Assistant Professor @UCSF | Prior @Ci2iFellowship @StanfordMedRes @WUSTLmed
San Francisco, CA
Joined August 2018
- Tweets213
- Following976
- Followers321
- Likes2.2K
Excited to join @UCSF_DHM_SFGH!
It's official
The @robjgallo era has began with us @UCSF_DHM_SFGH
Couldn't be more excited for him, for us, and most importantly, for the betterment of medicine
Excited to join the JAMA Editorial Fellowship Program!
JAMA today announces a new cohort of ten academic physicians and nurses selected for the JAMA Editorial Fellowship Program.
The fellows come to us from around the US and internationally, with a broad range of clinical interests. Read about each fellow: ja.ma/4ho5xLW
Rob Gallo, MD retweeted
Provocative result we did NOT expect. We fully expected the Doctor + GPT4 arm to do better than Doctor + "conventional" Internet resources. Flies in the face of the Fundamental Theorem of Informatics (Human + Computer is Better than Either Alone).
Rob Gallo, MD retweeted
Our first RCT on using an LLM on diagnostic reasoning is out!
And the results are 🔥🌶️... adding ChatGPT did NOT improve diagnostic accuracy or reasoning, and the AI alone outperformed ALL the humans.
What does this mean? A 🧵⬇️
nitter.cf/EricTopol/status/18509…
A small randomized trial of generative #AI for diagnosis again (as seen in a few previous studies) shows higher performance for #AI than physicians + AI. May indicate that physicians need to be trained on how to incorporate AI.
jamanetwork.com/journals/jam…
Rob Gallo, MD retweeted
Only a fraction of the doctors realized they could literally copy-paste in an entire case history into the chatbot and ask it for a surprisingly smart and comprehensive answer.
nytimes.com/2024/11/17/healt…
Rob Gallo, MD retweeted
A small randomized trial of generative #AI for diagnosis again (as seen in a few previous studies) shows higher performance for #AI than physicians + AI. May indicate that physicians need to be trained on how to incorporate AI.
jamanetwork.com/journals/jam…
Rob Gallo, MD retweeted
🧐 Is early PEG after stroke a #TWDFNR ? Maybe observing with NG tube is a viable option. @robjgallo and coauthors review the pros and cons.
🔗 shmpublications.onlinelibrar…
#VisualAbstract by @PatriciaTranMD
Rob Gallo, MD retweeted
World's fastest 🏃♀️➡️ will be crowned tomorrow. How fast should you send stroke patients for PEG? @robjgallo et al. suggest 🛑 and assess dysphagia risk in this #TWDFNR
🔗 shmpublications.onlinelibrar…
#VisualAbstract by @PatriciaTranMD
Great point! Although, I reviewed a manuscript not so long ago where the authors of a cluster randomized study didn't incorporate clustering in their statistical model, so I worry that even in that setting it is often ignored
#Statistics thought of the day: In cluster randomized trials one pays a big price in effective sample size to account for non-individual-level-randomization. Why in observational studies we almost never see that price being paid, i.e. effective N is assumed to be observed N?
Rob Gallo, MD retweeted
Bottom line is, a more complex model isn't neccesarily more accurate or more like "causal effects" or more interpretable. Thanks to intermediates in the causal chain, colliders, Table 2 fallacy, bias-variance trade-off, model misspecification, etcetera
Rob Gallo, MD retweeted
Excited to share our insights on recent research uncovering racial disparities in readmission rates from Medicare data! @StanfordHospMed jamanetwork.com/journals/jam…
Rob Gallo, MD retweeted
Increase your sample size by asking LLM chatbot the same question 100 times. Wait, maybe not? @robjgallo diving in on evaluating generative #AI. Repeated prompting like asking the same person a question or like random sampling from a population of people?jamanetwork.com/journals/jam…
Rob Gallo, MD retweeted
Fascinating regression discontinuity study in @JAMAInternalMed of Epic's deterioration index (EDI) by @robjgallo
EDI alerts reduced rapid response & ICU transfers
Though sensitive to bandwidth choice below & above the EDI threshold when the alert fires
jamanetwork.com/journals/jam…
Rob Gallo, MD retweeted
In this quasi-experimental study, a deterioration model intervention was found to be associated with a decreased risk of escalations in care during hospitalization. ja.ma/4aqjG7V
Rob Gallo, MD retweeted
Mitigating the 'black box' of AI: LLMs can imitate diagnostic reasoning strategies when solving clinical cases & provide an interpretable means to assess if the generated answer is true/false based on the diagnostic reasoning's factual & logical accuracy.
nature.com/articles/s41746-0…
Rob Gallo, MD retweeted
Is General Internal Medicine as well funded as other IM subspecialties when it comes to early career development grant awards? Look at this paper by @robjgallo et al to find out! #internalmedicine #grantfunding buff.ly/411sduu
Rob Gallo, MD retweeted
@robjgallo using regression discontinuity causal Inference methods to determine whether Epic deterioration models actually improved patient outcomes. #AMIA2023
Rob Gallo, MD retweeted
Wonderful Hospital Medicine Grand Rounds today with Drs. @StephenMaMDPhD and Robert Gallo presenting on Design and Evaluation of Machine Learning Interventions in Clinical Contexts! @Neera_Ahuja
Rob Gallo, MD retweeted
Robert Gallo, a postdoc at Stanford, presents at the 2023 NLM T15 Conference hosted by Stanford DBDS on Friday morning.
Diagnosis codes significantly underestimate inpatient hypoglycemic events among patients with diabetes. @ADA_Pubs
Read Here➡️doi.org/10.2337/dc23-0053
ALT Baseline characteristics and diagnostic performance