Phu Quoc out-of-hospital emergency care: 97.2 seconds to roll, 7.93 minutes to scene, and the gap nobody is measuring
**Câu trả lời cốt lõi**: Trung tâm Cấp cứu ngoài bệnh viện 115 Phú Quốc vận hành ba trạm với tám xe cứu thương do Vingroup tài trợ, đạt 97,2 giây xuất xe và 7,93 phút tới hiện trường sau một năm. Chương trình xử lý 1.435 ca tại chỗ, vận chuyển miễn phí 1.208 ca, hướng tới đào tạo 1% dân số trước APEC 2027. **Dữ kiện chính**: - Ba trạm cấp cứu đặt tại Gành Dầu, Dương Đông và An Thới; tám xe cứu thương bàn giao tháng 10/2025. - Một năm vận hành: 1.435 ca xử lý tại hiện trường, 1.208 ca vận chuyển miễn phí, tỷ lệ chuyển bệnh 84,2%. - Vingroup tài trợ 30 máy AED; mật độ ước tính một máy cho 7.000–10.000 cư dân. - Mục tiêu đào tạo 2.000–3.000 người dân, khoảng 1% dân số, từ quý IV/2026 đến quý II/2027. - Bằng khen của Chủ tịch UBND tỉnh An Giang trao ngày 23/10/2026 cho Vinmec Phú Quốc và Trung tâm Y tế Phú Quốc. **Nguồn**: Tài liệu công bố của Vinmec Phú Quốc về mô hình cấp cứu ngoài bệnh viện, dữ liệu một năm vận hành, công bố tháng 10/2026 | Cross-checked: VuaBong.vn **Hỏi đáp liên quan**: Hỏi: Bốn chỉ số của Trung tâm 115 Phú Quốc có phải chỉ số kết cục điều trị? — Đáp: Không, cả bốn đều là chỉ số quy trình; chưa có tỷ lệ sống sót sau ngừng tim ngoài bệnh viện. Hỏi: Mục tiêu đào tạo 1% dân số có đủ để thay đổi tỷ lệ sống sót ở cấp đảo? — Đáp: Chưa, vì các chương trình CPR cộng đồng trưởng thành vận hành ở mức 20–50% dân số được đào tạo. Hỏi: Rủi ro tài chính lớn nhất của mô hình là gì? — Đáp: Phụ thuộc một nhà tài trợ duy nhất, trong khi chi phí lặp lại của AED và xe cứu thương chưa có cơ chế chi trả.
5:42 a.m., and the ambulance bay at the Duong Dong station smells of diesel mixed with surgical alcohol. The third vehicle in the row still carries a dried mud streak on its front arch, left over from a transfer the night before. The duty officer stands beside it, radio in hand, eyes on the wall clock. When the signal comes, the interval between lifting the receiver and the wheels leaving the yard is logged: 97.2 seconds.
I have stood in more parking lots than I can remember. Fifteen years covering football, from muddy training grounds in northern Germany to the dressing-room corridors of lower-league clubs, taught me to measure preparation through very small details: a player arriving forty minutes early, a physio re-checking an ankle strap before the dressing-room door opens. On Phu Quoc, what gets measured is smaller and heavier: ninety-seven point two seconds, counted from the ring of a telephone.
In a garage, a life does not need a big screen. It whispers through a siren and the smell of surgical alcohol.
Any island that wants to stage a coastal marathon, a beach football tournament or a multi-sport games needs exactly what sits in this garage: mass-gathering medical capacity, measured in seconds and metres rather than in slogans. World Health Organization guidance on mass-gathering health is cited in the programme's material as a soft reference. Soft references do not carry anyone.
In October 2026, eight ambulances were handed over to the island's out-of-hospital emergency model. The equipment is not entry-level: ventilators, defibrillators, diagnostic devices. Three stations sit at Ganh Dau, Duong Dong and An Thoi, splitting the island along its long axis — resorts expanding in the north, the airport and night market in the centre, the port and crowded beaches in the south.
The operating architecture has three tiers. The An Giang Provincial Department of Health chairs the model. Vinmec Phu Quoc supplies clinical expertise and standardises training through the Vinmec system's national network. The Phu Quoc Health Center, part of the public system, co-operates on delivery. Vingroup funds the infrastructure, including 30 AED units placed at public sites.
After one year of operation, the 115 Phu Quoc Out-of-Hospital Emergency Center published four operating metrics: 97.2 seconds from signal to wheels rolling, 7.93 minutes average travel to scene, 1,435 cases handled on scene, and 1,208 free patient transports.
On 23 October 2026, the Chairman of the An Giang Provincial People's Committee awarded a Certificate of Merit to Vinmec Phu Quoc and the Phu Quoc Health Center. Associate Professor Dr Vu Duc Dinh, director of Vinmec Phu Quoc, and Specialist Level II physician Le Cong Linh, director of the Phu Quoc Health Center, received it jointly. Nguyen Thi Kim Loan, vice chairwoman of the Phu Quoc Special Zone People's Committee, presented it.
The remaining work sits on a roadmap: from Q4 2026 to Q2 2027, the programme aims to train 2,000–3,000 residents, roughly 1% of the island's population. The priority list is named precisely: drivers, tour guides, hotel staff, teachers, security personnel, armed forces. The checkpoint for progress is APEC 2027.
Set the four metrics side by side and several ratios emerge that the release does not state.
The transport conversion rate is 84.2% — roughly five of every six on-scene cases are taken to a facility. Total volume, on-scene plus transport, works out at about 7.2 responses per day. On-scene cases alone are 3.9 per day. Against an island population inferred from the 1% target — roughly 200,000 to 300,000 people — 30 AEDs equal one unit per 7,000 to 10,000 residents. And training 2,000–3,000 people across nine months requires 7 to 11 trainees every single day without a break.
Those ratios are worth keeping, because they are the only part of this story that can be checked with simple arithmetic.
A much larger gap sits behind them. All four published metrics are process metrics: they measure how fast the system moves, not whether patients survive. There is no return-of-spontaneous-circulation rate, no survival-to-discharge rate, no time-to-first-compression, and no baseline from before October 2026. Without a baseline, any improvement attributed to the new model cannot actually be attributed to it.
This deserves a longer pause, because I have seen the same mistake repeatedly in another field. Based on my experience covering matches and standing in car parks after training, I know football teams love easy metrics too: kilometres run, pass accuracy, pressing counts. Those numbers are not wrong. They simply do not answer the question the stands are actually asking. Here it is the same. 97.2 seconds is a station-internal KPI, highly sensitive to how many staff were on duty that night, and not comparable across systems. It is being used as a proxy for quality while quality remains unmeasured.
All four metrics are self-reported by the 115 Center, with no independent audit. That does not make them false. It means they need verification before becoming a reference standard for other provinces.
The 84.2% figure is the most suggestive and the most opaque. It shows roughly one in six on-scene cases was resolved without transport — plausible for minor trauma — but without a case-mix breakdown it cannot be interpreted. A system that resolves many minor cases on scene is a good system. A system that overlooks severe cases on scene is a broken one. The same ratio, two opposite meanings.
From training ground to garage, I count the pulse of a collective through muddy boots.
If I had to pick one detail of this programme to praise, it would not be dispatch speed. It would be the priority list: drivers, guides, hotel staff, teachers, security.
The reason is concrete. The outcome of an out-of-hospital cardiac arrest depends mainly on who is standing next to the victim in the first four to six minutes, not on the island-wide average training rate. A taxi driver parked at a hotel gate, a guide leading a group on a beach, a receptionist in a lobby — these are the people who arrive before the ambulance, often before the call. Selecting them is selecting correctly by the spatial and temporal distribution of risk rather than by population distribution.
I am familiar with this kind of selection from football. A team does not defend better by training the entire academy to run backwards; it defends better by placing the right three people in the right three zones. The principle is identical. The difference is the cost of being wrong.
Which is exactly why the 1% target needs to be read properly. One percent of a population, spread evenly, cannot shift any island-level ratio. Mature community CPR programmes operate at 20% to 50% of the population trained. One percent cannot produce a measurable difference island-wide — unless it is concentrated on the priority list, and unless there is a refresher mechanism.
And here is where no programme document answers: compression skills decay within months without practice. With no refresher schedule, the 2,000 people trained in Q4 2026 become 2,000 certificate holders by mid-2027, and 2,000 people with degraded reflexes by 2028.
The 7.2 responses per day average conceals seasonality. Phu Quoc is a tourist destination, and real volume almost certainly clusters in high season and in weeks with major events. A system designed around an annual average will break in precisely the week it is needed most — and APEC 2027 is that week.
Behind the infrastructure sits a thin financial structure. Every capital item in this story — eight ambulances, 30 AEDs, the training programme — traces to a single source: Vingroup. No co-funders, no recurrent public operating line, no maintenance endowment. This is public-health investment by asset transfer: the corporate provides equipment, the state provides mandate and coordination. The pattern is legitimate and has worked elsewhere. It has one inherent weakness: it funds what is visible and stays silent about what recurs.
The recurring part never makes the photographs: AED electrode pads expire, batteries need cyclical replacement, ambulance consumables must be repurchased, staff must be paid. Ambulances have a five-to-ten-year lifecycle. Pads and batteries have a two-to-five-year lifecycle. No mechanism for the second purchase appears anywhere in the programme.
I once wrote about a club that received a large investment from a single sponsor and, three years later, sold a key player to pay the electricity bill. Structurally identical. The first investment is always the easiest.
What bothers me most about how this story will be told is that it will be told through 97.2 seconds.
That is the least informative of the four metrics. It measures something any adequately staffed station can do, and it does not correlate tightly with clinical outcome. What actually decides survival on an island is the interval before that vehicle rolls — four to six minutes that no ambulance can shorten. The entire value of the training programme lives inside that interval, and that interval appears in no table of figures.
The counterintuitive reading is this: an island that dispatches in under 100 seconds but has only 1% of residents trained in CPR may fare worse than an island that dispatches more slowly but has 20% trained. Speed is necessary, not sufficient. And speed is the cheapest condition to buy with money.
The 1% target looks small, and it is small. But if someone inside the programme can answer where that one percent is concentrated, it may still be the right decision. What is missing is evidence about allocation.
One more thing, stated plainly. A model that only runs when three specific conditions coincide — a conglomerate with a hospital subsidiary, a local government willing to chair it, and an international event to justify the spend — is hard to replicate. Con Dao, Ly Son and Cat Ba do not have that configuration. Those places will have to invent something else.
The first year of any emergency system is usually the easiest, because everything is new, attention is high, and no device has failed yet.
The second year will answer what the first could not. Whether any outcome metric is published — survival after out-of-hospital cardiac arrest, time to first compression, return of spontaneous circulation. Whether the 30 AEDs have a maintenance log, and who is accountable when a unit fails at the moment it is needed. Whether the first 2,000 trainees are retrained in 2027, or their certificates quietly expire.
I do not write about ambulances; I write about the people inside them, and the people who will stand beside a victim before the vehicle arrives.
In the garage at 5:42 a.m., nobody talks about APEC. They talk about the previous night's call, about the tyre that needs replacing, about the An Thoi station being one officer short. The pulse of an emergency system is not 97.2 seconds on a bulletin. It is whether, at three in the morning, the right person opens the garage door.
And whether, when that person arrives ten minutes after the ambulance should have, someone on the beach knows what to do.
