Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods

The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a sing...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2017-03
Hauptverfasser: Cheng, Ju, Combs, Mary, Lendle, Samuel D, Franklin, Jessica M, Wyss, Richard, Schneeweiss, Sebastian, Mark J van der Laan
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Cheng, Ju
Combs, Mary
Lendle, Samuel D
Franklin, Jessica M
Wyss, Richard
Schneeweiss, Sebastian
Mark J van der Laan
description The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a single prediction model, but uses the strengths of a variety of learning algorithms to adapt to different databases. While the SL has been shown to perform well in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of the SL in its ability to predict treatment assignment using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also considered a novel strategy for prediction modeling that combines the SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2074100582</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2074100582</sourcerecordid><originalsourceid>FETCH-proquest_journals_20741005823</originalsourceid><addsrcrecordid>eNqNjbFqAkEQQBchoBj_YcD6YN3z1F4UiwgB08u4O3or5845s1dY5s-9SIqUqV7zeG9gRq4sZ8Vq7tzQTFSv1lq3WLqqKkfm-1O4paQxP0A9C0ErFKLPkROcWYAa8lk4RQ81YZNrj70UMOMJlRQ6jekCh64lgQ9CST0xBdjFS12EePtpc8IG_owOr9Gecs1B383bGRulyS_HZrrdfK13RSt870jz8cqd9AU9Orucz6ytVq78n_UESrZSMA</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2074100582</pqid></control><display><type>article</type><title>Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods</title><source>Free E- Journals</source><creator>Cheng, Ju ; Combs, Mary ; Lendle, Samuel D ; Franklin, Jessica M ; Wyss, Richard ; Schneeweiss, Sebastian ; Mark J van der Laan</creator><creatorcontrib>Cheng, Ju ; Combs, Mary ; Lendle, Samuel D ; Franklin, Jessica M ; Wyss, Richard ; Schneeweiss, Sebastian ; Mark J van der Laan</creatorcontrib><description>The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a single prediction model, but uses the strengths of a variety of learning algorithms to adapt to different databases. While the SL has been shown to perform well in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of the SL in its ability to predict treatment assignment using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also considered a novel strategy for prediction modeling that combines the SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Algorithms ; Health care ; Machine learning ; Mathematical models ; Modelling ; Optimization ; Performance prediction ; Pharmacology</subject><ispartof>arXiv.org, 2017-03</ispartof><rights>2017. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>780,784</link.rule.ids></links><search><creatorcontrib>Cheng, Ju</creatorcontrib><creatorcontrib>Combs, Mary</creatorcontrib><creatorcontrib>Lendle, Samuel D</creatorcontrib><creatorcontrib>Franklin, Jessica M</creatorcontrib><creatorcontrib>Wyss, Richard</creatorcontrib><creatorcontrib>Schneeweiss, Sebastian</creatorcontrib><creatorcontrib>Mark J van der Laan</creatorcontrib><title>Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods</title><title>arXiv.org</title><description>The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a single prediction model, but uses the strengths of a variety of learning algorithms to adapt to different databases. While the SL has been shown to perform well in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of the SL in its ability to predict treatment assignment using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also considered a novel strategy for prediction modeling that combines the SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.</description><subject>Algorithms</subject><subject>Health care</subject><subject>Machine learning</subject><subject>Mathematical models</subject><subject>Modelling</subject><subject>Optimization</subject><subject>Performance prediction</subject><subject>Pharmacology</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2017</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNjbFqAkEQQBchoBj_YcD6YN3z1F4UiwgB08u4O3or5845s1dY5s-9SIqUqV7zeG9gRq4sZ8Vq7tzQTFSv1lq3WLqqKkfm-1O4paQxP0A9C0ErFKLPkROcWYAa8lk4RQ81YZNrj70UMOMJlRQ6jekCh64lgQ9CST0xBdjFS12EePtpc8IG_owOr9Gecs1B383bGRulyS_HZrrdfK13RSt870jz8cqd9AU9Orucz6ytVq78n_UESrZSMA</recordid><startdate>20170314</startdate><enddate>20170314</enddate><creator>Cheng, Ju</creator><creator>Combs, Mary</creator><creator>Lendle, Samuel D</creator><creator>Franklin, Jessica M</creator><creator>Wyss, Richard</creator><creator>Schneeweiss, Sebastian</creator><creator>Mark J van der Laan</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20170314</creationdate><title>Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods</title><author>Cheng, Ju ; Combs, Mary ; Lendle, Samuel D ; Franklin, Jessica M ; Wyss, Richard ; Schneeweiss, Sebastian ; Mark J van der Laan</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_20741005823</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Algorithms</topic><topic>Health care</topic><topic>Machine learning</topic><topic>Mathematical models</topic><topic>Modelling</topic><topic>Optimization</topic><topic>Performance prediction</topic><topic>Pharmacology</topic><toplevel>online_resources</toplevel><creatorcontrib>Cheng, Ju</creatorcontrib><creatorcontrib>Combs, Mary</creatorcontrib><creatorcontrib>Lendle, Samuel D</creatorcontrib><creatorcontrib>Franklin, Jessica M</creatorcontrib><creatorcontrib>Wyss, Richard</creatorcontrib><creatorcontrib>Schneeweiss, Sebastian</creatorcontrib><creatorcontrib>Mark J van der Laan</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Cheng, Ju</au><au>Combs, Mary</au><au>Lendle, Samuel D</au><au>Franklin, Jessica M</au><au>Wyss, Richard</au><au>Schneeweiss, Sebastian</au><au>Mark J van der Laan</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods</atitle><jtitle>arXiv.org</jtitle><date>2017-03-14</date><risdate>2017</risdate><eissn>2331-8422</eissn><abstract>The optimal learner for prediction modeling varies depending on the underlying data-generating distribution. Super Learner (SL) is a generic ensemble learning algorithm that uses cross-validation to select among a "library" of candidate prediction models. The SL is not restricted to a single prediction model, but uses the strengths of a variety of learning algorithms to adapt to different databases. While the SL has been shown to perform well in a number of settings, it has not been thoroughly evaluated in large electronic healthcare databases that are common in pharmacoepidemiology and comparative effectiveness research. In this study, we applied and evaluated the performance of the SL in its ability to predict treatment assignment using three electronic healthcare databases. We considered a library of algorithms that consisted of both nonparametric and parametric models. We also considered a novel strategy for prediction modeling that combines the SL with the high-dimensional propensity score (hdPS) variable selection algorithm. Predictive performance was assessed using three metrics: the negative log-likelihood, area under the curve (AUC), and time complexity. Results showed that the best individual algorithm, in terms of predictive performance, varied across datasets. The SL was able to adapt to the given dataset and optimize predictive performance relative to any individual learner. Combining the SL with the hdPS was the most consistent prediction method and may be promising for PS estimation and prediction modeling in electronic healthcare databases.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2017-03
issn 2331-8422
language eng
recordid cdi_proquest_journals_2074100582
source Free E- Journals
subjects Algorithms
Health care
Machine learning
Mathematical models
Modelling
Optimization
Performance prediction
Pharmacology
title Propensity score prediction for electronic healthcare databases using Super Learner and High-dimensional Propensity Score Methods
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T21%3A25%3A34IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Propensity%20score%20prediction%20for%20electronic%20healthcare%20databases%20using%20Super%20Learner%20and%20High-dimensional%20Propensity%20Score%20Methods&rft.jtitle=arXiv.org&rft.au=Cheng,%20Ju&rft.date=2017-03-14&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2074100582%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2074100582&rft_id=info:pmid/&rfr_iscdi=true