Informative missingness in electronic health record systems: the curse of knowing

Abstract Electronic health records provide a potentially valuable data source of information for developing clinical prediction models. However, missing data are common in routinely collected health data and often missingness is informative. Informative missingness can be incorporated in a clinical...

Full description

Bibliographic Details
Main Author: Rolf H. H. Groenwold
Format: Article
Language:English
Published: BMC 2020-07-01
Series:Diagnostic and Prognostic Research
Subjects:
Online Access:http://link.springer.com/article/10.1186/s41512-020-00077-0
_version_ 1818190709540257792
author Rolf H. H. Groenwold
author_facet Rolf H. H. Groenwold
author_sort Rolf H. H. Groenwold
collection DOAJ
description Abstract Electronic health records provide a potentially valuable data source of information for developing clinical prediction models. However, missing data are common in routinely collected health data and often missingness is informative. Informative missingness can be incorporated in a clinical prediction model, for example by including a separate category of a predictor variable that has missing values. The predictive performance of such a model depends on the transportability of the missing data mechanism, which may be compromised once the model is deployed in practice and the predictive value of certain variables becomes known. Using synthetic data, this phenomenon is explained and illustrated.
first_indexed 2024-12-12T00:03:01Z
format Article
id doaj.art-a309e76703cb468ab77e85559b905fa5
institution Directory Open Access Journal
issn 2397-7523
language English
last_indexed 2024-12-12T00:03:01Z
publishDate 2020-07-01
publisher BMC
record_format Article
series Diagnostic and Prognostic Research
spelling doaj.art-a309e76703cb468ab77e85559b905fa52022-12-22T00:45:11ZengBMCDiagnostic and Prognostic Research2397-75232020-07-01411610.1186/s41512-020-00077-0Informative missingness in electronic health record systems: the curse of knowingRolf H. H. Groenwold0Department of Clinical Epidemiology, Leiden University Medical CentreAbstract Electronic health records provide a potentially valuable data source of information for developing clinical prediction models. However, missing data are common in routinely collected health data and often missingness is informative. Informative missingness can be incorporated in a clinical prediction model, for example by including a separate category of a predictor variable that has missing values. The predictive performance of such a model depends on the transportability of the missing data mechanism, which may be compromised once the model is deployed in practice and the predictive value of certain variables becomes known. Using synthetic data, this phenomenon is explained and illustrated.http://link.springer.com/article/10.1186/s41512-020-00077-0Prediction modellingMissing dataRoutine care data
spellingShingle Rolf H. H. Groenwold
Informative missingness in electronic health record systems: the curse of knowing
Diagnostic and Prognostic Research
Prediction modelling
Missing data
Routine care data
title Informative missingness in electronic health record systems: the curse of knowing
title_full Informative missingness in electronic health record systems: the curse of knowing
title_fullStr Informative missingness in electronic health record systems: the curse of knowing
title_full_unstemmed Informative missingness in electronic health record systems: the curse of knowing
title_short Informative missingness in electronic health record systems: the curse of knowing
title_sort informative missingness in electronic health record systems the curse of knowing
topic Prediction modelling
Missing data
Routine care data
url http://link.springer.com/article/10.1186/s41512-020-00077-0
work_keys_str_mv AT rolfhhgroenwold informativemissingnessinelectronichealthrecordsystemsthecurseofknowing