IO, die-O

PotatoMedic

Has no idea what I'm doing.
2,703
1,541
113

rescue1

Forum Asst. Chief
587
136
43
The discussion section (I have a subscription, I don't know if everyone else can see it?) says that one of their suggested reasons is the unknown effectiveness of drugs given IO during arrest (how much of it just hangs out in the marrow). However, since there's a lot of data saying that epi in arrest may be harmful, I'm not sure why you'd see such a big difference since 85% percent of the matched group had a non shockable rhythm and weren't getting any other meds.

I guess it's weird in general because all the "new" studies are saying "BLS saves lives!", so why would drug therapy cause such a big effect?

The plus side of this is that now my agency's policy of going for an IV before an IO seems evidenced based and smart, instead of based on saving money.
 

ATFDFF

Forum Crew Member
41
2
8
IMHO:

Short version: correlation does not prove causation

Long-ish version: While we all know that no published article "proves" anything, and there are limitations to any pieces of literature (particularly retrospective data analyses) this seems to have some glaring holes which make me severely doubt its conclusions. One particular line which stuck out to me (which is not even addressed in their "limitations" section) reads: "Patients with intraosseous access had a higher proportion of nonshockable initial rhythms, fewer public location and witnessed arrests, and shorter times from call to first paramedic and ALS paramedics arrival." Of the items listed, 75% have been shown to correlate with poorer outcomes for patients. When one does read through their six-items limitations section....I feel they far too quickly dismiss several complicating variables.

Anecdotally, patients on whom I (and I'd venture most paramedics) initiate an IO tend to have physical abnormalities which preclude or make traditional venous access difficult. Most commonly these are patients with large body habitus, poor vascular compliance due to prior drug abuse, etc. however they are all associated with a myriad of health complications at baseline....much less in a post-arrest recovery.

I would hope that your medical directors do not use *only* this to make any concrete changes to policy or practice.
 

rescue1

Forum Asst. Chief
587
136
43
They did use a matched dataset to try and control for those variables you mentioned, though apparently with no change in the results.

The patient population is a good point. At least where I worked, patients with an IO were many times more likely to have poor access or a failed IV attempt.
 

VFlutter

Flight Nurse
3,728
1,264
113
Looks like IO group had 86% (vs 73% IV) Non-shockable rhythms with other variables being somewhat similar.

From the article "the type of vascular access was not randomly selected in our data; paramedics may have selected patients with certain characteristics for intraosseous access: although we excluded those with unsuccessful attempts at either route and applied propensity score-matching techniques to adjust for this selection bias, residual confounders may have affected our estimates"

As others have stated it is hard to do a retrospective study with IO vs IV since many times those who get IOs tend to be non-shockable rhythms, unwitnessed, obese, or otherwise unhealthy.

I will continue to place IOs unless there are obvious adequate veins. Go Humeral and make sure you use a pressure bag. Get a large bore PIV when able.
 

Summit

Critical Crazy
2,691
1,312
113
I agree with the well written posts above. I'd had similar thoughts when I saw this study recently.

The data say ACLS drugs probably don't do much of anything most of the time... therefore, the PIV/IO outcomes should be equal unless there is a lurking variable or selection bias. It is hard to imagine a lurking dependent variable, but potential selection bias is easily identified and imagined with relation to patient acuity and other known prognostic factors, as mentioned above.

The only lurking dependent variable I can imagine is that if intraarrest Epi is bad, (and I see why it would be the way we give it in arrest), then IO could magnify the deleterious effect of the epi by delivering it more rapidly and thus in higher serum concentration (vs PIV), particularly from the humeral site... but the data don't suggest this... and if they did, the solution isn't to avoid IO's.
 

NPO

Forum Deputy Chief
1,831
897
113
Why did the IO group enroll 660 patients, but the IV group had nearly 13,000? That makes me think that the IO data came from one agency/location or something.
 

VFlutter

Flight Nurse
3,728
1,264
113
Why did the IO group enroll 660 patients, but the IV group had nearly 13,000? That makes me think that the IO data came from one agency/location or something.

Looks like they only included patients whom had IOs as primary access and excluded those got IOs after failed IV attempts. As mentioned the former group probably being "sicker" patients and the later group that was excluded being the more common situation.
 

MedicMcGoo

Forum Ride Along
9
6
3
Two things that question my faith in the validity of the study:

1. The data used in 10 years old.
2. The analyzed population who received an IO is a fraction of the IV population analyzed.
3. The study said "Patients with intraosseous access had a higher proportion of nonshockable initial rhythms, fewer public location and witnessed arrests, and shorter times from call to first paramedic and ALS paramedics arrival."

As stated in the prior posts, the need for further investigation citing other studies with a larger IO population and more current data would be suggested prior to making a treatment decision changes. At the end of the day, they're the medical director, who was hired to make these judgment calls. It's important however to question and critically think about the decisions that are being made.
 
Last edited:
Top