You read that right. San Francisco has a cardiac arrest survival rate of 100%.
Does that mean that everyone who suffered a sudden cardiac arrest survived?
Of course not.
But our survival rate is still 100%.
That's because today my numbers for witnessed Asystole with ROSC and cooling measures look really good. So, our rate is officially 100%.
Hogwash you say? How is that any different than some communities who bend their inclusion criteria to give the impression that they have an amazing rate of survival when their overall numbers are a complete unknown.
I rub my temples when editing our CARES registry, not because the data fields seem unending, but because there is one that I'm not sure we should be editing until long after the event:
"Suspected Cardiac."
This term sits in a line with others such as drowning, respiratory, trauma etc. Since our cardiac arrest patients seldom tell us what led to their arrest, we have to make a HUGE assumption and hope the hospitals update the information accurately. That is also assuming that the hospital was able to discover the cause of the arrest.
It would be refreshing to be able to determine the cause of cardiac arrest and only count the patients we were going to be able to help at all, but that should not be taken into account when reporting survivability.
Imagine it this way: Our major trauma survival rate is 99% because we don't include patients who had a BP of less than 50/P in the field since our efforts are unlikely to impact them.
Insanity, right? If trauma care and survivability from injuries is our population, why exclude those who are less likely to make it?
Because it makes us look bad.
We can argue the semantics of statistics, reporting and the like for decades, we have and we will, but make sure when someone reports 60% survivability from a condition that kills more than 90% of those who suffer it, ask what they mean by "survival" "cardiac arrest" and "witnessed."
It's all in who you place in the denominator that decides your final answer.
So when I neglect to tell you our survivability only includes tourists with witnessed arrest who received bystander CPR and an ALS intervention within 4 minutes, am I really telling you the chance of you surviving cardiac arrest in my community?
Nope, but it sure is fun to say.
What is our actual rate?
We don't know, some of our 2012 patients are still in the hospital. Some communities might count them as survivors and move on, but we're looking for total survivability, not simply a target some time after admission. It will also help us guide future interventions if we can wait to learn exactly what happened to cause the arrest in the first place.
Sub point being that putting all your efforts into community CPR may seem like a fantastic idea, but if none of your survivors received it, will it make a difference in your community?
None of your survivors received cooling measures...does it still have an impact in your community?
No one who had a transport time of less than 5 minutes survived...should we slow our response?
Asking questions like this comes from looking at the data with too fine a comb.
1/2 of 1% (.5%) of the patients seen by my system are in cardiac arrest. Of that group upwards of 90% do not survive, many of them being beyond our help before we arrive. So now look at the subset of the population we're observing. 10% of .5% means .05%.
That's 50 cents out of a $100 bill.
Are you willing to change your system, your community, your children's graduation requirements... for .05% of your patients?
While I'm a big fan of community CPR, I'm also a big fan of community asthma programs, community hypertension screenings and community programs to reduce unnecessary 911 calls.
Those programs impact a far greater population and while there are not immediate results, like in ROSC, we are preventing far more cardiac arrests 30-50 years from now.
Which is better: Preventing an arrest, or classifying one so as to show success?
Hard to prove a negative.
EDIT - I was contacted by someone I respect very highly who informed me my stance has been misinterpreted (ie I didn't get my thoughts out the way I thought I had. Not the first time either, go figure). I am not calling for the muddying of the waters when it comes to Cardiac Arrest Survival rates, simply that agencies know what they are reporting and, more importantly, WHY! I'm aware I come off as a non-believer in this post, implying that all the new fangled stuff doesn't work, I just want to make absolutely sure we don't all focus on one outcome at the possible expense of others. My views may be confrontative, ill-advised and downright wrong, I've never claimed to have the answers, but in my current position of gathering and interpreting data I see how easy it is to steer the results in my favor. It's tempting when someone survives cardiac arrest and goes home mere days later...yet doesn't fit the Utstein numbers, so doesn't "count." It's also frustrating when we have an Utstein case arrest in the ambulance and not survive. The one person we think we have the best shot with and there's little we can do or learn from it. Agencies like Medic One and Hilton Head Island, with impressive Ustein results should be admired. They're using their data, learning from it and applying it. Good for their patients (and good for them) but if you are not part of the CAREs registry make damn sure you know your metrics and are reporting apples to apples, otherwise you're not only fooling yourselves, but doing your patients a disservice.
-HM
Does that mean that everyone who suffered a sudden cardiac arrest survived?
Of course not.
But our survival rate is still 100%.
That's because today my numbers for witnessed Asystole with ROSC and cooling measures look really good. So, our rate is officially 100%.
Hogwash you say? How is that any different than some communities who bend their inclusion criteria to give the impression that they have an amazing rate of survival when their overall numbers are a complete unknown.
I rub my temples when editing our CARES registry, not because the data fields seem unending, but because there is one that I'm not sure we should be editing until long after the event:
"Suspected Cardiac."
This term sits in a line with others such as drowning, respiratory, trauma etc. Since our cardiac arrest patients seldom tell us what led to their arrest, we have to make a HUGE assumption and hope the hospitals update the information accurately. That is also assuming that the hospital was able to discover the cause of the arrest.
It would be refreshing to be able to determine the cause of cardiac arrest and only count the patients we were going to be able to help at all, but that should not be taken into account when reporting survivability.
Imagine it this way: Our major trauma survival rate is 99% because we don't include patients who had a BP of less than 50/P in the field since our efforts are unlikely to impact them.
Insanity, right? If trauma care and survivability from injuries is our population, why exclude those who are less likely to make it?
Because it makes us look bad.
We can argue the semantics of statistics, reporting and the like for decades, we have and we will, but make sure when someone reports 60% survivability from a condition that kills more than 90% of those who suffer it, ask what they mean by "survival" "cardiac arrest" and "witnessed."
It's all in who you place in the denominator that decides your final answer.
So when I neglect to tell you our survivability only includes tourists with witnessed arrest who received bystander CPR and an ALS intervention within 4 minutes, am I really telling you the chance of you surviving cardiac arrest in my community?
Nope, but it sure is fun to say.
What is our actual rate?
We don't know, some of our 2012 patients are still in the hospital. Some communities might count them as survivors and move on, but we're looking for total survivability, not simply a target some time after admission. It will also help us guide future interventions if we can wait to learn exactly what happened to cause the arrest in the first place.
Sub point being that putting all your efforts into community CPR may seem like a fantastic idea, but if none of your survivors received it, will it make a difference in your community?
None of your survivors received cooling measures...does it still have an impact in your community?
No one who had a transport time of less than 5 minutes survived...should we slow our response?
Asking questions like this comes from looking at the data with too fine a comb.
1/2 of 1% (.5%) of the patients seen by my system are in cardiac arrest. Of that group upwards of 90% do not survive, many of them being beyond our help before we arrive. So now look at the subset of the population we're observing. 10% of .5% means .05%.
That's 50 cents out of a $100 bill.
Are you willing to change your system, your community, your children's graduation requirements... for .05% of your patients?
While I'm a big fan of community CPR, I'm also a big fan of community asthma programs, community hypertension screenings and community programs to reduce unnecessary 911 calls.
Those programs impact a far greater population and while there are not immediate results, like in ROSC, we are preventing far more cardiac arrests 30-50 years from now.
Which is better: Preventing an arrest, or classifying one so as to show success?
Hard to prove a negative.
EDIT - I was contacted by someone I respect very highly who informed me my stance has been misinterpreted (ie I didn't get my thoughts out the way I thought I had. Not the first time either, go figure). I am not calling for the muddying of the waters when it comes to Cardiac Arrest Survival rates, simply that agencies know what they are reporting and, more importantly, WHY! I'm aware I come off as a non-believer in this post, implying that all the new fangled stuff doesn't work, I just want to make absolutely sure we don't all focus on one outcome at the possible expense of others. My views may be confrontative, ill-advised and downright wrong, I've never claimed to have the answers, but in my current position of gathering and interpreting data I see how easy it is to steer the results in my favor. It's tempting when someone survives cardiac arrest and goes home mere days later...yet doesn't fit the Utstein numbers, so doesn't "count." It's also frustrating when we have an Utstein case arrest in the ambulance and not survive. The one person we think we have the best shot with and there's little we can do or learn from it. Agencies like Medic One and Hilton Head Island, with impressive Ustein results should be admired. They're using their data, learning from it and applying it. Good for their patients (and good for them) but if you are not part of the CAREs registry make damn sure you know your metrics and are reporting apples to apples, otherwise you're not only fooling yourselves, but doing your patients a disservice.
-HM
Comments