Cardiac arrest gives no warning — and most of the time it happens at home. The person you may one day need to resuscitate is far more likely to be a family member than a stranger on the street.
1. What is at stake. Once the heart stops, brain injury begins at 4 to 5 minutes — longer than the average ambulance response. A bystander's chest compressions raise survival two to three times.
2. Where Korea stands. Bystander CPR rose from 12.9% in 2014 to 30.3% in 2024 — 2.35 times. Still only three people in ten, against 33,034 cases a year, about 90.5 a day.
3. What to do. Six steps is the whole thing. Note that the choking guidance was updated — the order you were taught years ago is no longer it. And an AED talks you through the entire process out loud.
Why the first 4 minutes decide everything
When the heart stops, oxygen to the brain stops with it. Brain damage begins within about 4–5 minutes, and survival odds drop steeply with every minute after that. Ambulances almost never arrive that fast. The bystander — you — pressing on the chest is what doubles or triples survival.
Three in ten — and the difference those three make
Korea's disease control agency (KDCA) runs an annual out-of-hospital cardiac arrest survey. These are the 2024 figures.
| Measure | 2024 | Ten years earlier (2014) | Change |
|---|---|---|---|
| Cases | 33,034 (64.7 per 100,000) | — | 90.5 a day (our calculation) |
| Bystander CPR rate | 30.3% | 12.9% | +17.4pp (2.35×) |
| Survival rate | 9.2% (highest on record) | 4.8% | 1.92× |
| Brain-function recovery | 6.3% | 2.7% | 2.33× |
A 30.3% bystander rate is “three in ten.” Turn it around and seven times out of ten nobody touched the person before the ambulance arrived.
Do not read this as “CPR multiplies survival by 2.4.”
Cases where a bystander acted were more likely to happen where people were around, to be spotted sooner, and to be reported sooner. The published figures put two groups side by side; they do not claim causation.
What does stand is the decade: as the bystander rate climbed from 12.9% to 30.3%, survival climbed from 4.8% to 9.2%. Dispatch, transport and hospital care all improved over the same years, so this cannot be pinned on one cause either.
The demographics are not exotic. 31.8% of cases were people aged 80 or older, and the split was 21,237 men and 11,766 women. Checking the arithmetic, that sums to 33,003 — 31 short of the total, and the published summary does not say what those 31 are (sex not recorded, most likely).
The picture in the US is not very different. The Centers for Disease Control and Prevention writes that “more than 356,000 people have an out-of-hospital cardiac arrest in the United States every year” and that “about 60% to 80% of them die before reaching the hospital.” But the two countries report the same event with different measures — Korea publishes a survival rate, the US a pre-hospital death share. Different denominators and different moments; they cannot be lined up.
What this statistic actually counts
Before reading those numbers it helps to know what is being counted. The KDCA National Injury Information Portal sets out what it counts and how is now read from the agency.
| Item | What the agency page states |
|---|---|
| Statistic | Acute Cardiac Arrest Survey — approved national statistic no. 117088 |
| Scope | All patients with cardiac arrest occurring outside a hospital. A case qualifies when the 119 ambulance activity log records cardiac or respiratory arrest as the main symptom, or shows CPR having been performed, or has an emergency-treatment detail sheet filled in |
| Data sources | Activity logs written by 119 ambulance crews and the medical records of the hospital the patient was taken to |
| Purpose | To “provide baseline data for health policy and for evaluating the effect of investment in emergency medicine” |
| Publication | Half-yearly from the 2022 data onward; raw data released annually |
Knowing the data source changes how the number reads. This statistic counts events where 119 was called. An arrest an ambulance never reached does not enter it. Which is why it should not be read as “everyone in Korea whose heart stopped this year.”
2022 beside 2024 — fewer cases, better outcomes
The portal page carries the survey through 2022. Placed beside the 2024 figures used above:
| Measure | 2022 — agency source | 2024 — via press | Change (our calculation) |
|---|---|---|---|
| Cases | 35,018 “the highest since the survey began in 2006” | 33,034 | −1,984 (−5.7%) |
| Survival | 7.8% | 9.2% | +1.4 pts |
| Neurological recovery | 5.3% | 6.3% | +1.0 pts |
| Bystander CPR rate | 29.3% | 30.3% | +1.0 pts |
| Share of cardiac origin | 73.1% | — | Injury (falls, transport, etc.) about 20% |
2022 was the highest since the survey began, and two years later the count is below it — while survival, recovery and the bystander rate all rose. But 2022 is agency source text and 2024 is press coverage, so the two were not confirmed the same way. That is why the table labels the origin of each column.
The agency also defines “cardiac origin”: arrests caused by failure of the heart itself, plus those where no clear cause was established and the event falls under no specific disease category. Undetermined cases sit in the same bucket, so “73.1% cardiac origin” must not be reread as “73.1% heart attack.” The signals that can arrive before the heart stops are covered in heart attack warning signs.
CPR in 6 steps
| Step | Action | Key point |
|---|---|---|
| ① Check response | Tap the shoulders and shout “Are you okay?” | Tap, don't shake |
| ② Call for help | Point at someone: “You — call emergency services and bring an AED” | Alone? Put your phone on speaker with the dispatcher |
| ③ Check breathing | Look for chest movement for up to 10 seconds | Gasping, irregular breaths count as NOT breathing |
| ④ Chest compressions | Center of the chest (lower half of the breastbone), 100–120 per minute, about 5 cm (2 inches) deep | Arms straight, use your body weight, let the chest fully recoil |
| ⑤ Rescue breaths (if trained) | 30 compressions : 2 breaths, repeat | Never trained? Hands-only CPR is fine — just don't stop pressing |
| ⑥ When the AED arrives | Turn it on and follow the voice prompts | Place pads as pictured; hands off during analysis |
Key point. For untrained rescuers, hands-only CPR is the recommended standard. Hesitating because rescue breaths feel daunting costs lives; uninterrupted compressions save them. The right tempo matches the beat of “Stayin' Alive” or “Baby Shark.”
How it changes by age
The principle is identical, but hand position and depth scale with body size. In children, cardiac arrest is more often breathing-related, which makes rescue breaths more valuable if you know them.
| Age | Position & hands | Depth | Rate |
|---|---|---|---|
| Adults (post-puberty) | Center of chest, two hands interlocked | ~5 cm / 2 in (not over 6 cm) | 100–120/min |
| Children (1 to puberty) | Center of chest, one or two hands | 1/3 of chest depth (~5 cm) | 100–120/min |
| Infants (under 1) | Just below the nipple line, two fingers | 1/3 of chest depth (~4 cm) | 100–120/min |
AED: the machine talks you through it
Defibrillators hang in subway stations, gyms, offices, and shopping malls. Turn one on and a voice guides every step. The device analyzes the heart rhythm itself and will only deliver a shock if one is needed — you cannot hurt someone by attaching it. Keep compressions going while the pads go on.
Choking: the guidance was updated
If someone can't speak, can't cough, and clutches their throat, the airway is fully blocked. Under the 2025 American Heart Association guidelines, the recommended response is no longer abdominal thrusts alone — it's alternating 5 back blows with 5 abdominal thrusts.
| Who | Method | Caution |
|---|---|---|
| Adults & children (1+) | ① From behind, give 5 firm back blows between the shoulder blades with the heel of your hand → ② wrap your arms around, fist above the navel, and pull up and in 5 times (abdominal thrusts) → repeat ①② until the object comes out | If they can still cough, let them keep coughing — don't interfere |
| Infants (under 1) | Face-down along your forearm, head low: 5 back blows → flip over: 5 chest thrusts with two fingers just below the nipple line → repeat | Never do abdominal thrusts on an infant |
| Pregnant or larger-bodied | Thrust at the chest instead of the abdomen | - |
| If they pass out | Lower them to the floor and start CPR immediately; call emergency services | Only remove an object you can actually see — no blind finger sweeps |
A real-world example: holiday dinner, and grandpa stops talking
Mid-meal, he suddenly goes silent and grips his throat. ① Ask: “Are you choking? Can you speak?” No voice, no cough — that's a full obstruction. ② Send someone to call emergency services while you stand behind him alternating 5 back blows → 5 abdominal thrusts. ③ When the food pops out, let him rest — but if abdominal thrusts were used, he should be checked at a hospital for internal injury. ④ If he goes limp, lay him down and start chest compressions until help arrives. The whole scenario plays out in under three minutes. Rehearse it mentally once, so your body remembers.
Prepare before you ever need it
- ☐ Learn where the AEDs are in your building, office, and gym
- ☐ Take a CPR class — Red Cross chapters and fire departments run cheap or free sessions
- ☐ Elderly family at the table: cut rice cakes and meat small; swallow first, talk second
- ☐ Small children: no whole nuts, hard candy, or whole grapes (top choking causes under age 4)
- ☐ Review home emergency responses and stock a first-aid kit
- ☐ Traveling soon? Save the destination's emergency number — see our travel emergency guide
Questions that remain
I'm scared of breaking ribs if I push that hard.
Ribs sometimes do crack during CPR — and they heal. A stopped heart is far more fatal than a broken rib. Five centimeters is deeper than most people expect; lean in with your full body weight.
Could I be sued for doing it wrong?
Most countries — including Korea, the US, and much of Europe — have Good Samaritan laws protecting people who give emergency aid in good faith. Standing by and watching is the worst legal and moral position of all.
They're unconscious but seem to be breathing. Compress anyway?
If breathing is clearly normal, don't compress — roll them into the recovery position (on their side) and monitor until help arrives. But irregular gasping is a sign of cardiac arrest, not breathing: start compressions. Dispatchers can talk you through the call.
What if I'm alone and choking?
Call emergency services first — even if you can't speak, the call can be traced. Then press your upper abdomen forcefully against a chair back or table edge to perform abdominal thrusts on yourself.
CPR isn't an expert skill — it's a basic skill for anyone with a family. Tap, call, then push hard, fast, and without stopping. That alone can double or triple someone's chance of coming home.
Sources
- Korea Disease Control and Prevention Agency — National Injury Information Portal, “Acute Cardiac Arrest” (page updated 18 March 2024, 2022 survey basis). Source for 35,018 and “the highest since the survey began in 2006,” 7.8% survival, 5.3% neurological recovery, a 29.3% bystander rate, 73.1% of cardiac origin and its definition, about 20% from injury, and approval number 117088.
- Korea Disease Control and Prevention Agency — National Injury Information Portal, survey overview. Source for the case definition (all out-of-hospital patients; qualified via the 119 log's main symptom, a CPR record, or an emergency-treatment detail sheet), the data sources (119 activity logs and receiving-hospital records), the stated purpose, and half-yearly publication with annual raw-data release.
- Korea Disease Control and Prevention Agency, 2024 out-of-hospital cardiac arrest survey (announced December 2025). That announcement itself could not be opened on an agency page, so we verified it through press coverage — agency figures, but not read on an agency page. Source for 33,034 cases, 64.7 per 100,000, 9.2% survival, 6.3% brain-function recovery, a 30.3% bystander rate, the ten-year trend (4.8→9.2 / 2.7→6.3 / 12.9→30.3), the outcomes by bystander action (survival 14.4% vs 6.1%; recovery 11.4% vs 3.5%), and 31.8% aged 80+, 21,237 men, 11,766 women.
- US Centers for Disease Control and Prevention — “Cardiac Arrest” (checked July 2026). Source for “more than 356,000 people have an out-of-hospital cardiac arrest… every year,” “about 60% to 80% of them die before reaching the hospital,” the signs to look for (collapse, unresponsive, not breathing or gasping, no pulse), and the order “call 9-1-1 right away, look for an AED, and give CPR until medical professionals arrive.”
- US National Heart, Lung, and Blood Institute (NIH) — “Cardiac Arrest” (checked July 2026). Source for “calling 9-1-1 first is very important” and “AEDs are not hard to use, but training is very helpful.”
- Our own calculation. The 90.5 cases a day (33,034÷365), the decade multiples (2.35×, 1.92×, 2.33×) and the 31-case gap between the sex split and the total are values we computed, not figures in the announcement.
Where to check further
- Getting compression depth and rate into your hands. The figures here follow widely used international guidance and were not matched to an agency page. More to the point, reading about compressions is not the same as doing them — a free session with the Red Cross or your local fire service is worth far more than this page.
- Year-by-year survival and neurological recovery figures. The National Injury Information Portal publishes that series as a chart image, so the individual years could not be read as text; only values the page states in words appear here. The KDCA's annual acute cardiac arrest survey releases carry the year you want.
- Causation between bystander CPR rates and survival. The published material places two groups' outcomes side by side and no more, and this article does not assert causation. Korea and the US are likewise not compared directly, because they publish different measures (survival rate vs pre-hospital death share).
Updated August 2026. The Korean statistics come from the KDCA National Injury Information Portal (case definition, data sources, 2022 figures) and the 2024 survey announcement ( via press), the nature of cardiac arrest and the response order from CDC and NHLBI source text, and the multiples and the daily average are ours. This is general information and does not replace hands-on training. Take one practical class — your hands will remember what your eyes read here. For when symptoms need an emergency room see ER or clinic?, and for what to keep at home see the home first-aid kit.
