Machine learning for identifying patients at risk of severe sepsis. Talk by @suchisaria at #IDWeek2022. Input + Output -> Rules. pubmed.ncbi.nlm.nih.gov/32166234/
This table highlights why we are not going to accurately identify sepsis in real-time without machine learning. pubmed.ncbi.nlm.nih.gov/31939780/ #IDWeek2022 @suchisaria Image
"Machine learning-based clinical decision support tools for sepsis create opportunities to identify at-risk patients and initiate treatments at early time points." 89% of ~9800 alerts reviewed by physician/APP over 2 yrs -> 2hr less time to 1st abx. pubmed.ncbi.nlm.nih.gov/35864251/
Those whose machine learning-based early warning system for sepsis was confirmed within 3 hours had a reduced mortality rate, organ failure, and length of stay. pubmed.ncbi.nlm.nih.gov/35864252/ #IDWeek2022

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Michael Calderwood, MD, MPH

Michael Calderwood, MD, MPH Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @CalderwoodMD

Oct 22
1/10 Most influential infection prevention papers presented at #IDWeek2022 (non-COVID). 1. In >300 HCP caring for monkeypox patients, only 23% wore all recommended PPE, yet no HCP infections in 21 days. Risk of transmission in health care settings is low. pubmed.ncbi.nlm.nih.gov/36136939/
2/10 Near ZERO viral detection when swabbing nostrils of HCWs exposed to a nebulized virus in a sealed clinical room for 40 mins wearing a fit-tested N95 (that passed) plus HEPA filtration. Wearing a non-fit-tested or failed fit-tested N95 not protective. pubmed.ncbi.nlm.nih.gov/35535021/
3/10 "In patients with symptom resolution of C. difficile infection AFTER treatment with standard-of-care antibiotics [PO Vanco or Fidaxomicin), oral administration of SER-109 was superior to placebo in reducing the risk of recurrent infection." pubmed.ncbi.nlm.nih.gov/35045228/
Read 10 tweets
Oct 21
1/8 Most influential stewardship papers presented at #IDWeek2022. 1. Women in rural settings MORE LIKELY to get antibiotic treatment for uncomplicated UTI for a duration longer than recommended by guidelines, compared with urban settings. pubmed.ncbi.nlm.nih.gov/33622432/
2/8 Why does this matter? "Each day of antibiotic therapy was associated with 4% increased odds of experiencing an adverse event." pubmed.ncbi.nlm.nih.gov/34775072/
3/8 ED specific urine antibiogram, UTI guideline, education, department specific feedback followed by re-education & provider specific feedback -> initial 15% improvement in guideline-concordant abx followed by additional 3% improvement Q2wks during study. pubmed.ncbi.nlm.nih.gov/35100621/
Read 8 tweets
Oct 21
Clinical trials in bacterial diseases that may change your practice. #IDWeek2022 1. De-escalation of antibiotics for VAP based on Gram stain was non-inferior to waiting for culture. pubmed.ncbi.nlm.nih.gov/35394515/
2. Obtaining follow-up blood cultures in patients with Gram negative bacteremia associated with decreased mortality. pubmed.ncbi.nlm.nih.gov/36136334/
3. SABATO trial presented at ECCMID 2022. Oral switch to Bactrim 1 DS PO BID (or clindamycin 600mg PO Q8hr or linezolid 600mg PO BID) non-inferior to IV course for Staph aureus bacteremia secondary to venous catheter or skin and soft tissue infection. clinicaltrials.gov/ct2/show/NCT01…
Read 8 tweets
Oct 21
At #IDWeek2022, @hboucher3 presented data that ~1.3 million deaths were directly attributable to antimicrobial resistance in 2019. pubmed.ncbi.nlm.nih.gov/35065702/
Although @IDSAInfo 10 by ‘20 initiative resulted in 14 new ID drugs coming to market, “the majority…have been modifications of existing chemical classes of antibiotics [and] larger pharmaceutical companies continue to abandon the field.” Nobody’s buying. pubmed.ncbi.nlm.nih.gov/30715222/
“The antibiotic market is broken and won’t fix itself,” says @hboucher3 at #IDWeek2022.
Read 4 tweets
Oct 20
1/4 Dr. Klompas presented at #IDWeek2022 on a big change in the @SHEA_Epi Compendium for VAP Prevention. STOP oral care with chlorhexidine and switch to oral care with toothbrushing. pubmed.ncbi.nlm.nih.gov/35589091/ Image
2/4 Data from double-blind RCTs doesn't show lower VAP rates with chlorhexidine oral care, there is no reduction in ICU length of stay or duration of mechanical ventilation, and there is concern about higher mortality. pubmed.ncbi.nlm.nih.gov/24663255/, pubmed.ncbi.nlm.nih.gov/24687313/ ImageImageImageImage
3/4 Risk may be from increased oral health dysfunction with chlorhexidine which can be caustic to oral mucosa. pubmed.ncbi.nlm.nih.gov/34609548/ Dr. Klompas also discussed potential ARDS related to aspiration of chlorhexidine. Image
Read 4 tweets
Dec 7, 2021
1/8 Let's talk New Hampshire. We need to have an honest dialogue about how bad things are right now with #COVID19. We currently are averaging more cases per day than at any point in the pandemic (66x higher than last summer). @SteveAhnen @NHHospitals @NHMedSociety @DartmouthHitch
2/8 The PCR positivity rate is back to where it was in April 2020, with 3 counties/cities reporting PCR positivity rates >15% (Coos, Sullivan, Nashua) and some clinics in the south of the state are at 25%. As this increases, true >> reported infections.
3/8 The entire state of NH is at substantial levels of community transmission. In fact, we are currently 10x higher than the level that determines substantial transmission, with Sullivan Country 18x higher.
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(