Scott Cohen

460 posts

Scott Cohen banner
Scott Cohen

Scott Cohen

@cohensa

MD-PhD Student | PhD in Epidemiology Infectious Disease Epidemiology + AMR + Causal AI

Katılım Şubat 2015
503 Takip Edilen185 Takipçiler
Scott Cohen
Scott Cohen@cohensa·
@PulmCrit One is a tool, the other a design framework, often used together. Yes, observational and limited by unmeasured confounding, but good design reduces real bias, not just decorates it. We need these studies for questions RCTs can’t or won’t ever answer, so why not do them better?
English
1
0
0
141
𝙟𝙤𝙨𝙝 𝙛𝙖𝙧𝙠𝙖𝙨 💊
propensity matching is so last year target trial emulation is the new hotness but fundamentally- although some methods are better than others - these are all retrospective observational studies dressed up in fancy statistics
GIF
Brad Spellberg@BradSpellberg

@ABsteward @DrToddLee I wish people would stop with this “target trial emulation” business. You did a retrospective observational study with propensity matching. Period. Be honest & call it that. Result won’t affect my practice. If atypical pna possible give azithro. If it’s a typical pathogen, don’t.

English
1
1
20
3.8K
Scott Cohen
Scott Cohen@cohensa·
@Papa_Heme Not so fun fact - Ertapenem still selects for Pseudomonas multi-drug resistance despite in vitro inactivity.
English
0
0
8
929
Papa Heme
Papa Heme@Papa_Heme·
Ertapenem = Meropenem - Psueodomonas Why would anyone use ertapenem if no pseudomonas coverage? Works great for ESBLs and is dosed once per day (easy to send someone home on this antibiotic)
English
4
22
193
64.4K
Scott Cohen
Scott Cohen@cohensa·
@alexolegimas The median number of **views** per paper is 0-1, let alone citations
English
0
0
4
398
Scott Cohen
Scott Cohen@cohensa·
@AnilMakam Nothing wrong with improving observational design. RCTs cannot be the only standard we rely on for causal questions.
English
0
0
3
124
Brad Spellberg
Brad Spellberg@BradSpellberg·
@ABsteward I don’t think sterilizing blood cultures of msaa within 72 hours is that impressive. MRSA? Yes. Mssa? No. Folk are gonna need to do a heckuva lot better than that to get me to stop telling people not to use erta for GPCs.
English
3
2
35
4.1K
Antibiotic Steward Bassam Ghanem 🅱️C🆔🅿️🌟
This is the largest study to date of patients who received ertapenem combination therapy for MSSA bacteremia. The authors found that 85% of patients sterilized their blood cultures within 72 h of treatment, despite nearly half of the cohort having definitive endocarditis. Awaiting for the definitive answer from CERT RCT #IDXposts #IEWiki
Antibiotic Steward Bassam Ghanem 🅱️C🆔🅿️🌟@ABsteward

🆕🔥🧫 Three Days to Clear? Ertapenem Combo Therapy and Rapid MSSA Blood Culture Sterilization, AAC, 2025 🚀 What they did 🏥 Retrospective multicenter study of MSSA bacteremia patients receiving ertapenem combination therapy ⏱️ Defined rapid sterilization as blood cultures negative within ≤72 hours after ertapenem start 🔍 Compared rapid vs delayed clearance and explored predictors (severity, MICs, endocarditis) 🔥 Key findings 🚀 85% (167/197) achieved blood culture sterilization within 72 hours ⏰ Median time to ertapenem start: 4 days after initial blood cultures 🫀 47% had definite endocarditis — yet rapid clearance remained common ⚠️ Delayed sterilization linked to: Higher Pitt bacteremia score 📈 Oxacillin MIC ≥0.5 µg/mL 🤔👀 💊 Suggests early, potent synergy with ertapenem combination therapy ⚠️ Limitations Retrospective design No comparator group without ertapenem Clinical outcomes beyond culture clearance not fully assessed 💡 Takeaway For MSSA bacteremia started on combination therapy, most patients clear bacteremia within 72 hours — supporting a short, 3-day ertapenem course as a pragmatic antimicrobial stewardship strategy, rather than prolonged combination therapy CERT RCT is ongoing @DrToddLee journals.asm.org/doi/10.1128/aa…

English
1
10
51
8.9K
Scott Cohen
Scott Cohen@cohensa·
@PulmCrit @AnilMakam Are you both unconvinced they properly controlled for baseline clinical improvement up until enrollment? The subgroup analysis and post-adjustment tables make it seem the weighting worked well.
English
1
0
0
25
Anil Makam
Anil Makam@AnilMakam·
I like the question and comports with my priors But i dont like the design because its fatally confounded People who get deescalated represent both real clinical improvement & practice variation Former cant be adjusted despite "target trial emulation"
JAMA Internal Medicine@JAMAInternalMed

Among adults hospitalized for community-onset #sepsis, broad-spectrum #antibiotic de-escalation at day 4 resulted in similar mortality, fewer antibiotic days, and shorter hospitalizations vs continued therapy. ja.ma/3LaeCxr

English
3
5
19
6K
Scott Cohen
Scott Cohen@cohensa·
@David_Ouyang @boulware_dr I had posted chart of success rates/applications by degree from NIH Reporter but deleted it because it wasn’t research awards. Do you know if a change in number of MD-only applications can explain the decrease in proportion of awards?
English
0
0
0
24
Scott Cohen retweetledi
Open Forum Infectious Diseases
Open Forum Infectious Diseases@OFIDJournal·
Dynamic networks of methicillin-resistant Staphylococcus aureus in communities drive hospital transmission ✅ Just Accepted 🔗 bit.ly/3RHuGqa
Open Forum Infectious Diseases tweet media
English
0
4
7
1.3K
Scott Cohen
Scott Cohen@cohensa·
@jbcarmody Declines can be explained almost entirely by decrease in repeat applicants (presumably due to higher enrollment). First time applicants are up.
English
3
0
14
2K
Scott Cohen
Scott Cohen@cohensa·
@PMurawala Had a nearly identical issue on a grant. It is a problem with shared first-author papers when your name just happens to be second. Fortunately, it was entirely resolved on resubmission. There really needs to be a better system to avoid inadvertent penalty on shared authorship.
English
0
1
9
3.3K
MurawalaLab
MurawalaLab@PMurawala·
Dear Reviewer, You have one job and that is to read grant properly. In case you are on X, here is my first author publication during my postdoc. science.org/doi/abs/10.112…
MurawalaLab tweet media
English
56
39
738
257.1K
Scott Cohen
Scott Cohen@cohensa·
@statsepi What about when trials are not possible? Wouldn’t that be “novel” treatment effects?
English
1
0
0
266
Scott Cohen retweetledi
Eric Topol
Eric Topol@EricTopol·
More than 1 million people are dying each year from antibiotic resistance, considered a global health emergency. The pharma industry is doing little to respond. Thankfully academic labs, using A.I. and structure-based discovery, are kicking in. My summary Table of recent progress
Eric Topol tweet media
English
1
284
1K
257.3K
Jonathan
Jonathan@IcyTonyan·
@SethWalder Just wondering what was the expected conversion % in that specific situation
English
2
0
0
275
Seth Walder
Seth Walder@SethWalder·
35%. That's another way to know there's only one move. No justification for the PAT.
Scott Cohen@cohensa

@SethWalder What is the breakeven conversion percentage?

English
2
0
9
16.9K
Seth Walder
Seth Walder@SethWalder·
Win probability swing for the Bucs 2-point decision down 8 in the fourth quarter WP PAT: 8.1% WP Go for 2: 9.3%
English
5
1
18
14.2K
Manni Mohyuddin
Manni Mohyuddin@ManniMD1·
This Christmas, I leave you all with a gem of an article. This industry written cost-effectiveness analysis claims denosumab (priced $2,500 per dose) is more cost-effective than zometa ($50 per dose) 😂 tandfonline.com/doi/full/10.10… (Yes, the earth is flat).
Manni Mohyuddin tweet media
English
11
6
60
23K