2015
DOI: 10.1002/sam.11262
|View full text |Cite
|
Sign up to set email alerts
|

A predictive framework for modeling healthcare data with evolving clinical interventions

Abstract: Medical interventions critically determine clinical outcomes. But prediction models either ignore interventions or dilute impact by building a single prediction rule by amalgamating interventions with other features. One rule across all interventions may not capture differential effects. Also, interventions change with time as innovations are made, requiring prediction models to evolve over time. To address these gaps, we propose a prediction framework that explicitly models interventions by extracting a set o… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
5
0

Year Published

2016
2016
2021
2021

Publication Types

Select...
6
2
1

Relationship

0
9

Authors

Journals

citations
Cited by 10 publications
(5 citation statements)
references
References 61 publications
0
5
0
Order By: Relevance
“…Dynamic Bayes Networks: While survival models are by far the predominant type of models, other methods that can incorporate temporal information also exist. Dynamic Bayesian networks (DBN) [Melnyk et al 2013] have been used to model temporal relationships among EHR variables [Rana et al 2015]. Nachimuthu et al [Nachimuthu et al 2010] used DBN's to model temporal relationships between insulin and glucose homeostasis.…”
Section: Cohort and Case-control Study Designmentioning
confidence: 99%
“…Dynamic Bayes Networks: While survival models are by far the predominant type of models, other methods that can incorporate temporal information also exist. Dynamic Bayesian networks (DBN) [Melnyk et al 2013] have been used to model temporal relationships among EHR variables [Rana et al 2015]. Nachimuthu et al [Nachimuthu et al 2010] used DBN's to model temporal relationships between insulin and glucose homeostasis.…”
Section: Cohort and Case-control Study Designmentioning
confidence: 99%
“…On the other hand, there exist statistical techniques, which utilize statistical models to mathematically formalize the generative process of the data under assumptions. They can elucidate hidden mechanism of the target system and also incorporate the effect of time such as dynamic Bayesian networks (DBNs) [1,2] and the state space models (SSM) [3][4][5]. For example, Nachimuthu et al [6] used DBNs to model temporal relationships between insulin and glucose homeostasis and predicted the future glucose levels of a patient admitted in an ICU.…”
Section: Introductionmentioning
confidence: 99%
“…They used the Random Forest and no temporal representation. Rana et al (2015) had introduced a framework that models the change in interventions over time to predict outcome events considering the temporal evolution of the events, which was shown to be useful.…”
Section: Introductionmentioning
confidence: 99%