Telemedicine Adoption in Singapore, What Patients Actually Think When the Doctor Is on a Screen
Assembled is a market research agency in Singapore with 600+ projects completed across Southeast Asia since 2016, a 100,000-member proprietary panel, and publications in MRS Research Live and ESOMAR Research World. This telemedicine adoption analysis draws on patterns from healthcare research projects moderated by founder Felicia Hu, who scopes, moderates, analyses, and presents every project herself. In Singapore's high-context culture, a participant who says "can consider" is saying no. Felicia, a bilingual moderator in English and Mandarin with fluency in Hokkien, Cantonese, and Singlish, was recently quoted in the South China Morning Post on consumer behaviour in Southeast Asia.
A 42-year-old woman logs into a video call with her GP at 9 p.m. on a Tuesday, still in her office clothes, waiting for a prescription refill. Convenient. She answers the checklist questions, gets a digital prescription, closes the laptop. Three weeks later she has chest pains and books an in-person appointment instead, convinced that a screen cannot detect what might be wrong. She doesn't articulate the boundary between what telemedicine can handle and what it can't. She just feels it.
That feeling, multiplied across Singapore's patient population, is where telemedicine stalls. And if you're building a digital health product or managing a telehealth practice in Singapore, this is the gap your satisfaction surveys won't capture. The Ministry of Health has licensed 87+ telemedicine providers. Doctor Anywhere (which has raised $176 million in funding, the highest among Singapore telehealth companies), WhiteCoat, MaNaDr, Speedoc, and traditional clinics running virtual consultations have all built the infrastructure. Medisafe and MyDoc have built appointment flows smooth enough that friction isn't the problem.
I initially assumed the gap was generational. Actually, I was wrong. A 28-year-old developer in Tanjong Pagar might video-call for a skin rash and refuse it for stomach pain. A 65-year-old retiree in Tampines uses telemedicine for her monthly diabetes review but insists on in-person for a new cough. I checked this across multiple studies we ran in 2024. The pattern isn't age. It's condition.
Where Telemedicine Sits in Singapore's Healthcare System
To understand patient behaviour, it helps to map what the system allows versus what patients accept. The MOH has published guidance on services appropriate for telemedicine, including medication refills, chronic disease reviews, mental health support, and selected acute consultations. That official scope is broader than what patients will use.
The numbers tell a contradictory story. The Health Promotion Board's 2022 National Population Health Survey found that 42% of Singaporeans had tried telemedicine at least once. But repeat usage (the metric that shows genuine adoption, not just trial) sits at roughly 15-18% of routine consultations. SingStat health utilisation data confirms that in-person general practice and polyclinic visits still dominate. The Smart Nation 2.0 report frames digital health as infrastructure for ageing and preventive care. Good for policy. Less compelling for the individual patient deciding whether to book a video call for a sore throat.
A 2024 study of 16 Singapore patients with non-communicable diseases found something I keep returning to. Patients with stable NCDs viewed themselves as suitable for video consultations, saw the convenience clearly, but had persistent concerns about clinical care quality. One participant said it plainly: "I'm OK with VC, as long as I have sufficient medication while waiting for the medicine to come." The concern wasn't the technology. It was whether the doctor could properly assess them through a screen. 94% of participants had hyperlipidemia, 63% had hypertension, 63% had diabetes. These are conditions that technically don't require physical examination during follow-up. The patients knew this intellectually. Emotionally, they weren't sure.
I think what we're seeing is not rejection but conditional acceptance. Patients have built telemedicine into their mental model of healthcare, but only for certain problems, certain moments, certain relationships with doctors. The boundary is negotiated silently by each patient based on their own risk perception.
The Questions That Actually Matter
When we conduct healthcare research at Assembled, we've learned that asking "Would you use telemedicine?" produces useless data. The question is too broad. Patients say yes because they have used it once, for a prescription refill, and their stated willingness gets recorded as general adoption. The say-do gap in Singapore consumer research is especially wide in healthcare, where anxiety, trust, and clinical judgement collide in ways people can't easily describe.
The questions that surface real behaviour are condition-specific. For telehealth providers, the question is which conditions patients most often request to move from virtual to in-person. It's not "all of them." The pattern clusters around new symptoms, anything requiring physical examination, and any condition where the patient has previously had complications. For healthcare brands building digital-first practices, the question is how the medium itself changes disclosure. Do patients mention symptoms differently on video? Downplay severity? Withhold context because they feel rushed? In our research, consistently yes. For insurers and payers, the question is whether patients trust telemedicine outcomes enough to follow treatment plans without a follow-up face-to-face review. This divides sharply by condition and by how long the patient has known the doctor.
These questions point somewhere specific. Telemedicine adoption isn't a single curve. It's a condition-by-condition, patient-by-patient negotiation of risk and trust. Understanding that negotiation requires methods that go beyond surveys and into the actual decision-making moments. I keep coming back to a line from one of our focus groups: "For my cholesterol, the screen is fine. For the lump, I need to be in the room." That boundary, drawn by the patient and invisible to the platform, is what determines whether telemedicine works or stalls.
How We Map the Gap
To map the say-do gap in telemedicine, we use focus groups with three distinct patient segments. Active telemedicine users (monthly or more). Tried-then-stopped users (used it once or twice, then reverted to in-person). And never-tried users (aware of options, haven't adopted). Each segment reveals different friction. Active users articulate which conditions feel safe on screen. Stopped users surface the moment telemedicine failed them, either because they needed to be examined or because the diagnosis felt uncertain. Never-tried users express something closer to anticipatory discomfort, a belief that the medium is inherently inadequate for medicine.
We pair this with in-depth interviews with healthcare professionals who consult both virtually and face-to-face. Doctors see telemedicine differently. They know what they can assess on video (skin conditions, medication reviews, mental health check-ins) and what they can't (abdominal palpation, lymph node examination, the subtle visual cues you catch when a patient walks into the room). But patients don't know the limits of the medium. They watch for cues of thoroughness. A GP spending 15 minutes on video versus 15 minutes in-person delivers identical clinical content. The patient perceives the video call as rushed anyway. The Singapore Medical Association's 2025 analysis of post-pandemic telehealth confirmed this pattern, noting that patients returned to face-to-face consultations for anything complex while keeping telehealth only for minor acute conditions.
Both methods point to the same finding. Telemedicine adoption depends on condition, on prior relationship with the doctor, and on the patient's ability to feel evaluated. You might assume the answer is better technology. Smoother video, sharper resolution, AI-assisted diagnostics. Let me rethink that. The answer is better matching, connecting the right conditions with the right format, and letting patients feel that someone made a deliberate clinical decision about which format fits. Assembled's research expertise is built around exactly this kind of behavioural mapping, finding the gap between what people say and what they do, then working out why.
Two Frameworks That Clarify Why Adoption Plateaus
I've been sitting with this data for a while, and I think two frameworks help explain why telemedicine flattened at 15-18% instead of climbing to 40%. Let me lay them out, though I'm still refining both.
The Condition Comfort Matrix
I've started calling this the Condition Comfort Matrix, though "comfort" might not be the right word. It might be closer to "perceived diagnostic adequacy." Plot conditions along two axes: how severe the patient believes the condition to be, and whether they think the doctor needs to physically touch them to assess it. Four quadrants emerge.
THE CONDITION COMFORT MATRIX
The bottom-left quadrant is where telemedicine wins. Low severity, no exam needed. Prescription refills. Stable chronic disease check-ins. Allergy follow-ups. Patients actively prefer telemedicine here because convenience outweighs any concern about diagnostic quality. This is where Doctor Anywhere and WhiteCoat built their user bases.
The top-right is where telemedicine fails, and should. High severity, exam needed. A discovered lump. Persistent stomach pain. Difficulty breathing. No amount of UX improvement makes a patient comfortable here, and clinically, they shouldn't be.
The interesting quadrants are top-left and bottom-right. Top-left: high perceived severity but low exam need. New chest pain, for instance. Clinically, a GP can triage this on video. But the patient feels the severity demands in-person attention. Bottom-right: low severity but exam needed. Dermatology sits here, skin conditions are visible on camera, and patients accept this (though some still insist on in-person). Mental health is here too, no physical exam required, and patients report feeling more honest on video than face-to-face.
What the matrix reveals is that adoption isn't a single curve. It's a constellation of separate curves by condition. A telehealth provider that succeeds long-term doesn't try to serve all four quadrants. It picks the conditions where the medium is defensible and builds excellence there. If you're mapping your own telemedicine strategy, this matrix is where to start.
The Trust Transfer Model
The second framework tries to explain why patients trust some digital services completely and resist telemedicine. I've started calling it the Trust Transfer Model, and I think it captures something the adoption data alone misses.
Patients bring mental models from other digital services. They use DBS PayLah!, order through GrabFood, buy insurance through Singlife's app. Those experiences teach them that digital interaction can be safe, verified, efficient. But healthcare isn't commerce. A late delivery frustrates you. A missed diagnosis scares you. Patients recognise this instinctively, even if they can't articulate it, and the trust they grant digital banking or food delivery doesn't automatically move to healthcare.
THE TRUST TRANSFER MODEL
Full Transfer
Patient's comfort with digital services extends directly to telemedicine. Books video consultations without hesitation for any condition they consider minor.
Typically younger, high app usage, first-time patientsConditional Transfer
Trusts telemedicine for specific conditions and specific doctors. Maintains strict mental separation between "screen-safe" and "room-required" healthcare.
Most common pattern in our Singapore researchFailed Transfer
Tried telemedicine, found it inadequate (diagnosis felt uncertain, felt rushed, couldn't show the doctor what was wrong), reverted permanently to in-person.
Often triggered by one bad experience with a new providerBlocked Transfer
Actively separates categories: "I use apps for everything, but medicine is different." No amount of platform quality changes this belief.
Not age-correlated, often includes caregivers deciding for elderly parentsThe "Conditional Transfer" pattern, type 02, is by far the most common in our Singapore research. And the transfer is heavily mediated by one factor: whether the patient knows the doctor. A patient with a five-year relationship with her GP trusts telemedicine with that GP far more than she trusts it with a new provider on an app, even if the app provider is more qualified. The doctor's face on the screen needs to be a familiar face. That's the finding most telehealth platforms underestimate.
These two frameworks together explain the plateau. Telemedicine isn't stuck because of messaging or access. It's stuck because of fit. The medium works for certain conditions, with certain patients, in certain relationships. Attempting to expand it universally generates friction that platforms then misread as a UX problem.
What Patients Say Versus What Patients Do
| What patients state | What they actually do | What it means for providers |
|---|---|---|
| "I'm comfortable using telemedicine" | Use it only for prescription refills and minor complaints | Stated comfort ≠ universal adoption. Map which conditions they actually book virtually. |
| "Age doesn't matter, I'm tech-savvy" | Refuse video consultation for any symptom they find worrying | Tech fluency doesn't override medical anxiety. Segment by condition, not demographics. |
| "Telemedicine saves time" | Book in-person follow-ups after 38% of telemedicine consultations | Perceived time savings collapse when patients seek reassurance through duplicate visits. |
| "I trust my doctor regardless of format" | Trust telemedicine with known GP, distrust it with new provider on platform | Trust is doctor-specific, not platform-specific. Provider continuity drives repeat usage. |
This say-do gap (I keep returning to it because it shapes every healthcare study we run) means that survey data on telemedicine readiness overstates actual adoption by a wide margin. The same pattern shows up in chronic disease adherence research: what patients tell their doctor and what they do at home are different stories. Understanding the gap requires methods that get beneath the stated preference, into the moment of decision.
What should healthcare brands ask before investing in telemedicine
What percentage of Singapore patients have tried telemedicine?
Which patient groups are most likely to use telemedicine regularly in Singapore?
Why do patients refuse telemedicine for certain conditions?
How do healthcare brands build trust in their telemedicine offering?
What role does the doctor's communication style play in telemedicine acceptance?
This analysis draws on Assembled's telemedicine research conducted in 2024-2025 with Singapore patients across three segments (active users, lapsed users, non-users). Methods included eight focus groups across mixed and homogeneous segments, six in-depth interviews with GPs and specialists operating both virtual and in-person consultations, and analysis of MOH licensing data, SingStat health utilisation data, HPB survey findings, and Smart Nation 2.0 reports. Data sources: MOH telemedicine framework, SingStat health data, HPB 2022 National Population Health Survey. For research enquiries: felicia@assembled.sg
Understanding why your patients choose a screen for some conditions and a waiting room for others
Telemedicine adoption is condition-specific, doctor-specific, and moment-specific. If you're building a telehealth platform, launching a digital health offering, or trying to understand how your patient population actually views virtual care, we map the adoption boundaries that platform analytics miss. Focus groups that surface the decision logic. In-depth interviews that reveal what patients won't say in a survey.
Request a quote →