S.I.N.A.R (Sembanglah, Ini Aman dan Rahsia) - translated to Speak Freely, It is Safe and Confidential

The problem

Felda Bersia Health Clinic (FBHC) is a rural primary care clinic in Hulu Perak, Perak, Malaysia, serving traditional villages and Indigenous Orang Asli settlements, including communities accessible only by boat or helicopter. Mental healthcare delivery in these settings is challenged by multiple intersecting barriers, including stigma, low mental health literacy, workforce limitations, geographical isolation, and delayed help-seeking.

Prior to the project, many individuals were reluctant to seek mental health support due to fear of judgement, cultural beliefs, concerns regarding confidentiality, and limited awareness of available services. Access to psychologists and mental health professionals was also limited, particularly in rural and underserved areas. As a result, many patients delayed seeking help until symptoms became more severe, while healthcare workers faced difficulties providing timely psychosocial support within existing workforce constraints.

A pre-implementation survey conducted in August 2023 highlighted these gaps:

• Only 39% knew where to seek mental health support
• Only 29% had previously sought mental health assistance
• 51% reported concerns about social judgement or stigma
• 63% identified lack of family or social support as a barrier to seeking care.

The limitations of traditional face-to-face mental healthcare models became even more apparent during crisis situations, including the Gerik floods in 2024 and a mass casualty bus crash involving university students in June 2025. These events required rapid and scalable psychosocial support for survivors, families and frontline responders across geographically dispersed settings, where conventional service models alone were insufficient.

The team recognised that the challenge was not solely the availability of mental health services, but the absence of accessible, low-stigma and scalable entry points into mental healthcare, particularly within rural primary care systems.

The solution

To address these gaps, FBHC developed S.I.N.A.R. (“Sembanglah, Ini Aman dan Rahsia” or “Speak Freely, It Is Safe and Confidential”), a hybrid digital mental health ecosystem designed to provide anonymous, accessible and early psychosocial support within a rural public healthcare setting.

The initiative combined low-tech and AI-enabled approaches into a single integrated access model:

• an anonymous physical drop-box placed outside the clinic
• QR code-based digital forms accessible via smartphones
• a 24/7 AI chatbot accessible through web and mobile devices.

This multi-access approach was intentionally designed to accommodate varying levels of digital literacy, internet connectivity and personal comfort with help-seeking across different communities.

The AI chatbot functioned as a first-line psychosocial support and navigation tool. It was designed to:

• provide immediate mental health information and supportive engagement
• guide users towards appropriate healthcare and support services
• reduce barriers related to stigma, accessibility and waiting times
• widen access to early psychosocial support in areas with limited mental health workforce availability.

The chatbot was not intended to replace healthcare professionals or provide clinical diagnosis. Instead, it functioned as an early access and navigation pathway within the broader mental healthcare ecosystem. Escalation pathways to healthcare providers and mental health teams remained in place, particularly for users requiring further assessment or support.

The project was developed using human-centred design and design thinking principles, informed by community engagement sessions, baseline surveys and local needs assessments. The initiative was intentionally designed to be low-cost, scalable and adaptable within existing government primary healthcare workflows.

The impact

S.I.N.A.R. improved access to early mental health support within a rural primary care setting while helping reduce barriers related to stigma, awareness and accessibility.

1. Improved awareness and willingness to seek support
A pre- and post-implementation survey conducted between August 2023 and January 2024 demonstrated measurable improvements in mental health awareness and help-seeking attitudes among service users and community members. Awareness of mental health services available at the clinic increased from 40% to 65%, knowledge of where to seek help increased from 39% to 76%, willingness to seek mental health support increased from 29% to 55%, and willingness to use anonymous support channels increased from 69% to 88%.

2. Reduced barriers to care
The initiative was also associated with reductions in several commonly reported barriers to seeking mental healthcare. Fear of judgement decreased from 51% to 35%, cultural and belief-related barriers decreased from 43% to 25%, and lack of family or social support as a reported barrier decreased from 63% to 30%.

3. Expanded access without substantial additional workforce requirements
Operationally, the initiative created an additional digital access pathway into mental healthcare without requiring substantial expansion of the existing workforce. The AI chatbot enabled continuous access to supportive engagement and mental health information, while reducing dependency on face-to-face interactions for early psychosocial support and service navigation.

4. Adaptability during emergency and crisis response
Beyond routine clinic operations, the system was also deployed during emergency and crisis situations, including the Gerik floods in 2024 as part of Psychological First Aid (PFA) efforts, and during a 2025 mass casualty incident involving university students, survivors, families and frontline responders. These experiences demonstrated the flexibility of the model in supporting rapid psychosocial engagement during public health emergencies and disaster-related responses.

5. National and international recognition
The initiative has since been presented and shared at national and international platforms, adapted or replicated across healthcare and non-healthcare settings, and highlighted during a World Health Organization (WHO) visit to the clinic in March 2025.

Lessons learned

Technology alone does not resolve access barriers in mental healthcare. Trust, anonymity and ease of access were equally important in influencing engagement. Some users preferred the physical anonymous drop-box, while others were more comfortable using QR forms or the AI chatbot. Providing multiple access pathways helped accommodate differences in age, digital literacy, connectivity and personal comfort with help-seeking.

The project also reinforced that AI should support healthcare professionals rather than replace them. The chatbot functioned most effectively as a first-line support and service navigation tool, particularly in settings with limited access to psychologists and mental health professionals.

Implementation within a public healthcare system required balancing innovation with privacy, safety, operational practicality and workforce realities. Building confidence among healthcare workers and users required time, particularly around concerns related to confidentiality, trust and the role of AI in mental healthcare delivery.

What worked well

• Low-cost implementation using existing infrastructure

  The initiative was implemented without major additional investment by leveraging existing digital tools, workflows and available infrastructure within the clinic.

• A hybrid physical-digital access model

  Combining physical tools such as the anonymous drop-box with digital pathways including QR forms and the chatbot allowed the initiative to remain accessible across varying levels of digital literacy and technology access.

• Adaptability during crisis situations

  The model proved sufficiently flexible to support psychosocial engagement during the Gerik floods and a mass casualty incident, demonstrating value beyond routine clinic-based operations.

Challenges

• Digital literacy and connectivity limitations

  Variations in digital literacy and internet connectivity in rural areas meant that not all users were equally able to engage with digital pathways.

• Safe escalation for high-risk cases

  Ensuring that individuals identified as potentially high-risk could be escalated safely and appropriately to healthcare providers remained a key operational consideration.

What the team would do differently

• Enhanced triage and escalation capabilities.
  Future iterations of the platform could incorporate more advanced triage and risk-flagging mechanisms to support earlier identification of users requiring urgent assessment or escalation to healthcare professionals, while maintaining appropriate human oversight and clinical governance.

• More inclusive and accessible user engagement features. Future development could include multilingual support, voice-assisted interactions and avatar-based conversational interfaces to improve accessibility and engagement across different age groups, literacy levels and communities with varying digital confidence.

• Population-level planning dashboards. Developing population-level dashboards to monitor mental health trends, engagement patterns and service demand could support longer-term service planning, workforce allocation and public health decision-making.

S.I.N.A.R. was developed within the Malaysian public healthcare system under the Ministry of Health Malaysia (MOH), primarily within a rural primary care setting. The design and implementation of the initiative were shaped by multiple regulatory, operational, ethical and clinical governance considerations.

Patient confidentiality and privacy were key priorities throughout development. The system was intentionally designed to minimise the collection of personally identifiable information, with users able to remain anonymous unless they voluntarily disclosed information for follow-up, referral or escalation purposes. Digital interactions and data handling were guided by principles of patient confidentiality and aligned with Malaysia’s Personal Data Protection Act (PDPA).

Operationally, the initiative was designed to function within existing MOH workflows, workforce limitations and public-sector resource constraints. Rather than requiring enterprise-level infrastructure or major procurement processes, the system was intentionally developed as a low-cost and lightweight solution using existing digital tools, low-cost hosting and scalable implementation approaches suitable for rural primary healthcare settings.

From a clinical governance perspective, the AI chatbot was positioned as a first-line psychosocial support and service navigation tool rather than a replacement for healthcare professionals or a diagnostic system. Escalation pathways to medical officers, family medicine specialists and mental health teams remained in place, particularly for users identified as potentially high-risk or requiring further clinical assessment. Human oversight remained an important component of the overall care model.

Future development considerations, including enhanced triage capabilities, voice-assisted interactions and avatar-based engagement features, would require continued attention to safety, accessibility, governance and appropriate clinical oversight.

The initiative was also aligned with several national policy directions, including Malaysia’s National Strategic Mental Health Plan 2020-2025, the Ministry of Health Digitalisation Strategic Plan 2021-2025, and broader national priorities related to improving access, early intervention and community-based mental healthcare.