RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @AlexJohnLondon: Awesome to see leading groups deepening development trajectories for #ML systems in #medicine. It’s not enough to have…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @AlexJohnLondon: Awesome to see leading groups deepening development trajectories for #ML systems in #medicine. It’s not enough to have…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @AlexJohnLondon: Awesome to see leading groups deepening development trajectories for #ML systems in #medicine. It’s not enough to have…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @MattBMcDermott: This was an excellent session, and the Q&A after dove much deeper into how we should think about evaluation and the var…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
This was an excellent session, and the Q&A after dove much deeper into how we should think about evaluation and the various associated trade-offs.
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
RT @GSCollins: Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it wi…
Nice (reminder) slide by @kdpsinghlab at @SymposiumML4H - just because you can build a 'good' model doesn't mean using it will be effective - highlighting the importance of (RCT) evaluation. -> https://t.co/ethoaWPUMU (development) -> https://t.co/K
RT @CostAnEffect: RCT of case management with connection to social services to reduce ED visit use in @JournalGIM by @Andy_Ryan_dydx et al.…
RCT of case management with connection to social services to reduce ED visit use in @JournalGIM by @Andy_Ryan_dydx et al. No effect on ED use. https://t.co/2Vw1v2MNDf
@suchisaria Am speaking based on general experiences from this project, where high-risk patients were identified using a model and targeted for care management in an RCT via phone. Null result was partly bc of ppl not picking up the phone, possibly due to
RT @JessHohman: @MagedRizkMD @DrLesWong Food for thought with some of the models we are using now to reduce unnecessary utilization
How it started: https://t.co/3pOcrHovr0 How it’s going: https://t.co/oji9kqa7EU https://t.co/LUBxDpFvgp
@MagedRizkMD @DrLesWong Food for thought with some of the models we are using now to reduce unnecessary utilization
Our new analysis from @BradyJPost and @Andy_Ryan_dydx evaluates predictive model-driven hotspotting to decrease emergency department visits in @JournalGIM @UM_CEHR https://t.co/2qDlATsCFu https://t.co/xZjjYjQqo6
RT @BradyJPost: Case management didn’t reduce ED utilization among complex patients in our randomized controlled trial, just published at @…
RT @BradyJPost: Case management didn’t reduce ED utilization among complex patients in our randomized controlled trial, just published at @…
RT @BradyJPost: Case management didn’t reduce ED utilization among complex patients in our randomized controlled trial, just published at @…
Case management didn’t reduce ED utilization among complex patients in our randomized controlled trial, just published at @JournalGIM (a thread) https://t.co/TuOyVlricd @NUBouve @Northeastern @UM_IHPI @UM_CEHR 1/9
RT @kdpsinghlab: The Camden "hotspotting" RCT in NEJM found that utilization could not be reduced (https://t.co/DTGsFxPIlA). In a spiritua…
The Camden "hotspotting" RCT in NEJM found that utilization could not be reduced (https://t.co/DTGsFxPIlA). In a spiritual successor study led by @Andy_Ryan_dydx, we used an ML model + an RCT to try again. Did it work? It's complicated. Will return to t
RT @JournalGIM: Predictive Model-Driven Hotspotting to Decrease Emergency Department Visits: a Randomized Controlled Trial https://t.co/43Z…
A community-based case management program was not associated with reduced ED utilization among complex patients. #hotspotting https://t.co/XWyXMtJQfB