January 2026

Clincial question

When we're not sure were the infection is should we be getting a CT?

Title

Suspected infection of unclear origin at the emergency department: diagnostic yield of thoraco abdominal-pelvic CT in non-severe patients

Author

Marine Dumon et al in the BJR 2025

Background

Infections are very common causes of presentation to EDs. Sometimes it's very obvious and a CXR or a urine sample or the obvious cellulitis tells the tale. But it is not unusual to have someone where the answer is not entirely clear or to lay a little too much weight on a tiny bit of fluff on the CXR or the really dodgy urine sample and call the diagnostic search quits.

This approach can cause serious issues when the infected gall bladder perforates 12 hrs later on the medical ward or the obstructed infected kidney stone is left unmanaged

The NICE guidance for several years now has had a statement suggesting that in sepsis that if the source is not immediately apparent then a CT scan of the belly should be considered. These guys want to look at the yield of the CT scan in suspected infection but in what they call the non severe cohort

Methods

• This is a chart refiew in a single centre in France.

• They looked at patients who had CT. This causes immediate methodology problems as we have no idea what happened to those who didn't get CTs

• They say that they know infection was suspected because someone had written it on the request. And of course we all know that we always write entirely honest and accurate information on our CT orders... wink wink...

• They did look at some SIRS criteria and labs to identify the "non severe" component and effectively enroll people without sepsis criteria

• They did categorise the CT scans in a blinded fashion with the reading radiologist not knowing final diagnosis or if they had a surgery etc...

• They did not do interreader correlation which is a weakness

Results

• 500 pts

• 310 infections found (55% of scans)

• mostly pneumonia, followed by belly, followed by urological

• 20% of scans had other major findings of clots or cancer issues

Thoughts

The CT scan here is pretty high yield. You could make an argument that the CXR may get you the same answer most of the time but you'd still be missing a lot and presumably the clinicians were only ordering the CT here because clinical exam and CXR hadn't nailed the diagnosis for them.

• And this is putatively in the non sepsis population.

• I am a big fan of the CT scan as you might be able to tell, especially for the sickies as you will end up relabelling a lot of your pneumonias and UTIs as surgical disease.

• We have excellent access to CT where we work typically and we have massively increased the amount of CTs in the 20 years i've been working but it's hard to ignore such a great test.

Clinical Question

Does bicarb help in metabolic acidosis?

Title

Sodium Bicarbonate for Severe Metabolic Acidemia and Acute Kidney Injury The BICARICU-2 Randomized Clinical Trial

Author

Boris Jung et al 2025 JAMA

Background

An acidaemia on a blood gas is eye catching the emergency department. Typically, if it wasn't already obvious, you have just identified a sick patient and the resus monkey in you should hopefully kick in.

The old adage of "treat the cause" is of course very important but does simple giving an alkaline fluid to an acidotic patient to fix the numbers actually helpful?

The same group did this before in BICARICU-1 for all comer severe metabolic acidoses with no obvious benefit. Now we have the bicarb strikes back, this time focussing on the AKI population which I will admit in advance is the group i find myself doing this in a lot.

Methods

• Enrolled ICU patients with pH<7.2 and an AKI. The obvious exclusions were DKA patients but also toxic alcohol poisoning. who both have well established pathways.
• 4.2% bicarb was used (which is half the strength of the stuff commonly available in ireland) typically given over 60 mins. equivalent to about an amp of our 8.4% stuff.
• Unblinded
• Primary outcome was mortality at 90 days which is a relatively robust outcome
• The secondary outcomes included need for dialysis but I'd suggest that this is believe it or not quite a subjective decision and given the trial isn't blinded is therefore much less robust
• Looking for a 10% absolute mortality decrease here but in a very sick population. Needed ~650 pts for this.

Results

• 640 pts
• Mid 60s, mostly medical with comorbidities, half had septic shock as main cause at enrolment which would seem typical to me for acidaemia. pH 7.15 and lactates of 6 emphasising how sick these patients were
• I would note that only 8% had an isolated AKI as cause of acidaemia
• 62% mortality each group
• 30 hrs vs 15 hrs to start dialysis suggesting the bicarb delayed that decision but remember that is potentially quite biased in an open label trial like this.

Thoughts

• Bicarb has also been one of those therapies looking for a good evidence basis and yet again struggling to find it. Treating the sepsis and getting source control is likely to be the more important focus rather than just fixing the number.
• I continue to use this on the very small population of patients with a nasty AKI, a nasty acidaemia who also need volume

The Droopy Dribbler: Paediatric Presentations of Bell’s Palsy - RCEMLearning

RCEM Learning Methaemoglobinaemia Reference

Genitourinary Foreign Bodies - RCEMLearning