I finally caught up with my 8-week backlog of @criticalcarereviews.com emails. I'm unsurprised to say that nothing has changed. This is totally not a sub-tweet directed at the Andromeda-Shock-2-stans.

This definitely comes with experience and feeling comfortable interpreting the data by yourself. I always skip the introduction and go straight into the methods, followed by the results, last paragraph of the discussion ± supplement.

Yes. If you're worried about a lung fungus, mica/eraxis ain't it. People reach for it because it's easy (where as vori/isavu/ampho aren't easy), but it's not useful here. And if you think you found candida because your BAL said so....no you didn't.

What I value most about institutions of higher education is their moral compass, their natural inclination to speak truth to power, and how unflappable they are in the face of changing political winds.

Dexmedetomidine-associated fever happens WAY more often than this, particularly in cardiac surgery patients. Such an easy way to trick yourself into an unnecessary infectious workup and antibiotic course.

The Hypotension-Prediction Index--a convoluted way of predicting hypotension by...checking whether the MAP is low--fails once again. Shocker. "HPI-guided hemodynamic therapy did not reduce the incidence of postoperative AKI or overall complications [vs] standard care."

I'm just here for the DL/VL drama. (this is pretty classic confounding by indication)

open.substack.com/pub/sensible... #EMIMCC #Cardiosky #Medsky

Merry Christmas to everyone except for the people being unnecessarily pedantic about hypoxia and hypoxemia.

1) If you're looking for an excuse to deescalate (i.e., procalcitonin), just deescalate 2) I really want to see us move away from magic numbers (5, 7, 10, 14, etc.) but not sure that fixating on a different magic number is the way to go

“Professional ethics must guide us precisely when we are told that the situation is exceptional. Then there is no such thing as ‘just following orders.’” ~ Timothy Snyder

This is incredibly disturbing.

A patient presented in need of an urgent but relatively minor surgery. Vitally stable. Surgery goes fine, but ICU is called because their “sat is too low to extubate.” No prior pulmonary or cardiac conditions. #emimcc

I'm just waiting for the day when we finally stop teaching people that pressors don’t work in acidotic patients.

I don’t have strong opinions about the result, but this study is grossly underpowered. I would not draw any definitive inferences from it.

I add a couple of people every day but there are many more of you who work with critically ill patients. Drop a comment here and I'll add you! But check to see if you're already on it first 😁 go.bsky.app/NC7iD2K

Renal: "We must keep the correction rate below 12 mEq/day, but to avoid overcorrection/this is a high risk patient, we're going to recommend 8 mEq/day." Primary team: "Wait—ODS is (really) bad, so we'll do 6 mEq/day." Maybe this is will be the impetus for some high-quality prospective studies.

Spiciest opinion of the day! Can we still use negative binomial if we don't know how to spell poission though?

One of the most amusing things in medicine is seeing how people dismiss high-quality evidence when it challenges their preferred practices (e.g., robotic surgery, regional anesthesia, TAVRs, TTM, etc.).

This common belief isn't entirely accurate. In acute blood loss, fluid shifts happen quickly—hemoglobin decreases in < 30 minutes, even without IV fluids. The levels won't reflect the magnitude of the blood loss, but someone who lost 40% TBV will not have a normal hemoglobin. https://t.co/59...

Without exaggeration, the font change on PubMed Central by @NCBI is probably the worst thing to have ever happened to science in the history of science.

Here's your friendly reminder that Plasmalyte is NOT affected by the IV Fluid Shortage, just in case you needed another reason to switch.

Looks like we’re still trying to make REBOA happen. 🙃 Regardless, this graph is a bit misleading. Neither B nor C survived to discharge, but not mentioned in the graph. https://twitter.com/JAMASurgery/status/1838518771805233592

At least it’s not Xuebijing again. https://twitter.com/giovannilandoni/status/1837176108585996518

It's #WorldSepsisDay. Did you remember to order your 30mL/kg normal saline boluses?

If you wait long enough, everyone dies. (p < 0.001) https://twitter.com/DFCapodanno/status/1830265384546545770

Time to update the Epic dot phrases, I guess https://twitter.com/drahmedmohsen85/status/1829593512914133229

Waiting for the randomized controlled trial so we can agree to use LR and move on with our lives. https://twitter.com/PharmacoJournal/status/1823345767420944565

I am convinced people don't understand delirium and instead use it for any behavioral disturbance they don't understand. Textbook delirium is often missed, while something that is clearly not delirium gets labeled as such.

I remain befuddled by the amount of Plavix prescribed without CYP2C19 testing. We keep burying our heads in the sand, ignoring the fact that 30% of patients will not respond to the drug. It's baffling.

The only way to win the D-dimer game is by never playing it. https://twitter.com/dr_shai/status/1806788473128415386

Don’t trash your med students publicly. Ask questions to find an entry point to learning and not to “pimp”. Again, don’t trash your med students publicly. https://t.co/TryUswvy9f

Pro tip: Add the date to the beginning of the file name. You can set up shortcuts (e.g., ":date") that automatically expand to today's date (e.g., 2024_06_11). https://twitter.com/NC_Renic/status/1800436101192274112

Yes! Please stop posting information patient stories/information from the chart. Regardless of the *actual* potential for identification, it’s a bad look. https://twitter.com/soupvector/status/1800109994383434018

Or, in a shell: mkdir -p clean_files/fourth_grade clean_files/fifth_grade cp data_collection/observation/**/*4th*.csv clean_files/fourth_grade/ cp data_collection/observation/**/*5th*.csv clean_files/fifth_grade/ https://twitter.com/Cghlewis/status/1798020819026870511

Round-to-even should always be the default rounding method. It's internally consistent and less biased. See generally https://www.clivemaxfield.com/diycalculator/popup-m-round.shtml#A5. https://twitter.com/Cghlewis/status/1795807870224679028

My editorial accompanying @ClementGakuba's work in @CritCareMed is out. I answer two questions: 1. Can #ChatGPT be used for #CriticalCare outcome prediction? No. 2. Does it fail in potentially dangerous ways? Yes. Read more: https://t.co/Cjz964Etnl @SCCM @SapnaKmd #AI https://t.co/J6Nqf8Bbj8

Amazing, we have now taught computers how to do addition "nearly perfectly". /s https://twitter.com/SeanMcleish/status/1795481814553018542

Incredibly valuable viewpoint! It challenges the often-used, but ethically problematic notion that "the best management of any patient with cancer is in a clinical trial". https://twitter.com/JAMA_current/status/1792560165130022975

The replies that are trying to justify poor, convoluted writing are something else. https://twitter.com/Michael_J_Black/status/1791392139642835030

No change in functional outcomes or 6-month mortality. Yikes. https://twitter.com/CCPharmacists/status/1790871392155570539

One can act in a "rational" manner and still be evil. Healthcare is a public good and health insurance should be regulated in a way such that 1) profits are capped or 2) profits are redistributed to policyholders. https://twitter.com/DrDiGiorgio/status/1790240471002087722

This is, unfortunately, an incredibly prevalent attitude in many areas of medicine. Reality be damned; don't let data get in the way of a story driven by ideology. https://twitter.com/georgetolisjr/status/1789413649041207652

“that’s not unreasonable”, while one of the best phrases in medicine, should not be confused with “that’s reasonable”—the true compliment https://twitter.com/TBrnaughtTB/status/1788798198477390090

Don’t cut corners. Don’t lie. https://twitter.com/ross_prager/status/1787470061738819740

Generic physiologic/laboratory goals are just a heuristic. MAP is a surrogate and not always reflective of physiologic reality. https://twitter.com/DrToddLee/status/1783147450330468632

This is, by and large, correct. Many QI initiatives are done by people with clinical knowledge but without an understanding of QI as a discipline. It’s the scientific equivalent of trying to DYI your electrical wiring. Sure, you could probably do it, but should you? https://t.co/NBhPM1nmZe

Shuddering whenever I think about MICE https://twitter.com/ChelseaParlett/status/1780390651538047414

👇👇👇👇 https://twitter.com/ebtapper/status/1778200357593825709

This is the true headline result. Evidence in adults AND children shows: 1. Benzodiazepines cause delirium 2. They do so across the dose range The benefits of benzodiazepines may outweigh the risk of delirium for some, but the risk is real. https://t.co/z5VbFYK1pz https://t.co/8D8bDyVyAP