This study sought to compare ED doctor’s gestalt vs screening tools to identify sepsis. The tools included SIRS, qSOFA, SOFA, MEWS, and a machine learning one called LASSO.
​
We would intuitively believe that these screening tools are an improvement over gestalt. Why would we use them if they weren't better? Oddly enough, the screening tools have never been compared to gestalt. What if they are worse?

This study took place in a single well-functioning ED in the USA with over 100,000 patient visits per year. They included undifferentiated adults who were critically ill and went to the resuscitation area of their ED. They excluded trauma, obvious cause of illness, STEMI, strokes, and transfers.

At 15 and 60 minutes after presentation to the ED, a trained independent researcher asked the ED consultant a single question; “What is the likelihood that this patient has sepsis.” They rated their answer on an iPad displaying a slide bar from 0 (no infection) to 100 (infection).

The gold standard diagnosis of sepsis was determined by an ICD-10 code at hospital discharge. Gestalt and screening tools performance were appropriately compared by looking at using area under the curve (AUC). 

Results?
2,484 patients met eligibility. 275 (11%) had sepsis.

Which was better? Gestalt or screening tools?

Gestalt was crazy good with an AUC of 0.9. This is really exceptional. Other tools performed worse. Machine learning AUC 0.84 (pretty good), MEWS 0.72 (good), and qSOFA, SIRS and SOFA about 0.67 (not great.) Don’t forget, an AUC of 0.5 is considered random… like flipping a coin. 
Although I agree with the overall message of this paper, there are some important limitations. I’ll only mention a few. 

They only included very sick patients that went to resuscitation. Sepsis should be more obvious in this cohort. I’m guessing this is why gestalt performed so well. What I am really interested in are the subtle cases that may still be in the waiting room. One could argue that they studied the wrong patient population. 

This was a single centre that used the gestalt of a highly experienced consultant. In addition, they were very quick and efficient at getting diagnostic tests back including x-ray, bloods, vital signs, and POCUS. 

The paper relied on a hospital discharge diagnosis of sepsis as the gold standard. It is very likely that patients with medical conditions who presented to the ED without sepsis subsequently developed the condition while in hospital. This would bias the results against gestalt. 

What is my take home message?
​
Before we rely upon screening tools or clinical decision instruments, we must know that they are better than what we are currently doing. As demonstrated in this limited study, tools may perform worse!
 
Covering: Knack SK, Scott N, Driver B, et al. Early Physician Gestalt vs Usual Screening Tools for the Prediction of Sepsis in Critically Ill Emergency Patients. Ann Emerg Med;84:246-258. [Link to article]

These authors from the UK sought to answer this question by conducting a prospective multicentre cohort study of consecutive patients presenting to the ED with acute headache. 

Investigation, diagnosis and management were conducted using their standard practice. All patients were followed up for 28 days to see if they missed any haemorrhages. 

3663 patients were enrolled. About 90% got a CT and one third got a lumbar puncture after negative CT. Prevalence of SAH was 6.5% (n=237). 

What were the results?
A CT withing 6 hours of headache onset was 97% sensitive with a negative predictive value of 99%. CT within 24 hours had a sensitivity of 94% and a post-test probability of disease of less than 1%. The authors believe this data will inform clinicians and patients about the need (or not) for further investigation after a negative CT. 

Although I would very much like to believe these conclusions, there is one major concern with the methodology. 

Incorporation bias.

The head CT, the test that they are trying to determine was good or not, formed a crucial part of the gold standard. An alternative title to the paper could have been, “We did a head CT to excluded SAH, it was negative, and we believed it.”

To be fair to the authors, it would be unfeasible and not ethical to have performed angiograms and lumbar punctures on all patients to exclude SAH. And their 28 day follow up was probably a reasonable surrogate to find missed SAH’s. 

What should we conclude?
​
This study provides some evidence that we might be able to push the time window of CT a bit further. But I doubt the quality of this study will be enough to change guidelines. Unfortunately, there still is no right answer, and we will continue to engage in shared decision-making discussions with our patients. 
 
Covering:
Trainee Emergency Research Network (TERN). Subarachnoid haemorrhage in the emergency department (SHED): a prospective, observational, multicentre cohort study. Emerg Med J. 2024 Oct 4:emermed-2024-214068. doi: 10.1136/emermed-2024-214068. [Link to article]

The Oakland score was derived in 2015 but it somehow slipped my attention. The score is quite simple and mostly includes at a combination of age, vital signs and hemoglobin level to risk stratify patients [Oakland score MdCalc link]. ​

This new external validation study was a multicenter effort that was conducted in 140 US hospitals. It identified 38,067 patients admitted to the hospital with a primary diagnosis of lower GI bleed by ICD-10 codes. They retrospectively applied the Oakland score to these inpatients to see if this decision instrument could identify a group that was safe for discharge. Test characteristics of the score were reported at several cut-offs (8,9, and 10 points)

The primary outcome to determine “safe discharge” was a composite of an absence of the following: in hospital rebleeding, RBC transfusion, therapeutic colonoscopy, mesenteric embolization, laparotomy, in hospital death and readmission with subsequent lower GI bleeding 28 days. 

Results?
Average age was 70 years old and 68% of patients received a blood transfusion. Yikes! These were not low risk patients… probably why they were admitted in the first place.

Applying the rule to this inpatient cohort identified a measly 8.7% of patients considered safe for discharge at the 8-point threshold with a sensitivity of 98.4% and 16% specificity. If the threshold score was increased to 10, identified 17.8% with similar sensitivity but increased specificity to 32%. 

Overall, it’s a pretty bad decision instrument in this population as it is not discriminatory. It pretty much calls everyone positive. 

However, I think there is a bigger issue with this study. 

It was conducted on inpatients. The decision to admit was already made!

For overall utility, it should have been done in the ED. Nevertheless, the authors did state, “a prospective cohort study is needed, in which all patients presenting to the emergency department are included, regardless of their admission status.”

For what it is worth, I might see myself documenting a low-risk Oakland score to support my decision to send home a low-risk patient with lower GI bleed home. But in its current form, I fear the lack of specificity may result in more patients being admitted to hospital if we apply it without thinking.

Ultimately, yes, I would really like to see this study prospectively validated in an ED population and compared to gestalt. Anyone interested in a research project?
​
Covering:
Oakland K, Kothiwale S, Forehand T, Jackson E, Bucknall C, Sey MSL, Singh S, Jairath V, Perlin J. External Validation of the Oakland Score to Assess Safe Hospital Discharge Among Adult Patients With Acute Lower Gastrointestinal Bleeding in the US. JAMA Netw Open. 2020 Jul 1;3(7):e209630. doi: 10.1001/jamanetworkopen.2020.9630. PMID: 32633766; PMCID: PMC7341175. [link to full text article]

Have you ever ordered an ultrasound on a patient and had the report come back stating “appendix not visualised”? This seems to be a common and frustrating occurrence. We have historically considered this to be a result that doesn’t change our thought process, and we throw the report in the bin.
But what about the indirect signs of appendicitis that may be seen on ultrasound? If the appendix is inflamed, often so are the surrounding structures. Can the presence or absence of indirect signs change how we think about the post-test probability of disease?

Researchers from a single centre in beautiful Valladolid, Spain conducted a retrospective study of paediatric patients aged 3-13 who were thought possibly to have appendicitis and had a diagnostic workup.

There were 1,756 patient encounters. They used the Alvarado score to risk stratify the kids into low, intermediate or high risk. The positive appendicitis rate was ultimately found to be 18%.
A total of 60% of children had an ultrasound performed. About half the time the appendix was not visualised. In this cohort of patients, they went further to look at the ultrasound report for possible indirect signs of appendicitis. 

Indirect ultrasound findings independently associated with appendicitis were a small amount of free fluid (OR 5), abundant free fluid (OR 30), and inflammation of the periappendiceal fat (OR 7).
The authors conclude that the absence of these indirect signs ruled out appendicitis in patients with low or intermediate suspicion with a sensitivity of 85% and a negative predictive value of 99% even when the appendix was not visualised. The missed cases tended to be patients that presented early after onset of pain as they likely had less inflammation present. 

Some might argue that 85% sensitivity is not good enough. But appendicitis is not aortic dissection. I think a lower sensitivity is acceptable so that we are not causing harm by over investigation and admission.  A short delay to diagnosis of appendicitis is not likely to be harmful especially in early presenters that don’t have ultrasound signs of inflammation. Obviously, we must give good discharge instructions and return precautions for those that we send home.  

To be complete, there were a lot of limitations to this retrospective study. But there are other studies out there with similar results and conclusions. 

Next time you get an ultrasound report of “appendix not visualised” make sure you read further. Indirect signs might be your friend. 
​
Covering:
Pernía J, Cancho T, Segovia I, de Ponga P, Granda E, Velasco R. Predictive values of indirect ultrasound signs for low risk of acute appendicitis in paediatric patients without visualisation of the appendix on ultrasound. Emerg Med J. 2024 Jul 22;41(8):475-480. doi: 10.1136/emermed-2023-213466. PMID: 38729752. [link to article]

Nevertheless, these researchers from Australia put the answer in black and white with this retrospective cohort study. Looking at the metric of ED length of stay (LOS), they proved that haematoma block was faster in their two ED’s. The mean ED LOS was 187 minutes for haematoma block, 227 minutes for Bier’s, and 239 minutes for procedural sedation. 

I suspect that other ED’s might have more impressive differences if they have trouble accessing crowded resuscitation rooms and qualified staff that are busy with other tasks.

On the flip side, procedural sedation had a higher rate of successful first attempt reductionat 94% vs. 76% for haematoma block. But groups were not randomised and reductions under procedural sedation were twice as likely to have senior staff present and an orthopaedic registrar.

They also mention the cost of consumables and opportunity cost were less with haematoma block. No surprise here.

More patients had complications with procedural sedation 18%, Bier’s 14% and 13% with haematoma block. All but one of the complications of haematoma block were due to “block failure.” I don’t believe this to be a complication but rather an expected outcome. Sometimes haematoma blocks simply don’t work and one needs to proceed to plan B. No big deal. 

In their conclusion, the authors state haematoma block “should be the default anaesthetic technique” for closed reduction of closed forearm fracture reductions. Obviously, the choice is multifactorial and comes down to patient and local institutional factors. But generally speaking, I agree. 
​
Covering:
Pitman GR, Soeyland T, Popovic G, Thomson D. Haematoma block is the most efficient technique for closed forearm fracture reduction: a retrospective cohort study. Emerg Med J. 2024 Sep 25;41(10):595-601. doi: 10.1136/emermed-2023-213591. PMID: 39019579. [link to article]

Recent covid 19 RAT at home was negative.
The child had subjected fever with lethargy last night.
Usually fit and well with no significant past medical history.
 
Normal work of breathing with normal oxygen saturation.
You've ordered an CXR and proceed to POCUS while waiting:

So you think long and hard about all these grey images...

​Now you wonder if you really need CXR to make clinical decision if you have ability to do Lung POCUS….

You've loaded him with a therapeutic dose of clexane, calculate the wells score in your head and add on 100 points for the POCUS findings and organise a CTPA:

By the time it gets done, youve also got some of hte bloods back: 

• Troponin 121 ng/L (normal <15)
  
• NT BNP 840ng/L (normal<125)

PW Doppler

Blood Gas Home  |  Blood Gas 101  |  Previous  |  Next

Joe is a 28 year old male.
He was well yesterday with some 'sore muscles' that evolved into feeling 'generally unwell' overnight, not able to sleep with nausea, some vomiting and nondescript abdominal pain.
After not being able to keep anything down today, and being too weak to mobilise on his own to the bathroom, his partner called an ambulance. 

His past history includes morbid obesity and type 2 diabetes mellitus.
He's on ozempic and dapagliflozin.

He's seen by triage and assigned a CAT 2 and moved to resus. He looks dreadful. 
Heart Rate 120, Blood Pressure 110/70, T37.2, Sat 97%RA, RR42

Appears very dry, with a soft but generally tender abdomen and no other significant findings

A blood gas is done (below) and ketones were 8.0 mmol/L

Guidelines and local practice vary. When you see this variation, it invariably means there is not conclusive evidence to inform practice.

Perhaps this RCT from the New England Journal of Medicine might help?

These researchers from the USA conducted a pragmatic, multicentre, randomized non-inferiority trial. They included patients who had a fracture of an extremity (hip to midfoot or shoulder to wrist) that were treated operatively or who had pelvic or acetabular fracture.

Patients were randomized to aspirin 81mg BD vs. enoxaparin 30mg BD while in hospital but could get whatever after discharge “according to the clinical protocols of each hospital.” (For what it’s worth, BD aspirin is not a typo…)
Primary outcome was death from any cause and secondary outcomes looked at PE’s, DVT’s and bleeding complications.

Results?
12,211 patients were included! Mean age 44 years old and 87% had lower extremity fractures. Median BMI was 27. On average, patients received about 9 in-hospital doses of thromboprophylaxis and were prescribed a 21-day supply of whatever upon discharge.

Primary outcome… no difference at 0.7% mortality in both groups. There was statistically more DVT’s in the aspirin group by a whopping 0.8% margin (NNT 125). PE’s were no different and nor were complications.
What should we conclude in the ED?

Yes, this is the tricky. The patients included in this study were not patients discharged from the ED. Extrapolating the findings from this RCT to our population is problematic. Bummer…

In addition, I’m disappointed they did not standardize the take home thromboprophylaxis. This likely made it much more difficult to find differences between the groups as patients were likely to cross-over.

Further, this trial did not restrict enrollment to high-risk patients. These are precisely the patients I want to target… another bummer. It is very possible that enrolling all-comers diluted down the benefits & differences of therapy. Who knows… but it’s common sense.
​
In the end, I don’t think this should substantially change how we think about ED patients that we send home with fractures. Nevertheless, I anticipate this study will influence the guidelines… we might see some more aspirin being recommended. But I'm not sure if this is for the better or worse.
 
Covering:
METRC Consortium, Aspirin or Low-Molecular-Weight Heparin for Thromboprophylaxis after a Fracture. N Engl J Med 2023;338:203-13. [Link to article]