I thought it prudent to start the new ETM Journal Club with a local article (from my alma mater as it were), and from none other than the famed (ex) Red Rag, EMA.
We all know that bleeding is bad in trauma. Bleeding until you’re shocked is bad. If you’re bleeding and you’re shocked to the point where you need a blood transfusion, you are more likely to have more serious injuries, get sicker, and die as a result. The more blood you need to get you out of shock, the worse off you are. Wouldn’t it be great if we had a single, easy to calculate number that told us whether you were likely to need a massive transfusion? We could act on this from the pre-hospital notification, notify blood bank, get the MTP down and warm up the blood warmer. We may even detect your need for blood earlier in the ED resus phase. So welcome back Shock Index. Is this the magic number it’s being held up as? Read on to find out.
Reference:
What’s this article about?
[DDET Click for answer]
This article is about use of an easily calculated clinical sign, the Shock Index (SI), or heart rate divided by systolic blood pressure in predicting critical bleeding. Critical bleeding was defined as that which required at least 4 units of packed red blood cells or whole blood in 24 hours.
SI = HR / SBP
[/DDET]
Who did the study and can we trust them?
[DDET Click for answer]
This article has been written by well-known Australian researchers from Monash University, The Alfred Hospital (largest volume trauma centre in Australia) and Ambulance Victoria. Pretty reasonable credentials!
No conflicts of interest were identified.
[/DDET]
How was this study conducted?
[DDET Click for answer]
This is a systematic review article, which looked at 5 papers related to the use of the Shock Index in predicting critical bleeding. Interpreting advice from a review article can be difficult without pulling all of the individual references and analysing them.
In general, the higher the level of evidence (RCT > Cohort > Case Control > Case series) and homogeneity of the referenced articles, the higher the quality of the review, and the increased chance that meta-analysis of the studies will have meaning.
This is a review of mostly retrospective, heterogenous studies, therefore drawing conclusions from the data should be done with caution, and due to this, meta-analysis was not possible. This is acknowledged by the authors.
Heterogeneity issues in this paper include:
- Definitions of massive transfusion, shock index, and critical bleeding varied between studies
- The proportion of blunt vs penetrating trauma is not clear
[/DDET]
What did it find?
[DDET Click for answer]
In the reviewed studies, there was a positive correlation between degree of Shock Index, and critical bleeding/need for massive transfusion, however this varied from 3.4% to 39.6%.
A large number of limitations are identified in this paper, including:
Shock Index will be inaccurate in:
- Geriatric patients (reduced physiologic compensatory mechanisms)
- Pre-existing hypertension (baseline blood pressure will be higher, will have to drop more to raise SI)
- Athletes (lower resting heart rate, higher stroke volume/cardiac output)
- Pregnant (larger blood volume, faster resting heart rate)
- Illicit substance use (eg sympathomimetics – faster resting heart rate)
- Cardiopaths (beta-blockers/Ca2+ blockers, pacemakers, transplant)
- Spinal cord injury (neurogenic shock from sympathetic disruption)
- Hypotermia (causing bradycardia)
Eg:
Fit, healthy 24yo man, fall off horse onto right side, yelling in pain, RUQ tender ++, FAST positive, pink & well perfused (from all the yelling). HR = 100 BP = 120 SI = 0.83
75yo pedestrian vs car, amputated lower leg, you’re standing in a pool of blood that’s running off trolley. HR 110 SBP = 130 SI = 0.84 (On beta blocker, usual BP 180/110)
[/DDET]
Should this article change my practice?
[DDET Click for answer]
In short – No.
Whilst there are some interesting observations in this paper, the heterogeneity of the studies, the small number of papers reviewed (only 5 articles met inclusion criteria), their mostly retrospective nature, and the large number of variables and limitations that affect shock index means this “vital sign” should not be used to determine whether your patient has critical bleeding or will need massive transfusion, or at least it should be used with caution.
Until the defintions are tightened, and the end points clarified, (and potentially studied prospectively), I’d be cautious about using a single number to determine need for massive transfusion.
I think it falls into the “general interest” category, in that you knew your patient was bleeding (either externally or internally) from your excellent assessment, so you were already going to give blood instead of saline, and part of your decision included looking at their blood pressure and pulse, and their trends pre-hospital and while the patient was in the ED, and you weighed up other issues like haemostatic resuscitation principles (target SBP 90-100mmHg, regardless of pulse), time delay to surgery, co-morbidities, age, and blood results (e.g. haemoglobin which is usually available very rapidly from a blood gas).
i.e. “Look, my patient’s bleeding, I can’t stop it as quick as I’d like, they are clinically shocked so I better give them blood, and oh, by the way their SI is greater than 1.
It may come in handy as the authors suggest when comparing SI’s temporally, (eg from pre-hospital to ED, and during the ED resus. If the SI is going up over 1 or 1.1, especially if it rises despite fluid/blood administration, they have anywhere from a 3-40% chance of needing large volume transfusion, but hopefully you will have already realised that from other parameters, signs and detected injuries).
I’m also no good calculating percentages in my head. I’d have to get my calculator out on my phone, and that’s one more thing to remember to do in a busy resus!
[/DDET]
Posted in Uncategorized
