iiinova — where your digital self keeps evolving

Razvan Azamfirei @razvan.azamfirei.com · 31/10/2025

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.

000

Razvan Azamfirei @razvan.azamfirei.com · 04/10/2025

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.

010

Reposted by Razvan Azamfirei

Emily Fridenmaker @emilyfri.bsky.social · 14/07/2025

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.

051

Razvan Azamfirei @razvan.azamfirei.com · 20/03/2025

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.

100

Razvan Azamfirei @razvan.azamfirei.com · 16/02/2025

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.

101

Razvan Azamfirei @razvan.azamfirei.com · 01/02/2025

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."

journals.lww.com

Hemodynamic Management guided by the Hypotension Prediction ... : Anesthesiology

erate-to-high-risk elective abdominal surgery patients. Methods: This multicenter randomized trial was conducted from October 2022 to February 2024 across 28 hospitals evaluating HPI-guided managem...

101

Razvan Azamfirei @razvan.azamfirei.com · 28/12/2024

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

230

Razvan Azamfirei @razvan.azamfirei.com · 25/12/2024

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

010

Razvan Azamfirei @razvan.azamfirei.com · 09/12/2024

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

020

Reposted by Razvan Azamfirei

Jack Iwashyna @iwashyna.bsky.social · 07/12/2024

“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

041

Reposted by Razvan Azamfirei

Emily Fridenmaker @emilyfri.bsky.social · 30/11/2024

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

1439

Razvan Azamfirei @razvan.azamfirei.com · 26/11/2024

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

010

Razvan Azamfirei @razvan.azamfirei.com · 22/11/2024

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

000

Reposted by Razvan Azamfirei

Lakshman Swamy (he/him) @laxswamy.bsky.social · 17/11/2024

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

9114

Razvan Azamfirei @razvan.azamfirei.com · 19/11/2024

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.

110

Razvan Azamfirei @razvan.azamfirei.com · 19/11/2024

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

120

Razvan Azamfirei @razvan.azamfirei.com · 17/11/2024

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.).

110

Razvan Azamfirei @razvan.azamfirei.com · 08/11/2024

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. t.co/59...

000

Razvan Azamfirei @razvan.azamfirei.com · 16/10/2024

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.

000

Razvan Azamfirei @razvan.azamfirei.com · 07/10/2024

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

000

Razvan Azamfirei @razvan.azamfirei.com · 24/09/2024

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. twitter.com/JAMASurgery/status/1838…

000

Razvan Azamfirei @razvan.azamfirei.com · 13/09/2024

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

000

Razvan Azamfirei @razvan.azamfirei.com · 13/08/2024

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

000

Razvan Azamfirei @razvan.azamfirei.com · 28/07/2024

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.

000

Razvan Azamfirei @razvan.azamfirei.com · 17/07/2024

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.

000

Razvan Azamfirei @razvan.azamfirei.com · 29/06/2024

The only way to win the D-dimer game is by never playing it. twitter.com/dr_shai/status/18067884…

000

Razvan Azamfirei @razvan.azamfirei.com · 28/06/2024

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. t.co/TryUswvy9f

000

Razvan Azamfirei @razvan.azamfirei.com · 11/06/2024

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). twitter.com/NC_Renic/status/1800436…

000

Razvan Azamfirei @razvan.azamfirei.com · 10/06/2024

Yes! Please stop posting information patient stories/information from the chart. Regardless of the *actual* potential for identification, it’s a bad look. twitter.com/soupvector/status/18001…

000

Razvan Azamfirei @razvan.azamfirei.com · 04/06/2024

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/ twitter.com/Cghlewis/status/1798020…

000

Razvan Azamfirei @razvan.azamfirei.com · 29/05/2024

Round-to-even should always be the default rounding method. It's internally consistent and less biased. See generally www.clivemaxfield.com/diycalculator…. twitter.com/Cghlewis/status/1795807…

000

Razvan Azamfirei @razvan.azamfirei.com · 29/05/2024

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: t.co/Cjz964Etnl @SCCM @SapnaKmd #AI t.co/J6Nqf8Bbj8

journals.lww.com

ChatGPT and Neuroprognostication: A Snow Globe, Not a... : Critical Care Medicine

An abstract is unavailable.

000

Razvan Azamfirei @razvan.azamfirei.com · 28/05/2024

Amazing, we have now taught computers how to do addition "nearly perfectly". /s twitter.com/SeanMcleish/status/1795…

000

Razvan Azamfirei @razvan.azamfirei.com · 20/05/2024

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". twitter.com/JAMA_current/status/179…

000

Razvan Azamfirei @razvan.azamfirei.com · 20/05/2024

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

000

Razvan Azamfirei @razvan.azamfirei.com · 16/05/2024

No change in functional outcomes or 6-month mortality. Yikes. twitter.com/CCPharmacists/status/17…

000

Razvan Azamfirei @razvan.azamfirei.com · 15/05/2024

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. twitter.com/DrDiGiorgio/status/1790…

000

Razvan Azamfirei @razvan.azamfirei.com · 12/05/2024

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. twitter.com/georgetolisjr/status/17…

000

Razvan Azamfirei @razvan.azamfirei.com · 10/05/2024

“that’s not unreasonable”, while one of the best phrases in medicine, should not be confused with “that’s reasonable”—the true compliment twitter.com/TBrnaughtTB/status/1788…

000

Razvan Azamfirei @razvan.azamfirei.com · 25/04/2024

Generic physiologic/laboratory goals are just a heuristic. MAP is a surrogate and not always reflective of physiologic reality. twitter.com/DrToddLee/status/178314…

000

Razvan Azamfirei @razvan.azamfirei.com · 17/04/2024

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? t.co/NBhPM1nmZe

000

Razvan Azamfirei @razvan.azamfirei.com · 08/04/2024

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. t.co/z5VbFYK1pz t.co/8D8bDyVyAP

000