Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach

(1) Background: Predicting chronic low back pain (LBP) is of clinical and economic interest as LBP leads to disabilities and health service utilization. This study aims to build a competitive and interpretable prediction model; (2) Methods: We used clinical and claims data of 3837 participants of a...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Adrian Richter, Julia Truthmann, Jean-François Chenot, Carsten Oliver Schmidt
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	record linkage machine learning calibration best subset selection low back pain Medicine R
Acceso en línea:	https://doaj.org/article/db524eadd30046b9a458640c2de48848
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:db524eadd30046b9a458640c2de48848
record_format	dspace
spelling	oai:doaj.org-article:db524eadd30046b9a458640c2de488482021-11-25T17:50:22ZPredicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach10.3390/ijerph1822120131660-46011661-7827https://doaj.org/article/db524eadd30046b9a458640c2de488482021-11-01T00:00:00Zhttps://www.mdpi.com/1660-4601/18/22/12013https://doaj.org/toc/1661-7827https://doaj.org/toc/1660-4601(1) Background: Predicting chronic low back pain (LBP) is of clinical and economic interest as LBP leads to disabilities and health service utilization. This study aims to build a competitive and interpretable prediction model; (2) Methods: We used clinical and claims data of 3837 participants of a population-based cohort study to predict future LBP consultations (ICD-10: M40.XX-M54.XX). Best subset selection (BSS) was applied in repeated random samples of training data (75% of data); scoring rules were used to identify the best subset of predictors. The rediction accuracy of BSS was compared to <i>randomforest</i> and <i>support vector machines</i> (SVM) in the validation data (25% of data); (3) Results: The best subset comprised 16 out of 32 predictors. Previous occurrence of LBP increased the odds for future LBP consultations (odds ratio (OR) 6.91 [5.05; 9.45]), while concomitant diseases reduced the odds (1 vs. 0, OR: 0.74 [0.57; 0.98], >1 vs. 0: 0.37 [0.21; 0.67]). The area-under-curve (AUC) of BSS was acceptable (0.78 [0.74; 0.82]) and comparable with <i>SVM</i> (0.78 [0.74; 0.82]) and <i>randomforest</i> (0.79 [0.75; 0.83]); (4) Conclusions: Regarding prediction accuracy, BSS has been considered competitive with established machine-learning approaches. Nonetheless, considerable misclassification is inherent and further refinements are required to improve predictions.Adrian RichterJulia TruthmannJean-François ChenotCarsten Oliver SchmidtMDPI AGarticlerecord linkagemachine learningcalibrationbest subset selectionlow back painMedicineRENInternational Journal of Environmental Research and Public Health, Vol 18, Iss 12013, p 12013 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	record linkage machine learning calibration best subset selection low back pain Medicine R
spellingShingle	record linkage machine learning calibration best subset selection low back pain Medicine R Adrian Richter Julia Truthmann Jean-François Chenot Carsten Oliver Schmidt Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
description	(1) Background: Predicting chronic low back pain (LBP) is of clinical and economic interest as LBP leads to disabilities and health service utilization. This study aims to build a competitive and interpretable prediction model; (2) Methods: We used clinical and claims data of 3837 participants of a population-based cohort study to predict future LBP consultations (ICD-10: M40.XX-M54.XX). Best subset selection (BSS) was applied in repeated random samples of training data (75% of data); scoring rules were used to identify the best subset of predictors. The rediction accuracy of BSS was compared to <i>randomforest</i> and <i>support vector machines</i> (SVM) in the validation data (25% of data); (3) Results: The best subset comprised 16 out of 32 predictors. Previous occurrence of LBP increased the odds for future LBP consultations (odds ratio (OR) 6.91 [5.05; 9.45]), while concomitant diseases reduced the odds (1 vs. 0, OR: 0.74 [0.57; 0.98], >1 vs. 0: 0.37 [0.21; 0.67]). The area-under-curve (AUC) of BSS was acceptable (0.78 [0.74; 0.82]) and comparable with <i>SVM</i> (0.78 [0.74; 0.82]) and <i>randomforest</i> (0.79 [0.75; 0.83]); (4) Conclusions: Regarding prediction accuracy, BSS has been considered competitive with established machine-learning approaches. Nonetheless, considerable misclassification is inherent and further refinements are required to improve predictions.
format	article
author	Adrian Richter Julia Truthmann Jean-François Chenot Carsten Oliver Schmidt
author_facet	Adrian Richter Julia Truthmann Jean-François Chenot Carsten Oliver Schmidt
author_sort	Adrian Richter
title	Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
title_short	Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
title_full	Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
title_fullStr	Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
title_full_unstemmed	Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach
title_sort	predicting physician consultations for low back pain using claims data and population-based cohort data—an interpretable machine learning approach
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/db524eadd30046b9a458640c2de48848
work_keys_str_mv	AT adrianrichter predictingphysicianconsultationsforlowbackpainusingclaimsdataandpopulationbasedcohortdataaninterpretablemachinelearningapproach AT juliatruthmann predictingphysicianconsultationsforlowbackpainusingclaimsdataandpopulationbasedcohortdataaninterpretablemachinelearningapproach AT jeanfrancoischenot predictingphysicianconsultationsforlowbackpainusingclaimsdataandpopulationbasedcohortdataaninterpretablemachinelearningapproach AT carstenoliverschmidt predictingphysicianconsultationsforlowbackpainusingclaimsdataandpopulationbasedcohortdataaninterpretablemachinelearningapproach
_version_	1718411953365319680

Predicting Physician Consultations for Low Back Pain Using Claims Data and Population-Based Cohort Data—An Interpretable Machine Learning Approach

Ejemplares similares