BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:BSU Seminar: &quot\;Hyper-Localization and Predictive Modeling of 
 Rapid Lung Function Decline in Cystic Fibrosis&quot\; - Rhonda Szczesniak 
 and Emrah Gecili\, both from the Cincinnati Children's Hospital Medical Ce
 nter and the University of Cincinnati
DTSTART:20231212T140000Z
DTEND:20231212T150000Z
UID:TALK208030@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Neighborhood/built environments (the areas in which people liv
 e\, work\, and play) and community context as social and environmental det
 erminants of health have gained prominence with the changing care needs of
  people living with cystic fibrosis (CF) lung disease. Select measures of 
 these social and environmental determinants of health (referred to as “g
 eomarkers”) are also predictors of rapid decline\, which is clinically d
 efined as a prolonged drop in lung function relative to patient and/or cen
 ter-level norms. The extent to which hyper-localization (defined as increa
 sing the spatiotemporal precision of social and environmental exposures) a
 ids in prediction of rapid decline remains unclear. Linear mixed effects (
 LME) models have been historically used for predicting rapid decline in CF
 \, but there are few options to properly incorporate spatial correlation a
 nd induce simultaneous variable selection. The objective of this work is t
 o develop a Bayesian spatial linear mixed effects model to predict rapid d
 ecline using geomarkers.\n\nWe describe an application of the proposed mod
 el for predicting rapid lung function decline (measured as FEV1% predicted
 /year) in a Midwest U.S. cohort of pediatric CF patients aged 6-20 years. 
 We consider a breadth of demographic and clinical characteristics alongsid
 e geomarkers\, which focus on neighborhood/built environments and social/c
 ommunity context. Our innovative Bayesian model uses a “spike and slab
 ” prior\, accounting for spatial correlation based on ZIP code distances
 . We evaluate model fits and prediction accuracies. Our proposed model res
 ults in improved model fit and predictive accuracy\, compared to other Bay
 esian and frequentist LME models with different spatial correlation assump
 tions. We describe how a combination of demographic\, clinical\, and geoma
 rker variables can be selected as optimal predictors based on the posterio
 r inclusion probabilities and Bayesian false discovery rate controlling ru
 le. Our findings suggest that incorporating spatiotemporal effects and geo
 markers results in an improved prediction tool. We discuss how predicting 
 the timing and extent of rapid lung function decline can help clinicians t
 o proactively adjust treatment plans and improve patient outcomes.\n
LOCATION:MRC Biostatistics Unit\, East Forvie Building\, Forvie Site Robin
 son Way Cambridge CB2 0SR.
END:VEVENT
END:VCALENDAR