiness required to support it, (2) the need for AI ethics training, (3) the importance of developing AI systems tailored to local realities, (4) the role of human-centered AI governance, and (5) the value of securing multidisciplinary teams. Findings highlight critical gaps in AI literacy, ethical governance, and interdisciplinary collaboration, emphasising that AI solutions must be co-designed with local communities to be culturally and contextually relevant. This study underscores the urgent need for participatory AI development in LMICs and calls for investment in AI education, ethical oversight, and inclusive governance frameworks to ensure that AI serves as a tool for social equity rather than exclusion.