WARNING, Do NOT read further if you are sensitive about the realities of death and hospital services.
Why? Because doctors know that CPR only works in very rare cases, about 10% one-month survival with good neurologic outcome.
Unlike the TV shows where everyone rushes to a patient and gives CPR, an injection, and perhaps a big defibrillator shock, then leaves the hospital 35 minutes later in a wheelchair and hops into a car for a ride home, in actuality if you are sick enough to be in a hospital and stop breathing or your heart stops beating, or you collapse at a game, at home, or a concert, and need to go to a hospital, then there is about a 10% chance you will leave that hospital alive if you needed CPR.
There are a few instances when it can be useful, especially with younger adults and especially kids who tend to recover from illnesses and injuries much better than older adults. This should be mandatory in cases of apparent drowning or electric shock.
Most CPR training no longer emphasizes anything but chest compression for that very reason - no use coming in close contact with a potentially very ill person when it won't do any good.
But consider that when done correctly CPR often results in broken ribs.
So, if you are over 50, or have a bad heart, and especially if you have other medical conditions and are even older, there is almost no chance that you will ever walk out of a hospital if you had to receive CPR even in a hospital setting and done correctly.
But, what about a hospital you ask, don't they also give a shot of adrenalin with the miraculous effect recently seen in the Sherlock Holmes movie? Well, truth be told, and you seldom hear much truth when it comes to death, a recent study showed that giving adrenalin shots in addition to CPR and shock increases long term survival (meaning you ever leave the hospital on your own) by a whopping 1.0% or one in one hundred. https://www.sciencedaily.com/releases/2018/07/180720154915.htm
On the other hand, it nearly doubles the chance of living on with severe brain damage.
What will happen is that as you are dying you will have a stranger beating on your chest with clenched fists (in a vain attempt to start your already dead heart) and/or having large heavy people essentially jumping up and down on your chest.
How would you prefer to go? I don't want to go at all, and you probably don't either, but given a choice, I prefer peaceful to getting a beating severe enough to break ribs.
Why is the CDC and Red Cross and even civic organizations still promoting CPR? Because it makes people feel good to do something. But beyond what it does to the patient, consider how the person giving CPR will feel when they realize they have been beating on a dying person?
"two randomized trials comparing the short-term survival of CPR using chest compression alone or chest compression plus rescue breathing. Of 2 496 cases of adult OHCA in that study, 1 243 (50%) were randomly assigned to chest compression alone and 1 253 (50%) to chest compression plus rescue breathing. Follow-up revealed that in these 2 496 cases there were 2 260 deaths and 236 long-term survivors, and that chest compression alone was associated with a lower risk of death in comparison with chest compression plus rescue breathing (adjusted HR 0.91, 95%CI 0.83–0.99)."
Read that again, only 236 survivors out of 2496 CPR cases and that was a random sample so some people were young and healthy before some accident or injury such as a car crash. More people recover for a few hours or even a few days if recovery is measured as being in a hospital ICU bed.
Read what doctors choose for themselves and loved ones.
"A study...by the Stanford University School of Medicine, found most physicians surveyed would choose a do-not-resuscitate or “no-code” status for themselves if they were terminally ill even though they tend to pursue aggressive, life-prolonging treatment for patients facing the same prognosis."
How Doctors Die
"Do Unto Others: Doctors' Personal End-of-Life Resuscitation Preferences and Their Attitudes toward Advance Directives" https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0098246