Global Incidence of Frailty and Prefrailty Among Community-Dwelling Older Adults ja.ma/2GIRhOR

.Brown University @HealthyBrown study shows increasing staffing at long-term care facilities may reduce hospitalizations for pneumonia, influenza. Via @Contagion_live. bit.ly/2Xbdz1P

A third of older nursing home residents don't have statins or beta-blockers initiated after an acute MI. #geriatrics onlinelibrary.wiley.com/doi/10.1111/jg…

Thanks to Evidencio for this online implementation of the FRAiL Tool for hip fracture risk in NH residents! evidencio.com/public/show/916

Want to estimate the risk of hip fracture for your loved one in the nursing home? Try the new FRAiL tool: doi.org/10.1093/gerona…

Metformin Warnings and Diabetes Drug Prescribing Patterns for Older Nursing Home Residents jamda.com/article/S1525-…

dx.doi.org/10.1080/174501…

Faced With Unaffordable Drug Prices, Tens Of Millions Buy Medicine Outside U.S. khn.org/NjgyNzgz via Kaiser Health News

New paper! Association of β-Blockers With Functional Outcomes After Acute Myocardial Infarction ja.ma/2gFEQsj

aspph.org/brown-research…

@BrownCEbM AHRQ K12 Scholar Andrew Zullo, PharmD recently published on economic incentives to modify HIV-related behaviors in Journal of the IAS

Another feather in the collective cap! Andrew Zullo, PharmD @BrownCEbM AHRQ Journal of the IAS twitter.com/BrownCEbM/stat…

Economic incentives are cost-effective in reducing HIV risk behaviours amg the vulnerable Andrew Zullo, PharmD #endpoverty
bit.ly/1NJx1M1

Four days of antibiotics produces similar outcomes to 8 days in patients w abdominal #sepsis . nej.md/1Ejpqce

This morning, #acic2015 continues with Brown Biostatistics Joe Hogan joining session on Bayesian Inference and Causal Effect at UPennCCEB

Bayesian approach is useful bc it quantifies the uncertainty about unmeasured confounding -L. McCandless #acic2015

Unmeasured confounding requires content area expertise bc the problem is inherently qualitative; a challenge! -L. McCandless #acic2015

Separate effect definition from estimation -L. Stuart #acic2015

Both PATE and SATE are important; if you don't care about bias, use SATE for surveys bc weights add uncertainty -L.M. #acic2015

Sample generalizes to population average tx effect (SATE to PATE) under independence and constant tx effect -L.M. #acic2015