I’ve always assumed scanning the chest at the level of the nipples would be best for finding a pneumothorax. Recently, I had a trauma patient where I excluded pneumothorax after a negative ultrasound scan. The CXR was negative as well. But the CT showed a rather big right sided pneumothorax that I hadn’t spotted with US. A recent article in Journal of American Emergency Medicine shows we probably need to scan lower to catch the most pneumos.
The study
What’s the best area to scan for a pneumo? Do you need to scan several, or even all areas of the chest? This was a retrospective convenience sample of CT’s of 183 trauma patients with pneumothorax. The chest was divided into 13 zones shown in the image below. They looked for the zones affected by the pneumothorax and the zone for the largest anterior-posterior diameter.
The results
THere was increasing frequency of pneumothoraces from lateral to medial, and from apical to basal aspects of the lungs. The zones most frequently affected by pneumothorax in this study were 9, 11 and 12, the most anterior and basal zones.
Scan basally
Only larger pneumos need treatment and placement of a chest tube. According to this study, larger pneumos will also affect the more apical zones that I usually scan. But for the highest sensitivity, you need to scan the anterior, but most basal lung aspects. Time to change.
One misconception that is surprisingly common is that metabolic acidosis is the most common acid-base disorder in the ICU. Metabolic alkalosis is the most common acid-base disorder. True, in the acute phase metabolic acidosis is a lot more common but in the long run (due to chloride losses, diuretics and hypoalbuminemia) alkalosis is a lot more common.
After hearing the opposite one too many times, a norwegian professor of anesthesia and intensive care named Erik Waage Nilsen extracted 22.583 base excess measurements from all the arterial blood gasses obtained at his ICU for the last three years. He then published the frequency histogram below.
(Antall = Number N, baseoverskudd = base excess)
He also noted that if there is any bias, it will be towards acidosis as we obviously do more arterial blood gasses in the early acute phase.
Found it in a norwegian medical journal, here (in norwegian).
A study in anesth analg reminds us of why intubating some intensive care patients with succinylcholine can be a bad idea. It demonstrates how the potentially harmful potassium efflux and hyperkalemia we get after administering potassium increases with length of stay in the ICU.
Background
Succinylcholine (sux) is the muscle relaxant of choice for almost every rapid sequence induction intubation ever done in the history of mankind. Most of the time it is an incredibly efficient and safe drug that we all know how to use. However, there are some pit-falls associated with it. One of them is the potential for hyperkalemia caused by the potassium efflux we get when sux binds nicotinic acetylcholine receptors.
Typically serum potassium increases by 0,5-1 mmol/L with a peak 3-5 minutes after injecting sux. If this increase (ΔK) shoots serum potassium to more than around 6,5-7,0 mmol/L the patient risks having a potentially lethal arrhythmia. Which is why we are taught not to use sux in hyperkalemia, burns, rhabdomyolysis , crush syndrome (whatever that is) or other hyperkalemia states. Unfortunately we often forget a much more common reason for hyperkalemia. Up-regulation of cholinergic receptors.
Such up-regulation of receptors happen in diseases and states like anatomical denervation, prolonged administration of neuromuscular blocking drugs, burn injury and prolonged immobilisation. The body is fooled into thinking it needs more cholinergic receptors and decides to make more of them. More receptors means more potassium ions are purged in depolarisation. This in turn means the ΔK risk being a lot bigger than the expected 0,5-1 mmol/L.
Or something along those lines. As a purveyor of the miracle of anesthesia I am perfectly happy knowing stuff happens without me know how or why.
Anyway, this is important as we intubate a whole lot of critical care patients. Many of them suffering from critical care neuropathy/myopathy which results in prolonged immobilisation and denervation. Prior studies have shown up-regulation and exaggerated potassium efflux in these patients.
ICU immobilisation, critical care neuropathy and myopathy are progressive diseases. This means the ΔK abnormality should increase with time, and that is exactly what this study looks at.
The study
A prospective observational study where serum potassium was measured before and after intubation was measured. The study enrolled a general ICU population of 131 patients who were intubated for a total of 153 times. The increase in serum potassium, ΔK, was plotted against the length of ICU stay.
Results
The ΔK from all the intubations are plotted against length of ICU stay in the graph to the right.
ΔK increases with increased length of ICU stay.
At around day 15 the average potassium surge is more than the expected 0,5-1,0. Some of the patients had massive ΔKs in the order of 3-5 mmol-L after 10 or so days.
Take home message
Think twice before using succinylcholine when intubating patients who have long ICU stays or who have developed critical care neuropathy.
We have looked into this issue before, with several succesful case reports and case series being reported. Fuehner et al who has previously released case series, have now compared 26 awake ECMO bridgings to lung transplant (LTx) to historical controls with mechanical ventilation. ECMO comes out in front.
Higher survival, better survival.
The ECMO group seemed to be in better shape, as they needed less time on a ventilator after LTx and had shorter post operative stays. The main finding was that survival after 6 months was 80% in the ECMO group and 50% in the traditional mechanically ventilated group. This could be due to the ECMO group being awake and able to move, eat, receive physiotherapy and generally be in better shape to endure the LTx and the recouperation.
ECMO as standard of care?
ECMO has traditionally been used as a rescue intervention in patients refractory to standard therapy. For the pro-ECMO people, this has been listed as a cause for the bad outcomes of early ECMO trials, combined with the complexity and need for intense anticoagulation of the old ECMO systems. In the newer, easier to use ECMOs, it seems more important to get patients on ECMO before they crash, not as a rescue attempt after they’ve already burned half their organ systems. So this article had a rather new way of doing things, putting patients on ECMO as the first choice when they became to sick to support their own ventilation. There’s a good editorial on the article in Am J Respir Crit Care Med: “Extracorporeal membrane oxygenation as bridge to lung transplantation: what remains in order to make it standard of care?” commenting on ECMO as the possible near future as standard of care for LTx and other patient groups.
Oxygen toxicity has been getting a lot of attention since a few years back. The evidence of how high FiO2 is harmful for patients who are critically ill is somewhat solid and has to some extent changed resus guidelines. An article I found in Aesthesia & Analgesia suggests high peroperative FiO2s increase mortality in a normal surgical population too. Significantly so in cancer patients.
Background
Improved surgical wound oxygenation, through increased oxygen content in the air patients inhale during surgery (FiO2), has been thought to protect against per- and postoperative wound infections. One proposed mechanism was improved oxidative killing by neutrophils.
Studies have showed mixed results. Some studies have showed decreased wound infection with higher FiO2s while others could demonstrate no difference.
One of the largest and most recent trials was the PROXI trial in which general surgical population was randomised into breathing either 80% or 30% O2 during surgery. The PROXI trial could demonstrate no reduction in surgical site infection. More troubling was the increased mortality in the high-O2 group. 30 day mortality in the 80% group was 4,4% versus 2,9% in the 30%.
Study
From the PROXI-trial, it would seem high O2-concentrations are harmful not only in critical illness but is also harming a ´healthy´ population of elective and emergency laparotomies. So the researchers decided to do a follow up on the PROXI-trial. This time looking at the same patients´ long-term mortality and also to see if some patients were harmed more than others.
Results
A total of 1383 patients were randomised into breathing either 80% or 30% preoperatively and for two hours post-op.
After a median follow-up of 2,3 years
23,2% in the 80% O2 group had died.
18,3% in the 30% O2-group had died
Even more striking was mortality in the abdominal cancer patients, the patients who had abdominal cancer surgery.
33,5% of the cancer patients in the high O2 group were dead at follow-up, compared to 24,6% in the physiological 30% O2group.
Take-home message
There is a possibility for higher inhaled O2-concentrations harming even reasonably healthy (from a critical care perspective) patients. Even if only exposed for a few hours. Cancer patients could be exceptionally susceptible to O2-toxicity.
More reasons to maintain anti-O2 vigilance in theatre, in the ED and in the ICU…
A secondary study using data from the ROC PRIMED study suggests there could be something to the warnings about using supraglottic airways (LMAs) for airways in out-of-hospital cardiac arrests. Supraglottic airways were associated with increased mortality.
Background
For the last five years or so, the supraglottic airways (SGAs) have increasingly been used as the preferred airway in out-of-hospital cardiac arrests (OHCA).
Then, in early 2012, Resuscitation published a small animal study by Segal et al. demonstrating significantly (15-50%) reduced carotid blood flow in pigs being resuscitated with a SGA for an airway as compared to traditional ETI-resuscitation.
That decreased carotid blood flow could result in increased mortality and worse functional outcome in the OHCA patients who get SGAs instead of traditional endotracheal intubation.
The study, by Segal et al, generated some controversy on the medical blogs. For the most comprehensive and accessible discussion about SGAs in OHCA you should listen to the EMcritpodcast here. Don’t forget to read the comments field.
The study
ROC PRIMED was a large OHCA trial comparing early vs late ECG analysis as well as ITD vs placebo. The trial was stopped early for futility and became yet another huge, expensive and inconclusive cardiac arrest study that remains controversial. Interestingly, the study protocol did not dictate what airway the EMS were to use. As a consequence there is outcome data from OHCA airway management using both ETI and SGA. The secondary study used that data to explore if the airway used affects outcome.
The secondary analysis included OHCA patients who had successful out-of-hospital airways. Either they had traditional ETI one of three SGAs (King LT, Combitube and LMA). Primary outcome was survival to hospital discharge with a satisfactory functional status, defined as a Modified Rankin Scale score of 3 or more.
Results
The study identified 8487 (81,2%) successful ETIs and 1968 (18,8%) successful SGAs. Of the SGAs 63% were King LTs, 20,5% were Combitube and 16,6% were LMAs.
Survival to hospital discharge with satisfactory functional status was 4,7% in the ETI group compared to 3,9% in the SGA group. This was considered a moderate association.
There were also stronger associations between ETI and ROSC and 24h survival.
Take home message
Supraglottic airways in out-of-hospital cardiac arrests might actually be associated with increased mortality and worse functional outcomes.
Studies like this have huge limitations. Even so, paying attention might pay off. ALS algorithms have in the past changed as a result of weaker evidence than this.
A study in Crit Care Med suggests sepsis patients that, on admission, are on β-blockers could have a survival advantage over patients that are not on β-blockers. Despite the β-blocker patients having more prior cardiac disease, hospitalisations and cardiac risk factors.
Background
The hypothesis that β-blockers might have protective effects in sepsis is not new. At least one clinical trial is underway to explore how beta-blockers influence sepsis.
The theory is sound. Early stage sepsis results in a cathecholaminergic overdrive that could result in myocardial injury or dysfunction. Septic patients who get cardiac dysfunction or injury have a two- to fourfold increase in mortality.
Another aspect is how that same cathecholamine storm stimulate excess cytokine production. By putting a β-blocker lid on that cathecholamine storm perhaps we could put a lid the cytokine storm that then results in inflammation and multi-organ failure.
The study
Retrospective study of hospital records, databases and ambulance documentation. Between 2003 and 2008 the authors identified a population of 9465 patients diagnosed with sepsis. 1061 (11,4%) patients were on β-blockers while 8404 (88,8%) remained unexposed to β-blockers.
Results
For obvious reasons the patients on β-blockers had a higher prevalence of risk factors, cardiac disease and hospitalisations than the β-blocker-naive group.
Despite that, the patients who were on β-blockers had a lower mortality.
Overall mortality 28 days post-ICU-admission was 2045. Patients previously prescribed β-blockers had a mortality of 17,7% while the non-exposed control groups 28 day mortality was 22,1%.
Take-home message
This study, much like this blog, comes with more limitations and disclaimers than the swiss navy. Still, chronic β-blocker prescription may result in a survival advantage in patients with sepsis.
This could happen through attenuation of the early catecholamine surge. Perhaps by putting less stress on the heart.
A more intriguing aspect is how β-blockers could modulate the inflammatory response, central to sepsis pathogenesis, that is in part a result cathecolamine surge.
Perhaps this means the cathecholamine storm should be modulated as early as possible. Even in the emergency room or prehospitally? Administering β-blockers to a pre-shocked patient sounds terribly counterintuitive, but who knows where this will lead?
FOAMed – Free Open-Access Medical Education – as coined by Mike Cadogan, has been the recent label put on the medical blogging world’s posting and discussing of medical topics online. Do a #FOAMed search on Twitter, and you’ll get a lot of interesting links. The blogging world of medicine is a great way to learn, and one example is the recent publishing of the IST-3 trial on thrombolysis in stroke, and the discussion that followed.
The IST- 3 trial
“The benefits and harms of intravenous thrombolysis with recombinant tissue plasminogen activator within 6 h of acute ischaemic stroke (the third international stroke trial [IST-3]): a randomised controlled trial” was published in the Lancet 23rd of June 2012. It was the largest and arguable most prestigeous trial for thrombolysis in stroke. Everyone was cheering on them – hoping thrombolysis worked. And the IST-3 trial concluded it did!
The blogging world roars
The sad part is that the IST-3 numbers doesn’t really support that conclusion. The IST-3 numbers couldn’t really prove any benefit. Smart docs commented on this fact online, and a big debate was sparked on twitter and various blogs. One of the best summaries were posted here: Thrombolysis in Stroke – Are We Done Yet? – slaughtering the IST-3 collaberative’s conclusion. A highly recommended read, incuding the pasted comments.
The Lancet follows
Now, some months later an editorial in the Lancet, who published the article in the first place, says the same thing – in a slightly nicer way. Normally, this kind of revelation in a Lancet editorial would have been a small bomb for me. Now, it was old news. Viva la revolution!
Thrombolysis in Stroke – Are We Done Yet?, EDTCC blog, 13th July 2012.
Thrombolysis in acute ischaemic stroke – editorial, The Lancet, 23rd Sept 2012.
Bleeding down in hemoglobin is a kind of limbo dance for the patient involved. How low can you go? I found a case report on trauma patient with a hemoglobin of 0,7 g/dL(!) yielding a hematocrit of 2.2%(!). O neg blood wasn’t available, and for some reason, blood matching wasn’t ready before 10 hours into his surgery before finally getting blood transfusions. Not only did he survive those 10 hours – his arterial blood gas numbers remained largely normal, with a base excess of -4.
This remarkable survival and numbers were achieved with crystalloids and colloids to keep the patient near normovolemic, and oxygen delivery was achieved through 100% O2. The patient was 34 degrees celcius. Anaesthesia helped lower the metabolism and oxygen demand. The body’s physiological responses also helped keeping the patient alive: hemoglobin doesn’t just transport O2, it’s also the most important carrier of CO2, so the low hemoglobin level also led to CO2 accumulation in peripheral tissue and was likely helping with tissue oxygenation by shifting the oxygen disassociation curve. He did also receive some matched plasma to help him coagulate, and a great plasma expander. Still, the main points is keeping intravascular volume adequate and giving 100% oxygen if you don’t have blood available. This patient survived and is doing well. The case report is well worth a read.
So, we don’t need blood transfusions anymore?
Well, this was a special case. But, yes, often patients can be kept alive at a very low Hb. For hours, maybe days. But eventually, it will kill them. They get organ failure. Of course, it’s hard to make good studies on extreme anaemia, but we do have one group that guinea pigs for this stuff. Jehova’s Witnesses. So some smart guys pooled all Jehova Witness patients in the States over the past years and looked at the ones with an Hb
Hb 1-2: Mortality 100%
Hb 2-3: Mortality 54%
Hb 3-4: Mortality 27%
Hb 4-5: Mortality 34%
Hb 5-6: Mortality 9%
Hb 7-8: Mortality 0%
It was not a surprising finding, but interesting to get an idea of where the cut-off is. Below 5 g/dL is not where you want to be. And above 7 g/dL is looking really good. This fits pretty well with standard transfusion practice with a cut-off for healthy people of 8 g/dL and patients with heart disease 10 g/dL. Even if these levels are also subject to continuous debate.
Another interesting finding with this study was that the anaemic patients didn’t die straight away. I would have though they would end up with acute organ failure and/or cardiac arrest. But most of them managed to stay alive for days, or even over a week, with extremely low Hb levels. So anaemia isn’t the biggest emergency – hypovolemia is. And coagulopathy might also be a more pressing danger than low hemoglobin. But to secure a good outcome in the long run, it is important to keep hemoglobin levels adequate. So, for massive, uncontrollable bleeding the saying still holds true: If they’re bleeding – give blood.
Time course and etiology of death in patients with severe anemia, Transfusion, 2009.
Don’t know who to credit for this amazing image. It could be from that travelling ‘Human Body’ exhibition? Got it on facebook.
If you ever feel bad over a misplaced central line, just go read this case report:
Accidental Puncture of the Pulmonary Artery during a Subclavian Central Venous Catheterization, Case Reports in Critical Care, Volume 2012 (2012).
Seriously? The pulmonal artery?! Respect for publishing that case.
Recently NEJM published the results from the german SHOCK II trial. The study suggests intraaortic balloon pumps (IABP) don´t work as well as we think. It has been interesting following the e-mail correspondance between our hospital cardiac intensivists and cardiologists. They went nuts.
Background
Via a guidewire an IABP is inserted so that the balloon is placed in the proximal descending aorta. It then works in synchrony with the ECG and invasive blood pressures. It is timed to inflate at the onset diastole and deflate in systole.
In theory the diastolic counterpulsation increases diastolic pressures which in turn should improve diastolic myocardial blood flow in oxygenation. In theory the sudden systolic deflation should reduce afterload and increase forward blood flow.
Commonly it is used in patients with cardiac shock (CS) patients due to ischemic disease. It is the most commonly used form of mechanical support until the CS patients can be revascularized with PCI or CABG.
Using IABP in these scenarios is in european and US guidelines given class 1B and 1C recommendations (strong recommendations). Despite that there is no strong evidence proving it works and definitely no adequately powered randomised trials.
Some studies demonstrate increased coronary perfusion and reduced afterload but, as far as I can tell, no studies prove it actually improves cardiac output more than marginally (at best).
The study
600 patients with CS after myocardial infarction were randomised. All were due for early revascularisation with either by-pass surgery or percutaneous coronary intervention. 301 received IABP-support while 299 received no IABP-support. 30 day mortality was the primary end-point.
Results
The vast majority of patients had revascularisation within 24 hours. On average the patients in the IABP-group were on IABP for three days for 3 days.
At 30 days 119 (39,7%) patients in the IABP-group had died while 123 (41,3%) patients in the no-IABP groups had died.
IABP counterpulsation did not significantly reduce mortality in ischaemia related CS due for revascularisation.
Likewise, there was so significant difference in secondary outcomes like time to hemodynamic stabilization, length of ICU-stay, lactate levels and catecholamine therapy.Interestingly there was no improvement of blood-pressure or pulse rate in the IABP-group.
Take-home message
There is still no evidence IABPs are really doing anything for patients in cardiogenic shock CS due for revascularisation. Between these groups of 300 there was no difference in mortality.
However, in the small subgroup of patients <70 years old there is an interesting difference.
Study lives here. It is open access.:
The forgotten access gets a revival thanks to ultrasound, which offers good control with in-plane cannulation. You might find this gets you the best view on kids.
Yoffa’s supraclavicular approach
In 1952 Aubaniac described “A new route for venous injection or puncture: the subclavian route”, using the now familiar infraclavicular approach. In 1965 Yoffa described subclavian vein cannulation via a supraclavicular approach. Through several articles it seems to be just as effective and safe as the infraclavicular approach – but the supraclavicular route is very seldomly used and seems largely forgotten.
Blind technique
It can be used as a blind technique using anatomical landmarks. Yoffa’s original approach involves the clavisternomastoid angle, where the sternocleid attaches to the clavicle. Insertion of the needle 1 cm lateral and 1 cm posterior to this point, aiming slightly upwards and towards the contralateral nipple.
In-plane Ultrasound
With ultrasound, everything feels safer. You can identify the internal jugular vein, the point where it meets the subclavian vein, and then the arch of the subclavian vein. This is easiest on children, where other access routes might be more difficult. With ultrasound, you can cannulate the vein in-plane, always seeing where your needle tip is placed, and where it is headed.
I can see the supraclavicular route as useful during surgery, where I have limited access to the patient. I usually use the internal jugular vein for that, but if the patient will be using the CVC long after the operation and he’ll be ambulating, it’s more comfortable to have a subclavian line. Also, if you for some technical reason can’t tilt the patient head down, the subclavian vein might be the bigger target, held open by surrounding tissue. It could also be a good technique to have in your arsenal as it’s described that swithcing between the supra- and infraclavicular route if one approach gives you problems yields a higher success rate.
There’s a good overview article of the supraclavicular approach found here:
Supraclavicular subclavian vein catheterization: the forgotten central line, West J Emerg Med, 2009.
Small kids
As mentioned above, small children and neonates might be particularly well suited for the supraclavicular route. Central vein access can be challenging on these tiny creatures. The infraclavicular route offers a partly obscured ultrasound view, and internal jugular vein can be a moving target on a small neck. So, enter the ultrasound guided supraclavicular approach. In the junction of the internal jugular and subclavian vein, you can align the linear ultrasound probe with the clavicle, and get a longitudal view of the arch of the subclavian vein. And visualise the pleura directly underneath it.
Thanks to Thanks to Dr. Jean-Paul Mission at Hôpitaux Pédiatriques de Nice in France for uploading the videos. Another one can be found here.
We’re back. The last post before the move turned out to be a bit too optimistic. But after a lot of major minor problems during our ISP and server switch, we’re finally up and running on our new servers.
We’ve learned that if you’re backing up a WordPress blog by backing up your home folder only, you’re missing out. Apparently, all your posts and content is in your SQL database. SQL database??? Yup. And when you switch servers you have to hack your wp_config.php file to get things to work again. wp_config.php??? Yup. A lot of trial and error on our part, after being nonchalantly arrogant about the backup and moving process.
Hopefully, we’re better doctors than web masters… More to come.
